not sure but in general our 2022 solar panels and inverters show such string behaviour that 1 partly shadowed module kills the production from about 2500 Watt to 250 Watt and then 30 minutes later the same for the next string which is twice as big from 4500 W to 450 W caused by leaf tree. Each day the same time this can be observed where we moved from 1 bing string inverter to 3 smaller one phase and double string inverters which lead to 6 strings and therefore a much longer usage time of the full power . Our roof is now split into 6 strings instead 1 or 2 and the benefits are significant if you check the inverters string by string and not the inverters in general even though that will show the same effect . If you add a timelapse camera you can mark the lines of the shadow when the shadow is too big for the diodes and the power will be lost. I am not sure about the optimizers but the 4 different Growatt inverters we have had in the past 3 years for tests (like SPH series, 3000-S String, Mic 3000, Min 3000 and Mic 4600). But they also come at a price, which means the price of the purchase of 3 smaller versus 1 bigger but also the lower production cause 6 strings instead of 1 big one means lower string voltages and too low voltages means that the inverter will start later than the 6 times bigger one. 3 Watt is not enough for a 4,6 kW inverter, but 18 Watt sure. Not sure what the shadow optimizer inverter will cost or how big the benfits will be they can gain but it is not that easy. I had played a lot with the simulators like that pv sol premium which Valentin Software offers for free for 30 days but I was able to gain more and more Watt hours by the right inverter size and amount strings and of cause the sting cabling cause our strings are build up vertical cause a tree is spending shadow from 4 pm in the evening which is now not so much wanted than 30 years ago. But it took quite a long time to find solutions for each roof, one with the chimney and the tree shadow and a lower angle while the other has a steeper angle but also its own issues caused by a GAUBE of 7 m widthspan . Gaube looks nice but 22 years later more or less useless nowadays cause a 7 m long segment of the south roof is lost and also causing shadows in the morning and the evening on one or the other side. Hope to see the update video soon, but I guess it will take till spring. Meanwhile we will simply add more panels and another inverter which will help a lot to get the 20 kWp in Spring and maybe in 2 years the 30 kWp as maximum using the roof area better with small microinverters around the 2 chimney and lower areas where the greenhouse is which has a steel roof too. Gets a lot of shadow but 5,5 m width and 2 m depth will offer a spot for 5 big 525 Wp panels that can produce during summer full till 4 pm before autumn and winter will start the time of long shadows no one wants. 3 kWp here and 4 kWp there can make 5000 kWh even in shady situations if the microinverters can proctect these modules close to the chimneys. The Min series from growatt with 2 mppt has helped us a lot like the Min 4600 we could get for 600€ , but the smaller ones like the 1000 W inverters are getting expensive about 350€. Unfortunatelly the cheap chinese GMI inverters or the SGB or many others we have all tested are not really working fully and throttle caused by overheating. Only the more expensive western ones get it done right but at a higher price, lot higher price of about 360€ per 500W - but at least without throttling. Maybe hoymiles might be able to deliver a better performance and having the EU certificates like VDE AR 4015
@@typxxilps Your situation seems to be quite complex! I do not know the products you use. But the inverters must have built-in shadow management. Otherwise, they can react completely wrong (as it is the case at the beginning of your comment). If you need more and more small inverters, optimizers or microinverters might be the better choice, at least for the shortest strings. But still: If a shaded panel does not produce energy, an optimizer will not fix it.
Hi, Andreas! It's been sad for me to see that you don't number you videos any more. It is a killer feature of your channel! Numbers make it so easy to navigate through your vids and I go through them a lot! Let's say I need the video about A6 GSM module, I would simply type 66 in your channel's search and here I have it. I hope you consider keeping on with this tradition! Cheers from Prague =)
Really useful! Thank you so much Andreas for thoroughly researching this topic, performing real-world experiments and explaining your conclusions so clearly.
Hi Andreas, Nice video. I did my PhD on optimizers and PV shading and my conclusions were similar to yours. We also did an article on how to maximize power production of a PV plant by interconnecting the strings in parallel. Passive optimization works about as well as an active one.
Interesting paper. Maybe one day they add a few transistors inside the panels to switch each cell accordingly. Because with the current fixed bypass diodes, most of the loss comes from the panel. Which is also lost for the optimizer...
@@AndreasSpiess Thank you for the reply. The bypass diodes are an "economic compromise". When one cell goes on the negative region, its voltage is around -17 times that of the other cells. When adding them up, this makes for a negative voltage that activates the bypass diode. This means that 18 (or around that) are the maximum number of cells you can still protect with a single bypass diode. If I'm not mistaken, space applications put one bypass diode per cell, to avoid this problem and maximize power production. I suppose it would make the PV modules too expensive to do so. :P Transistors would be too complicated.
could you help answer a (hopefully very simple for you) question - adding several higher power (much newer) panels in series with the old ones I should get less than rated from the new addition, but something proportional to the new panel voltage * old panel current right? thank you
@@tjunkieu2b Thank you for your question. I will try my best to answer it. I think you are right: as PV panels age, they produce less current (but not that much less). If you connect a newer panel in series with an old one, they will have the same current for a given point of operation, but they might have different voltages. Would you mind explaining a little more? Are you retrofitting stand alone/DC grid PV for grid injection?
@tjunkie Yes, but there are issues with mixed modules in a string, and if at all possible one should not do so, or you can significantly shorten the life of the lower power modules as they will be possibly passing more current through them than they are designed for. Best to have the same modules in each series string.
Micro inverters are maybe not useful in large installations, but in small ones they are a game changer. Not because of efficiency but by pure existence. With micro inverters small installations are easy to install, even by amateurs. Thanks to Micro Inverters balcony power plant are now very common and it will become more and more.
Absolutely. Cost of microinverters for a 3 panel system will be less than a string inverter. In addition, it is easy to install. But if you are going for above 1.5kW power like 3kW or more, they become expensive
This video is why I'm subscribed, real life experiments, a bit of theory to go along with it, excellent stuff! It's a shame so many installers have no clue how this works, not to mention installers that get bigger profit from installing micro inverters.
You and I are here for the same thing. I frequently encourage Andreas to focus on real world projects because, building stuff well is often remarkably difficult. I also keep encouraging him to use MicroPython on ESP32 instead of C/C++ ESP32 because MicroPython is both quicker an easier to work with, yet often just a performant-from the user's perspective-for many, many projects.
@@AndreasSpiess that would explain a thing or two ;) then again installers have access to hardware and putting up a small system to test shading wouldn't take them long but I guess many of them don't care that much. Luckily we have people like you that are willing to spend the time to show and explain how things work without marketing BS and I'm trying to do my part explaining "basic" electrical stuff to my friends.
Thank you very much, Andreas! We got our PV system installed back in august, and because of your "warning" in one of your earlier videos, I made the last minute decision to cancel the already ordered optimizers - not so much because of their additional cost (I still would have loved to monitor the performance of each module), but more because of the chance they could fail one by one over the years, causing a lot of cost to replace or remove them, especially since they are installed at a hard to reach place. Better to have a reliable, though less performant system for many years, than a little bit more performance with the risk of more problems later on - these modules shall stay untouched on that roof for the next 30 years, if possible. (And saving an extra 900 euros on the invest did not hurt, either... :-)
I added a simple "divided by the number of panels" in my Grafana board. Now I can compare the cells and see if one creates problems. The only thing is that I do not know which one it is ;-)
Pretty much everything I've read and watched about optimizers says they are a waste or worse in most situations. Unfortunately, since the US NEC 2017 & 2020 electric codes require Rapid Shutdown on a per panel basis, the math changes a bit. If you already have to put an additional RSD module on a panel, maybe putting on optimizers with RSD and monitoring instead makes more sense since you are no longer adding more connections or modules in the chain.
No one in the US seems to make (or admit) mistakes these days. You don't learn without making mistakes or failing at tasks... so, why not admit them. Love that you did just that.
Personal experience: I pay ~$50 for an SolarEdge S440-1GM4MRM with 99.5% efficiency (as per its data sheet). It's not "just" about increasing the overall performance of the installation, you debunked that quite nicely, but also about measuring individual panel performance. Helps with identifying misbehaving ones.
Agree. On my installation, one if the panels failed after three months. It was easy to detect as you get power per panel. Our system has 3.6 kWp facing South and 1.5 kWp facing West (both different inclined roof), on one giant string. Hard to say whether this works better than two independent MPPT.
Exactly, this line of “optimisers are all about shade” is typical of string manufacturers. It totally misdirects from the many other benefits of MLPE like many safety features, panel level visibility, simpler fault detection and repair work, higher efficiency, dc coupled storage, the list goes on. Making it all about shade, which is a confusing topic and difficult to properly compare, makes it simpler to push the traditional dangerous string inverters with uncontrollable high voltage dc.
@@BrentHeberAusNZ Ooh, dangerous string inverters! It's still the same amount of kilowatts coming down the string, which will still kill you, whether DC or AC. Who needs panel level visibility? What fault detection. Solar panels almost NEVER fail nowadays.Higher efficiency? Any proof of this?
Excellent and eye opening video. I was never fully convinced to micro-inverter idea and it is really nice to see that PV in strings are in fact more robust than the common tale says. Thank you so much!
Great video André. I thought i knew a lot about solar panels and optimizers, but this is actually a big thing. Thanks for adressing it. I think more tests and theory explanation should be done in the following years. Perhaps the days of the optimizers are counted.. (i myself have 32x 310Wp solar panels, in the same direction, without any shadow and sadly enough all with optimizers)
I currently test optimizers in an arrangement of four panels directing towards south and in a steep angle. I plan to add another 4 panels in the same direction but a different angle. 4 panels are not enough for a string for the Huawei (not enough starting voltage. So I might end up with optimizers in this configuration. At least I will test them.
Finally someone else who did the calculations. All the micro inverter salesmen have been very active lately. I always wonder why they lie about the non-micro inverter panels. Expensive and useless, microinverters are dumb. Also why hang them on the back of a solar panel, where temps often are 65°C. On flat roofs there is access, but on slanted roofs it's a mess if you have to replace a microinverter.
Great observations/tests to make us think about what to install Andreas. Hope you will go more into the use of the battery in the near future. Just my n=1 anecdote: I had a 110cm satellite dish on my flat roof that is North-South (about 165 degrees). My PV installation is 10 years old, 1 string, facing almost perfectly south. At just after noon there nearly always (on cloudless days) was a huge drop in my generated power for 45-60 minutes. Power "lost" was about 1 kWh which is about 10% of the daily power generated during summer. In the other seasons, due to high latitude, there is almost no significant loss. The dish was out of alignement, but I cannot easily access my roof to either fix of remove. When my neighbor installed a PV installation, I asked the installers to remove my dish, not interested in fixing it not watching TV anymore. Since then the drop is gone. Power generated is a nice sine like figure. The dish cast a shadow on 1 of the panels apparently. Don't know the make or manufacturer of my panels, but I guess they would have benefited from optimizers. But I have a feeling it wouldn't make economic sense to have had them installed.
The forecast of the „optimizer guys in your case would be that you lose the power of the whole string. So if you lost all power during the shadow phase, the chance is that your panels do not have bypass diodes…
Really excellent and hard to find information. There's lots of bad information on the internet (as always). Thanks Andreas! I have been agonizing over whether I made a mistake in not installing the micro-inverters/optimizers. Now I can finally tell my wife WE did not screw up!
I also forwarded your video to engineer and architect friends who are working with solar. You are the hero to all of us. Thanks again for your hard work.
Solar power is really big here in Australia for obvious reasons but most of the installers are dodgy and even electricians don't understand it. I'm about to put a solar plant on my roof so this is very useful.
Thank you so much! I have two small panels in parallel on my boat, and a very simple (and cheap) regulator. I tried an MPPV system, and it constantly drained 20mA from the battery for its processor, day or night, without providing any obvious benefit. I can't afford 20mA for 24h from my boat battery. I thought I must be doing something wrong. But now I feel I need not worry at all!
20mA indeed is not a lot for solar power systems where we try to produce MWh. But in your case, maybe it would be of benefit to find a space for an additional panel. This often is the better way.
Now I understand what the shade management algorithm is doing - a more global optimization in the MPPT controller. Older inverters don't have this and older panels don't have they half-cut and bypass diodes, so things are improving!
Really interesting video ! I have PV solar on my roof for over 14 years now ! In those days, it was rather expensive. But anyway, I'm glad we did it. Now, thanks to your information, I'm preparing a second installation on the roof of my workshop. A work in progress... but defenately very satisfying !
It's so interesting to see the progression in the PV world. I also started in 2006, with three Sharp 165Wp panels on a Mastervolt Soladin 600 inverter. That system would seem like a toy nowadays.
@@SolarWebsite Mastervolt is a really good brand. We have 10 panels for a total of 2.2 KW with an Infinity inverter. This brand does no longer exist, but it's quite good, so far, never had any issues.
It would be nice if some research could be done into modelling roof situations to better decide if optimizers/microinverters would be a good solution for special situations. I think the vast majority of PV installations don’t need optimizers, but there still are a some roofs that might get better yield with them. Whether they are cost effective is another discussion, but to get the best yield sometimes they are needed in my opinion.
This is exactly what I concluded before installing my panels... but I have roof-surface facing all sides so optimizers seemed a reasonable solution to have west/south/east in a single string. An alternative would have been a three string inverter and connecting them up separately, but that also creates some problems in terms of wire mess (and these inverters are pretty expensive). Another possibility would be micro-inverters to east and west, and a simple string south... but again wire mess and on top of that mixed solutions. The price differences for different solutions were not big enough to be the deciding factor. My main worry is indeed that an optimizer might break, it will be easy to find it but hard to access. 25 years warranty does give some confidence, but the price of the optimizer is not my main concern obviously - loss of production and man-hours to fix it are not covered.
@@AndreasSpiess And your work is appreciated. So many options out there gives a lot of space for installers to 'peddle' expensive solutions - information is key!
@QPW Chris Indeed, the implementation of any warranty beyond 5 years is always a bit of a question mark. My house is about 17 years old now and some things are starting to break (as expected)... good luck finding *new* parts to repair anything!
@@diatonicdelirium1743 Yup, older parts are very difficult to replace. It is always better to buy standardised unbranded/generic items which will remain standard for long time than buy fancy branded items which keep getting upgraded!
Thanks Andreas - another fantastic, clear video. I have optimisers fitted in my "professionally" installed system - seems like I fell for the bullshit! One good point about the optimisers (although possibly not making them worth while) - they do let you know the individual output of each panel. Like your brother, I also lost most of the summer's production due to slow delivery of the SolarEdge inverter. I wish I had considered my installation after you, and could have taken advantage of your superb research!!
as someone new to solar and trying to get up to speed you have helped me a great deal. All our panels are on the same southern facing plane of our roof and therefore microinverters are unnecessary but yet the company doing the install has used them. We will be questioning their usage and likely asking for their removal since they are ineffective and expensive. And requesting the appropriate price reduction according to their cost. Our situation is something of a nightmare, a company that takes advantage of its customers by being greedy, and overly accepting of what the equipment providers ad copy says. I dont buy half of it, and am becoming increasingly antagonistic towards my installer for that very reason. If we don't end up in court it will be a minor miracle. Thank you again for your common sense approach to this. I am very grateful.
Many installers are also "cheated" by the suppliers. Keep in mind: This is a growing market with many new entrants without deep knowledge. I would inform your installer, not punish him.
Great experiment, I had long thought "but what about the bypass diodes?" to these claims. BTW, I have an Enphase setup, not because I was sucked in, but because my panels are at different angles and alignments, 3 vertical, 5 South ,6 East and 6 West, and that would mean 4 strings, and probably more than 1 inverter... Instead I have no big central inverter at all, which is big bonus to me.
A fantastic video! The best solar optimizer is to take cash used to buy optimizers and buy 30% more solar panels! Being a fellow amateur radio operator, the last thing you want are a bunch of poorly made switching supplies on your roof connected by a long antenna.
Hi Andreas. When I did my research regarding my installation, I came accross an online huawei simulation tool. It allowed me to choose, panels, cables, etc. It also provided the option of adding the huawei optimizers. What I noticed in the simulation charts, is that the inverter started to work sooner in the winter with optimizers. As I understand, an inverter needs a treshold current before it starts? The optimizer would lower the voltage to increase current in order for the inverter to start. This would be interesting in overcast weather. I asked solar installers for their experience, but no one could help me verify the simulation claim of huawei. I also didnt find anything online about this particular scenario. Maybe your LTSpice simulation can verify this? If my simulation charts where correct, the optimizer was a good investment over a couple of years. Regards,Wim
Hi Wim, What were the numbers they told you in the charts? The inverters I know start at 100 mA or below, so that would be at a power of 50 W (at 500 V). You would have the same amount of loss in the optimizers at 2500 W (2% loss). It really depends on the inverter and specific optimizer, but it doesn't convince me... I also haven't seen any optimizer manufacturer advertising this as a big advantage of their product (I work in the PV industry, also with optimizers and rapid shutdown devices) Cheers
You might pick up a few minutes at the start and end of a day, where you are making maybe 3% of peak power.... you're still better off taking your cash and getting another solar panel.
The open voltage of the panels increases extremely fast and reaches the needed minimum voltage. And, as James mentioned, I do not care about a few percent of a kW when it only costs a few cents. One panel more gives you much more bang for the buck…
My solaredge system was installed 7 years ago in a single string split orientation system. 10 west 10 east. As i have nothing to compare it with i cannot know if I fell for the sales pitch. One benefit of the optimizer is that you can see when a panel in a string is not working and get it replaced under warranty. Not sure I would have noticed that one pannel in 20 was only giving about 50% without being able to see the output of each individually. Good video as always.
I agree in your case. So I assume the panel was defective and not the inverter. What was the issue? Maybe you would have discovered that the string is no more up to date. But for sure you would have to find the panel manually. I am also glad to read that your optimizers still work. An important information!
@@AndreasSpiess I don't know what was wrong with the panel, the supplier turned up with a new one and swapped it out. All panels look about the same still. The system output does not appear to be reducing over time, just fluctuates from year to year. Keep up the great work.
I agree that optimizers aren't really worth it compared to bypass diodes. But I do like the safety aspect of micro inverters. High voltage DC can cause arcs that do not extinguish. It is one of the major culprits in residential solar fires.
Thank you for this contribution to such a controversal topic. The YT channel MC Electrical has 2 vids "Fronius vs. Enphase" ca. one year ago with real measurements. Similar did "NRG Solar - National Renewable Group Australia" laso ca. one year ago. Some percent in difference in practical cases. I also simulated different parallel vs. serial configurations for older 200W modules using LT-spice. Results were also not soo straight forward. My 36 PV modules are on the roof now, but I also miss the Fronius und the BYD battery. Maybe next year... we will see. And the switch box for backup island operation has no delivery date up to now. Here we call it DDR 2.0. And do not refer to RAMs. ;-)
Thank you for the hint about the Australian video. I thought I had it included in my comment, but I forgot it. Now it is there. And good luck with your inverter and battery...
What I have read, optimizer is a good choice if you want monitor status of every single of solar module. It also has safety feature that reduce DC voltage to touch-safe levels and arc fault detection.
You are right. Keep in mind that after a few weeks, most users only check the power every month or so. And only for the whole plant and a whole day... The risk of having electronics in the heat and humidity of a roof stays forever...
At my latitude (42N) I found that South facing and tilted at 60 to 65 degrees was best for winter production while having the least effect on the total yearly output.
My biggest "a-ha" moment in this video was the "shadow-management" local-vs.-global optimum-seeking explained around 13:25 onwards. Once again, thanks, Andreas!
@@AndreasSpiess Dave from EEVblog did a video about a similar topic, ruclips.net/video/AbxHoQF4ADk/видео.html - i suspect his older inverter was not equipped with global-MPPT.
Maybe we should have all panels in parallel and use very thick cable or bars to minimize loss and the mppt as close to the panels as possible. What a test, hats of to you Andreas, taking speculation away and making it a real life test with the correct knowledge and measurements.
Having them in parallel isn't perfect either. A shaded solar panel can become a current sink without a blocking diode. Although counterintuitive because a solar panel is a diode, it's only a very weak one and reverse current can cause it to fail. Moreover, the thick cables will cost a lot more than having a mini inverter behind the panel and using classic 230 cabling.
@@matthewmaxwell-burton4549 I agree and value the response! I have a mix of different 18V max mppt panels just for testing. I actually wire them up in parallel with a Schottky diode in series of each panel. I know not 100% efficient but there is a best approach which each situation.
With small panels, maybe more arrangements are possible. If you work with higher power you have to take the capabilities of the inverter into consideration. I like if my inverter and battery use high voltages because it generally reduces loss and reduces the needed copper.
Thanks for the thorough testing. So the claim "one panel shaded halts all output of the entire string" began with earlier panels and now won't die. Glad mine are the latest Longi Hi-Mo. The spec sheet states "Half-Cell" but doesn't detail the implications or discuss shading. Glad you explained it. Seems I lucked out in my craigslist purchase. I thought my panels didn't output much until 10 am in Summer since one panel in each string of 7 is mostly shaded before that, so exploring optimizers. After your tests, I suspect more likely just due to the sun angle since my panels at tilted slightly to the southwest.
In my case I have 3 panels pointed east- (slightly north east), and 5 pointed west (slightly south west), all on a 45 degrees roof. 2 pointed west get an early shadow. For that reason I choose microinverters. A solar-edge system wouldn't have worked in my case, 3 panels wouldn't produce enough voltage for the inverter to work properly. They say Solaredge has a minimum stringlength of 8 panels or so. I only have 8 panels, but often only 3 of them get full sunlight. For that reason I choose microinverters. It made the installation a bit more expensive, but I'm still happy with it, especially now the energy prices went up so much. The nice thing about mictoinverters or (solaredge)optimizers is that you have per panel monitoring. For anyone interrested: the panel pointed east made 267kwh so-far this year, the ones pointed west 340 to 380.
Thank you, what an eyeopener. I would only want to point out that with the optimizers or microinverters you can monitor the power output per panel and that is a nice feature if you have to do troubleshooting in the future. Only you need to know the layout of your panels and which is which or you still don''t have any usefulness. But on the other hand, you really have a good point about the active electronics in the harsh environment on the roof during the hot summer sun. What is more likely to break anyway...
The monitoring is a nice feature but not needed. Even without it, the panels will function the same. Even if you manage to find out that 1 out of 15 panels is defective, the cost of optimiser would have been much more than the cost you saved by getting 1 panel replaced
What would I have to add to my string inverter (LUNA2000) to be compliant? Here, I had to add nothing. edit: I just found information: It seems you have additional specifications in the US. So you are right. Fronius sells such a RAPID SHUTDOWN box for 160 dollars. And it is needed for each string.
@@AndreasSpiess I think sometimes it's each panel if roof mounted,if ground mounted then don't need RS for each panel. I think Tigo brand is about $30 per panel for each shutdown module.
When you completely block a panel, the bypass diodes go to work to prevent blockage of the current. (All panels have bypass diodes) The value of optimizers is that they improve current flow during partial blockage. I use Tigo optimizers and they improve my partially shaded arrays power output by about 25%.
@@AndreasSpiess I put the optimizers on since I had a lot of (partial) shading at different times of the day. My "comparison" was before/after installing the optimizers and there is a clear boost in power. I like the Tigo optimizers since you don't have to install them on every panel, just the ones with shading. The software gives you a reading on what they call "reclaimed power" which is their calculation of the extra power "reclaimed" by the optimizers. For me, it's about 25% extra on my 4.5 kW array.
@@MarkSpohr That 25% is a value which is give by the company which may not be accurate. The actual benefit of optimisers is less than 10%. Remember that the optimiser themselves consume 2-3% power. Also, the shaded panels will activate the diodes regardless of optimisers as diodes are inbuilt to the panels. So, your 4.5kW system would perform the same as a 5kW system without optimiser. Unless the cost of optimiser was less than what is needed for 0.5kW additional panels, it is not worth it
@@MarkSpohr The data on what basis? Is its a theoretical data or based on actual difference with & without optimiser? The 10% is based on empirical evidence on how the diode works. The difference between the diode & optimiser is minimal and is in the margins
great video, thanks for bring this topic up and make it clear. it is ridiculous how some people mislead others in this field claiming incorrect result relying on unrealistic condition
Writing from the wet tropics of Australia: Yes I agree the use of optimizers is rarely worth the extra cost. Here we have issues with huge UV exposure and many systems get replaced after just a few years. Bypass diodes are a way to save infrastructure cost and our feed in credit is now so low it is hardly even worth having panels at all.
I had a hunch, and I'm glad a channel I know did the same test. I've come across a few videos saying the same thing, that micro-inverters while they technically do something, the cost/benefit might not be as significant thanks to diodes (or good placement).
Fantastic video -thank you for the useful insight! Some people have more money then sense - sadly some of these solar installers will believe the marketing hype and happy to accept the kick backs / commissions from these big vendors!
WOW! awesome video - I confirmed the exact same findings by trial and error when trouble shooting for a bad panel. In defense of the micro's and optimizers - older generations of solar panels probably did not have that pesky diode. But modern half cell solar panels just changed the game completely, they are amazing in dealing with shading. I wish there were more grass roots videos or communities like this, because most solar system installers have no clue about the intricacies of solar installations beyond wiring and city codes, hence the easy way out is to just simply offer a money making stop-gap measure. Good job, man!
What could really help is a system that can inform you, on an app, what cells are not getting unlight. Without that, you have to either go outside and check, or just guess, and it may not be that obvious whether you have one cell a leaf blown over (easily fixed), one that a bird crapped on (a bit harder), or a cloud drifting overhead, which you can't do anything about.
I divide the power by the number of strings. Then I see if something is wrong (which seems to be very rare). But if you want this feature, optimizers or microinverters are your solution.
Hi Andreas, I have a very important observation to the topic, where probably optimizers have an important role! I have installed a Huawei Sun2000-8KTL inverter with 2x9 Sharp 410 PV modules conencting to the 2 MPPT trackers. These PV panels are on a south directed 35° inclination roof in two parallel rows, where I have a tall tree in front. So like a solar clock, the shade wanders around. I was convinced by the installer that no optimizers are needed, along with an australian vlogger, who made tests on different mppt inverters. However! Today the sky is crystal clear, in Hungary, the heating is off, that it warms the house, but the 7.2kWpeak system gives only 0.6kW at 2x310V, compared to a foggy ugly diffuse day few days ago, when I had over 2kW. Also some other day, with clear sky, it happened, that in the morning I had 210V and 9 amps on one string, then jumping up to 310V and 1 A. So, my conclusion is, that if you totally block the light from given areas, then the diodes will bypass, and you get to an optimum, but if you have some partial light, casted by leaves on a tree, it will screw up the mppt, and give low currents at high voltages. I will do observations further, maybe you could test this feature, or we can discuss some experiments and data to see here more clear.
Thank you very much for this analysis. I always asked myself how the cells behave in such a situation, because this year I am planning to built a small solar balcony power plant with storage, where I can only install two panels in very different angles. So I will have very different sources with relatively low voltage and have to figure out the best converter to charge a battery so I can use the power when I am home, which is mainly at dark hours. Schöne Grüße aus München!
Dear Andreas, Hello from Geneva! I am so happy to have found your channel by accident, i am watching all your video and love them! Thank you for your work and very clear explanations. I really appreciate the fact they debunk the « sales speech » of certain manufacturers such as Enphase but also most video on RUclips still encouraging blindly to ise micro-inverters in any situations. (I noticed it especially in french video (in France they still encourage 2-3 kWp lower cost installation with expensive Enphase micro-inverters 😅) I just made up my mind today to go for a 14kWp 2 strings setup, i will get the Huawei SUN2000 hybrid inverter and get the battery in few years hopefully cheaper. Initially i ordered 6 optimizers (chimney shade) but i will now probably cancel them. Thank you!
Welcome aboard the channel! For small systems, microinverters might be a good solution because string inverters usually need a certain startup voltage. For your large system, a string inverter is a good choice.
Great video! I am planning on installing solar power on my roof, I have about 100 square meters of the roof facing south that can have panels. This gave me valuable information in deciding what to get. Since I'm also into Home Assistant and dabble in electronics, you are my perfect video guru😁. Best wishes from Sweden
That is a pretty big roof considering it is just one side! You can place 10kW panels easily on it Depending on latitude, climate and shading, you will get 30-50kWh/ day on average
I've been running 12 panels with SolarEdge for 7 years now. Last year I had an optimiser that stopped working. The main advantage of this system is that you can track the performance of your panels. I called the company that installed my system and they replaced the optimiser free of charge (Warranty). 12 panels should give me 3420 Wp but they outperform every single year. Got at least 3800kWh every year minimum. The only thing I don't like is that the main invertor is passively cooled. So I added some PC fans on a timer to cool down the backplate during the hot summer months.
Thank you for sharing your experience. You were happy that they replaced the device free of charge! Here they usually are not allowed to go to a roof just like that ;-)
You forgot two important advantages of optimizers: Safety and maintenance. Because we installed the system ourselves, we don‘t want to deal with 1000V on our cabling and risk light-arcs every time we unplug something. And we want to see defective panels without having to book a flight with a thermal imaging drone.
I don't know enough to understand how relevant your comments are. But it would be great if these points were addressed. Thank you for bringing them up.
My experience has been that the optimized systems also start producing earlier in the morning and keep going later in the evening, because the startup voltage of the system is lower.
He adressed maintenance. Don't put complicated electronics in a wet, cold and hot environment. The only thing you tend to monitor is a failing micro inverter, not a faulty panel. (solar panels are very robust, inverters are not).
@@AndreasSpiess that does not help. The full DC voltage is present on the whole string as long as there is sun. This is dangerous. And optimizers completely eliminate this risk and let only 1V per panel out.
Although microinverters contain an optimizer, I believe the real benefit they offer is redundancy. They connect in parallel. Thank you for the excellent video. I believe I have watched others of yours regarding antennas and vnas and thoroughly enjoyed those too.
Yes, microinverters contain an optimizer and they offer redundancy. The only thing I do not like is that they are exposed to extreme heat and humidity for 25 years. But this is a general remark...
@@AndreasSpiess As a relatively new enphase customer, the thing that bothers me the most is that I don't have full access to the very expensive equipment that I own. Some functionally is locked behind an installer login. Not a big fan of their marketing, either. As you have exposed in this video.
@@saltyexxer8253 I have a "balcony power plant with two panels and two Enphase inverters for a test. I just added a simple Sonoff to know what they produce. Sufficient for me. If it is only half, one is dead ;-)
The company I work for installed optimizers before. Now we don't except in very extreme cases. If we have a huge tree or different directions, we recommend it. On standard flat roof, not recommended.
Thank you for sharing! For extreme cases they are ok, I agree. With the tests of the video people should be able to see where these extreme situations are and where not.
Great video, as always! One thing that I do like about optimizers: in case of a fire, when you turn off output, the optimizers will drop the voltage on roof so there is less chance of electrical shock to firefighters. I didn't know about bypass diodes, makes sense the manufacturers implemented that :)
Yeah its a good idea on paper, but an incorrect installation, a factory defect, or like in most cases a bad connector tend to start these fires. Connectors failing are one of the leading causes of fires and when adding optimizers you triple the amount of connectors in a system. Not to mention that not all systems are equipped with such devices and they can fail closed leaving the system still energized. So in the end you have to train for the worst case scenario and treat each as it is live. The real solution would be to integrate into the panel but I don't see module manufacturers wanting to do that.
I do not know how dangerous it is for firefighters. As always, you have to hold both ends of a "battery" in your hands to experience a shock. Reading through recommendations for firefighters in this case, they list various things. None of them related to be electrocuted by the panels.
Have you checked out 'ideal diodes'? MOSFET configured as a diode. Allows bypassing solar cells without the diode drop if the cell isn't producing as much as the rest. Small and inexpensive. Could be used on each panel individually for maximum efficiency.
@@AndreasSpiess After reading the datasheet for MP6914 (an example ideal diode for solar panel bypass IC), I'm thinking the reluctance for using this circuitry is that it appears to require a separate power source. I was thinking they worked solely on the power provided by the solar cells (supposedly making them a simple drop in replacement for the bypass diode). This would complicate wiring if that is the case making them a pain to integrate. They also show it used once per solar panel string/group rather than each panel...which reduces its overall benefit over just a single cheap low voltage drop bypass diode. If bypass diodes were being used per panel I could see more benefits of replacing those with ideal diodes (especially if many panels get regularly shaded from moving clouds or foliage).
Optimizers are great for the company that installed your panels, if one of your panels get broken you can see that in de app / on line. So the install company will save time figuring out what is wrong.
Thank you for this analysis, I’ve been sceptical of such micro inverters, they probably made more sense on older panel technology, now I now what to look at for my system, need to Google shadow management now on inverters 😊
Sackstarch... Ich hoffe du lehrst an einer berufsschule oder fh eltech. Du erklärst sauber und extrem verständlich. Habe viel aus deinen früheren videos mitgenommen. Danke das du dein wissen teilst.
The bye pass diodes will route the current around any panel not performing as well as the others this has been my take on the situation. What I did not see any reference too. Was the case of using mismatched panels ie units of diferant outputs where the use of the optimiser will allow these to all add Energey to the total . I personally use two panels connected in parallel to one optimiser each panel is 300 w one facing east and one facing west at quite a steep angle. I have 8 panels set up this way these are to catch morning and late evening sun. Especially in the winter when sun is lower in the Uk . During the mid day sun both panels are producing but well below there optimum so still ad to the total . The idea is to flatten and widen the hump and make the power more usable . I also find the feed back from each optimiser hand to see what each panel is producing. Again this shows that all panels point same direction but they do not all produce the same kw over the same time period . Again a lot of this could be done utilising the blocking diode but each diode that operates adds a volt drop in the feed that costs power also . Interesting vid
Connecting two panels with different directions to the same MPPT creates a difficult problem. The best voltage for one panel is very different from the other. So, which voltage is best? It can only chose one voltage if the panels are in parallel.
@@AndreasSpiess hi yes that is true . But in practice I find that as they are east and west facing panels. In the morning the the east panel is seeing the sun and provides power as the sun comes round the east panel starts to reduce its output while the west panel output starts to rise as the sun starts to shine from the west till sundown. At about midday I have current clamped both panels and they both are supplying some current for a short duration until the west panel output is rising and taking the lead . For me it just saves the cost off a optimiser on a panel that will spend 1/2 of the day out of direct sunlight . I now have some more optimisers so out of interest I may just try adding an extra one and see if the return is a lot better
Installer and service guy here. SolarEdge breaks. ALL. THE. TIME. Save your $$ and get an SMA inverter and run your panels straight to each other. If you have issues and absolutely HAVE to have system optimization, get Enphase microinverters. If you're gonna spend the money on SolarEdge, just go all the way and get Enphase. I have about 12 service calls in queue here in central Texas, almost all SolarEdge.
Very interesting! I wanted a low voltage DC system for a water heater and some 135W half size panels. After fiddling with designs I realised it was more efficient / simpler to have separate optimiser / boost converters on each pannel - with high current / low voltage losses mount up very quickly. It seems to work quite well actually.
@@AndreasSpiess Yes sorry I didn't want to imply that! One thing occurred to me though.. for most applications isn't the selling point of the optimiser to convert directly to 230V AC at each panel? Not applicable in your own case, but I could then understand that it would shift and break up total losses from a central inverter to each optimiser so not quite as bad!
Andreas, great Video. Thank you. As you pointed out, there are situations where optimizers might be worth it: * Assume a situation where a panel is largely shaded. In this situation, the shottky diodes would bypass the whole panel. This woul not result in the whole string "tied down", but still, this one panel does not produce power anymore. Whith an optimizer, this Panel could still deliver it's small part to the system * Solaredge Optimizers allow to build very long strings (up to ~11.5kW within a single string), and you kind of don't need to care about elecrical Specs (VOC of the String in cold conditions or the like) or direction of panels. This might facilitate cabling a lot - you need to factor this in for the investment decision * Solaredge inverters tend to be cheaper than others as they don't need to perform the MPPT anymore - they operate at a constant voltage. I would be interested in how you "controlled" the optimizer in your test setup? To my knwoledge, you need some Solaredge device to tell the optimizer what current it should produce. Best regards
I did not want to condemn optimizers! I just wanted to show a false claim. In certain cases, they are good. However, if the bypass diodes kick-in, also the optimizers do not get energy from this part of the string... I cannot control the optimizers. It is done by the inverter.
The diodes are inbuilt to the panel. If the panel is largely shaded, the diodes will be activated even before optimiser can do anything. Even if the optimiser manages to salvage some power, it will be in few tens of watts which is miniscule when comparing panels of 300-500W. Keeping 10kW in a single string is a serious risk of safety and highly expensive due to high amount of current flowing. Even if you can do it, you must avoid doing it.
Extremely interesting this post of yours. I guess that common mortals like myself trust it when they say "You have to have it" but now I ask ... do I really? Thanks to, at least, planted the bug and make me research this more deeply.
Another great video by you! Thank you for all this relevant information that helps me now know what to expect when this situation arrises and now what to worry about!
The optim. very good, if You have to increase your panels, but you cannot get same watt . If your grid has 410 and next year you cannot buy just 450.. in case is very usefull. Also if you have a difficult shape of roof, with shading.. optimisers can win 10% of production.. and panel life as well. A simple roof, simple and nice situation you dont need.
Another reason towards micro inverters is, that the system gets slightly more reliable. When something like a Strom destroys parts of your array, your prodution isn't gone completely. Also a failing inverter is not as much of a problem with bigger setups. And the components are less expensive, as the requirements are lower ;)
I have had Enphase for a while and one thing I noticed is that with more microinverters the slower the response time on loads. When I only had 10 it would respond fairly quickly when the A/C turned on/off. When I upgraded to 24 there was a noticeable lag on both turn off and on.
@@AndreasSpiess The power line control interface also interferes with my old X10 controlled lights, is slightly annoying. But I hear there are solutions.
Interesting video! I recently had my PV installed with a SolarEdge inverter and 2 sets of 9x 370W solar panels with optimizers. I certainly did have to wait quite a while a while until the inverter for SolarEdge was available. I choose optimizers because 1 set is on the roof of my house and the other is on the roof of my "shed"/parking spot. I also thought I would need the optimizers due to potential shadows of trees a few moments during the day, I guess that was less true than I thought. Don't know exactly how much I paid for each component, but it wasn't necessarily cheap and let it install by professionals (they already came back to sort out an issue).
Run DC at home: Kettle, TV, computer and anything with universal motors or switchmode power supplies works just fine on DC You might only need to throw away your AC clock radio.
Thanks for the video. I hadn't previously considered that MPPT may not track to optimum due to shading. Perhaps because I just expect them to work as expected! Those bypass diodes are certainly worth their weight when used and sized properly.
There is one thing to consider, and that is partial shading of a single cell. My panels have 3 anti-parallel diodes, one diode for every 24 cells. If one cell gets partially shaded, it is reverse biased by the other cells in that series. That can lead to localized heating(often referred to as "hot spotting") of the surface of the cell still exposed to the sun. That permanently degrades the performance of that cell. A MPPT that does not globally scan, means that the string only produces the current of the weakest cell(the one partially shaded) and you lose a lot of output, but, you don't damage the cell/panel. Clearing the leaf, or the bird deposit is all that's necessary to restore full rated power of that panel. The ideal solution, would be a short barrier Schottky diode anti-parallel to each cell. That one cell gets by-passed by its diode, and you lose only ~5 watts of power in that whole string, and the cell is protected, until the panel gets cleaned. I agree 100%. If all the panels in a string are mounted on the same plane, the losses of the DC optimizers is greater than the differences in the outputs of the panels. Even as they age, I don't think the differences will be anywhere near that 2% used up by the extra electronic modules. They cannot compensate for the damage of the hot-spotted cell caused by the partial shading. the only hope is, to limit the current load on the panel, protecting the cell, but, instead of the whole string being de-rated, you only lose the rating of the one affected panel.
Optimizers have a safety advantage , like safe DC from solaredge where it cuts the voltage to 1volt when the string is broken / disconnected. In comparison to a string situation where the current/voltage keeps flowing.
My cables are insulated. So I do not see a big safety advantage. To be electrocuted you need a defective plus and a defective minus pole and touch both at the same time. Still, it is DC which is much less harmful than AC.
What Attix15 means i think is that when there is a damaged cable internal or exteranal there forms a ark because of the high voltage and a optimizor stops this. And at high voltage DC i would not say it is much less harmfull then AC.
@@kwynton8272 Most modern inverters have arcing prevention. And I did not read of a lot of accidents because of arcing. AC is more harmful to your heart because it interacts with nerve signals and can make your heart stop. Otherwise, the effects are the same.
@@AndreasSpiess Andreas how is it with the Firefighters in Switzerland? Here they have "issues" with extinguishing something with more then 400V and will not permit above 800V.
If you cut the cable between two solar panels and held both ends of the cable, then you’ll have a problem. But consider a few things: 1. You haven’t shutdown the solaredge inverter first, so there’s still a circuit voltage for a few seconds after a circuit interruption (400V capacitors inside the inverter). 2. If you isolate first at a string inverter, while your solar panels are still generating voltage, there’s no current because the circuit is interrupted on both poles. So you can’t get a shock if you insert your body into the circuit between two panels. 3. Solar Panels are generally not earthed, so there’s no return path for any circuit current via earth back to the inverter if your insert your body into the earth fault circuit. Dry rooftops at least don’t offer a good path to earth either so chances of suffering a shock due to your body being on the path of an earth fault are minimal. 3. Opening a connection via an MC4 connector is very safe. They are very well designed connectors leaving no naked conductors that can be reached with your fingers. The most important intervention to make before going onto the roof to address your solar panels is to isolate the dc string first.
Very interesting and thanks for Sharing your results. Bypass diode is the magic word here. In combination with an advanced Inverter (shadow Management). I worked in the PV industry some years back and experienced the rise of these companies. We had cooperations with these companies and also investigated what you showed. We had large Installations where we did what you investigated. I agree that it is not worth the money for a private small PV construction. It is more important to work with an expert to plan and Install the PV modules. Optimize to get the most of it is not trivial and needs good expert. Microinverters have their niche but is not the general solution like you said and showed.
Been looking at getting solar installed, all the panels are going to be same side facing can get a SMA 3600 string inverter for good price and 8x550w ish panels for over sizing as I am in the UK I don't expect the inverter to clip often even with 4.4kw total solar panels (they likely to be doing less then 400w per panel as it's the uk, so I am considing 10 panels as that brings the peak voltage the solar panels can make to 440v DC witch is maximum mpp but still way below 540v max DC) was looking at solar edge or enphase but adds £700+1300 inverter to £1500 (enphase) but as it seems it just add per panel monitoring really I could get tigo TS4 power optimizers at a later date if I wanted monitoring, as they work with any normal string inverter systems as they add monitoring only cost £50 each or £600-700 with cloud connect WiFi thing (seems wasteful really for monitoring really) we can do around 3600w > 16A@240v in the UK with only needing to notify the electrical provider that's it's now live, above that we require pre permission before switch on (g99 or g100+ 16A limiting device I believe if they won't allow g99)
And not only that. If 1 panel stops working, the rest will still remain in service. Provided that the Optimizer is not defective. Then the system is out. But that is also the case with a regular string. What's better now? That extra cost? For me, yes.
And what do you change for the better when you have those values? Data without action has no value in my world. And often it steels the sleep ;-) More panels or less shadow would change something, for example. But I know, we are nerds and love our gadgets. So no problem with your decision.
@@AndreasSpiess You can identify a faulty panel. I did not decide for installation of these optimizers. They were included by our consultant for the solar installation. Back then I did not have the knowledge I have now and I might decide against the installation of these. Tho I did not have a problem so far and even if I would have one, the replacement of these modules is very easy, at least in my installation.
@@shrujanamsyama9940 You can also detect degrading of the panel and determine the influence of obstacles on your roof. Idk there are pros and cons. If you want em install, em. If you dont, dont....
Very informative, great approach. What we need now is a bigger sample set. As you say, the optimiser may have a 2% energy cost and you've robustly debunked the necessity of them in a string. However, could they potentially have enough benefit to counteract the energy cost? There are of course other benefits, not least the per-panel monitoring. The reality is, they aren't as prohibitively expensive as made out. Our PV quotes ranged from £14,000-16,000 - the additional cost of optimisers, perhaps £500-600 is significant but almost in the noise. We ended up picking a middle ground option, that was SolarEdge. No regrets, a very efficient inverter, and as I say, amazing to be able to monitor each panel's performance.
Some of the Y-axes are wrongly labeled (W instead of kW and once V instead of A). But it should be clear what it needs to be.
not sure but in general our 2022 solar panels and inverters show such string behaviour that 1 partly shadowed module kills the production from about 2500 Watt to 250 Watt and then 30 minutes later the same for the next string which is twice as big from 4500 W to 450 W caused by leaf tree. Each day the same time this can be observed where we moved from 1 bing string inverter to 3 smaller one phase and double string inverters which lead to 6 strings and therefore a much longer usage time of the full power . Our roof is now split into 6 strings instead 1 or 2 and the benefits are significant if you check the inverters string by string and not the inverters in general even though that will show the same effect .
If you add a timelapse camera you can mark the lines of the shadow when the shadow is too big for the diodes and the power will be lost.
I am not sure about the optimizers but the 4 different Growatt inverters we have had in the past 3 years for tests (like SPH series, 3000-S String, Mic 3000, Min 3000 and Mic 4600).
But they also come at a price, which means the price of the purchase of 3 smaller versus 1 bigger but also the lower production cause 6 strings instead of 1 big one means lower string voltages and too low voltages means that the inverter will start later than the 6 times bigger one. 3 Watt is not enough for a 4,6 kW inverter, but 18 Watt sure.
Not sure what the shadow optimizer inverter will cost or how big the benfits will be they can gain but it is not that easy.
I had played a lot with the simulators like that pv sol premium which Valentin Software offers for free for 30 days but I was able to gain more and more Watt hours by the right inverter size and amount strings and of cause the sting cabling cause our strings are build up vertical cause a tree is spending shadow from 4 pm in the evening which is now not so much wanted than 30 years ago. But it took quite a long time to find solutions for each roof, one with the chimney and the tree shadow and a lower angle while the other has a steeper angle but also its own issues caused by a GAUBE of 7 m widthspan . Gaube looks nice but 22 years later more or less useless nowadays cause a 7 m long segment of the south roof is lost and also causing shadows in the morning and the evening on one or the other side.
Hope to see the update video soon, but I guess it will take till spring. Meanwhile we will simply add more panels and another inverter which will help a lot to get the 20 kWp in Spring and maybe in 2 years the 30 kWp as maximum using the roof area better with small microinverters around the 2 chimney and lower areas where the greenhouse is which has a steel roof too. Gets a lot of shadow but 5,5 m width and 2 m depth will offer a spot for 5 big 525 Wp panels that can produce during summer full till 4 pm before autumn and winter will start the time of long shadows no one wants. 3 kWp here and 4 kWp there can make 5000 kWh even in shady situations if the microinverters can proctect these modules close to the chimneys. The Min series from growatt with 2 mppt has helped us a lot like the Min 4600 we could get for 600€ , but the smaller ones like the 1000 W inverters are getting expensive about 350€.
Unfortunatelly the cheap chinese GMI inverters or the SGB or many others we have all tested are not really working fully and throttle caused by overheating. Only the more expensive western ones get it done right but at a higher price, lot higher price of about 360€ per 500W - but at least without throttling.
Maybe hoymiles might be able to deliver a better performance and having the EU certificates like VDE AR 4015
@@typxxilps Your situation seems to be quite complex! I do not know the products you use. But the inverters must have built-in shadow management. Otherwise, they can react completely wrong (as it is the case at the beginning of your comment).
If you need more and more small inverters, optimizers or microinverters might be the better choice, at least for the shortest strings. But still: If a shaded panel does not produce energy, an optimizer will not fix it.
Micro inverter and Optimizer The solar equivalent of Batterizer see EEVBlog
Hi, Andreas! It's been sad for me to see that you don't number you videos any more. It is a killer feature of your channel! Numbers make it so easy to navigate through your vids and I go through them a lot! Let's say I need the video about A6 GSM module, I would simply type 66 in your channel's search and here I have it. I hope you consider keeping on with this tradition! Cheers from Prague =)
@@romanalexandrov2880 I still number my videos. Only the last few get their numbers later...
Really useful! Thank you so much Andreas for thoroughly researching this topic, performing real-world experiments and explaining your conclusions so clearly.
Thank you. You are ahead of time! This video is not published yet ;-)
@@AndreasSpiess Hi Andreas,
Would it be possible to add diodes to old solar panels that do not have them integrated in the panels?
@@Conservator. As you see they have to be placed on the front side. So if you can remove the glad, maybe it is possible.
Hi Andreas,
Nice video. I did my PhD on optimizers and PV shading and my conclusions were similar to yours. We also did an article on how to maximize power production of a PV plant by interconnecting the strings in parallel. Passive optimization works about as well as an active one.
Interesting paper. Maybe one day they add a few transistors inside the panels to switch each cell accordingly. Because with the current fixed bypass diodes, most of the loss comes from the panel. Which is also lost for the optimizer...
@@AndreasSpiess Thank you for the reply. The bypass diodes are an "economic compromise". When one cell goes on the negative region, its voltage is around -17 times that of the other cells. When adding them up, this makes for a negative voltage that activates the bypass diode. This means that 18 (or around that) are the maximum number of cells you can still protect with a single bypass diode.
If I'm not mistaken, space applications put one bypass diode per cell, to avoid this problem and maximize power production.
I suppose it would make the PV modules too expensive to do so. :P
Transistors would be too complicated.
could you help answer a (hopefully very simple for you) question - adding several higher power (much newer) panels in series with the old ones
I should get less than rated from the new addition, but something proportional to the new panel voltage * old panel current
right?
thank you
@@tjunkieu2b Thank you for your question. I will try my best to answer it. I think you are right: as PV panels age, they produce less current (but not that much less). If you connect a newer panel in series with an old one, they will have the same current for a given point of operation, but they might have different voltages.
Would you mind explaining a little more? Are you retrofitting stand alone/DC grid PV for grid injection?
@tjunkie Yes, but there are issues with mixed modules in a string, and if at all possible one should not do so, or you can significantly shorten the life of the lower power modules as they will be possibly passing more current through them than they are designed for. Best to have the same modules in each series string.
Micro inverters are maybe not useful in large installations, but in small ones they are a game changer. Not because of efficiency but by pure existence. With micro inverters small installations are easy to install, even by amateurs. Thanks to Micro Inverters balcony power plant are now very common and it will become more and more.
I agree!
Very much so: single panel plug and play solutions are up for grabs now.
Absolutely. Cost of microinverters for a 3 panel system will be less than a string inverter. In addition, it is easy to install. But if you are going for above 1.5kW power like 3kW or more, they become expensive
This video is why I'm subscribed, real life experiments, a bit of theory to go along with it, excellent stuff!
It's a shame so many installers have no clue how this works, not to mention installers that get bigger profit from installing micro inverters.
You and I are here for the same thing. I frequently encourage Andreas to focus on real world projects because, building stuff well is often remarkably difficult. I also keep encouraging him to use MicroPython on ESP32 instead of C/C++ ESP32 because MicroPython is both quicker an easier to work with, yet often just a performant-from the user's perspective-for many, many projects.
@Daniel: I found out that "Solar" people often do not come from electronics. So they believe their suppliers.
@@AndreasSpiess It's "found out" not "fund out". Pro Tip: compose in Google Docs to get grammar checking "on the fly" (automatically).
@@AndreasSpiess that would explain a thing or two ;) then again installers have access to hardware and putting up a small system to test shading wouldn't take them long but I guess many of them don't care that much.
Luckily we have people like you that are willing to spend the time to show and explain how things work without marketing BS and I'm trying to do my part explaining "basic" electrical stuff to my friends.
Thank you very much, Andreas! We got our PV system installed back in august, and because of your "warning" in one of your earlier videos, I made the last minute decision to cancel the already ordered optimizers - not so much because of their additional cost (I still would have loved to monitor the performance of each module), but more because of the chance they could fail one by one over the years, causing a lot of cost to replace or remove them, especially since they are installed at a hard to reach place. Better to have a reliable, though less performant system for many years, than a little bit more performance with the risk of more problems later on - these modules shall stay untouched on that roof for the next 30 years, if possible. (And saving an extra 900 euros on the invest did not hurt, either... :-)
I added a simple "divided by the number of panels" in my Grafana board. Now I can compare the cells and see if one creates problems. The only thing is that I do not know which one it is ;-)
Pretty much everything I've read and watched about optimizers says they are a waste or worse in most situations.
Unfortunately, since the US NEC 2017 & 2020 electric codes require Rapid Shutdown on a per panel basis, the math changes a bit. If you already have to put an additional RSD module on a panel, maybe putting on optimizers with RSD and monitoring instead makes more sense since you are no longer adding more connections or modules in the chain.
No one in the US seems to make (or admit) mistakes these days. You don't learn without making mistakes or failing at tasks... so, why not admit them. Love that you did just that.
I agree. You only learn with making mistakes. If you get it right, you already knew it before ;-)
Personal experience: I pay ~$50 for an SolarEdge S440-1GM4MRM with 99.5% efficiency (as per its data sheet). It's not "just" about increasing the overall performance of the installation, you debunked that quite nicely, but also about measuring individual panel performance. Helps with identifying misbehaving ones.
Agree. On my installation, one if the panels failed after three months. It was easy to detect as you get power per panel. Our system has 3.6 kWp facing South and 1.5 kWp facing West (both different inclined roof), on one giant string. Hard to say whether this works better than two independent MPPT.
If you value the additional features of optimizers, you did the right choice. I only did not like when salesmen use the "single leave" argument.
Exactly, this line of “optimisers are all about shade” is typical of string manufacturers. It totally misdirects from the many other benefits of MLPE like many safety features, panel level visibility, simpler fault detection and repair work, higher efficiency, dc coupled storage, the list goes on. Making it all about shade, which is a confusing topic and difficult to properly compare, makes it simpler to push the traditional dangerous string inverters with uncontrollable high voltage dc.
@@cybersax2 Solar panels hardly EVER fail nowadays. Quality control is so good.
@@BrentHeberAusNZ Ooh, dangerous string inverters! It's still the same amount of kilowatts coming down the string, which will still kill you, whether DC or AC. Who needs panel level visibility? What fault detection. Solar panels almost NEVER fail nowadays.Higher efficiency? Any proof of this?
Excellent video Andreas - your experience and analysis has confirmed that micro-inverters and optimisers really aren’t economic in most applications.
Thank you!
Excellent and eye opening video. I was never fully convinced to micro-inverter idea and it is really nice to see that PV in strings are in fact more robust than the common tale says. Thank you so much!
That is what I wanted to show in this video. Microinverters are a good solution for small strings.
Great video André. I thought i knew a lot about solar panels and optimizers, but this is actually a big thing. Thanks for adressing it.
I think more tests and theory explanation should be done in the following years.
Perhaps the days of the optimizers are counted..
(i myself have 32x 310Wp solar panels, in the same direction, without any shadow and sadly enough all with optimizers)
I currently test optimizers in an arrangement of four panels directing towards south and in a steep angle. I plan to add another 4 panels in the same direction but a different angle. 4 panels are not enough for a string for the Huawei (not enough starting voltage. So I might end up with optimizers in this configuration. At least I will test them.
Finally someone else who did the calculations. All the micro inverter salesmen have been very active lately. I always wonder why they lie about the non-micro inverter panels. Expensive and useless, microinverters are dumb. Also why hang them on the back of a solar panel, where temps often are 65°C. On flat roofs there is access, but on slanted roofs it's a mess if you have to replace a microinverter.
Microinverters are ok for small systems with only one or a few panels, I think (we call them "balcony power plants").
Great observations/tests to make us think about what to install Andreas. Hope you will go more into the use of the battery in the near future.
Just my n=1 anecdote:
I had a 110cm satellite dish on my flat roof that is North-South (about 165 degrees). My PV installation is 10 years old, 1 string, facing almost perfectly south. At just after noon there nearly always (on cloudless days) was a huge drop in my generated power for 45-60 minutes. Power "lost" was about 1 kWh which is about 10% of the daily power generated during summer. In the other seasons, due to high latitude, there is almost no significant loss.
The dish was out of alignement, but I cannot easily access my roof to either fix of remove. When my neighbor installed a PV installation, I asked the installers to remove my dish, not interested in fixing it not watching TV anymore. Since then the drop is gone. Power generated is a nice sine like figure. The dish cast a shadow on 1 of the panels apparently.
Don't know the make or manufacturer of my panels, but I guess they would have benefited from optimizers. But I have a feeling it wouldn't make economic sense to have had them installed.
Did you have "shadow compensating" MPPT? Also, did you have adequate bypass-diodes on your panels?
The forecast of the „optimizer guys in your case would be that you lose the power of the whole string. So if you lost all power during the shadow phase, the chance is that your panels do not have bypass diodes…
Really excellent and hard to find information. There's lots of bad information on the internet (as always). Thanks Andreas! I have been agonizing over whether I made a mistake in not installing the micro-inverters/optimizers. Now I can finally tell my wife WE did not screw up!
So you can be her hero ;-) I am glad that I was able to contribute.
I also forwarded your video to engineer and architect friends who are working with solar. You are the hero to all of us. Thanks again for your hard work.
Solar power is really big here in Australia for obvious reasons but most of the installers are dodgy and even electricians don't understand it. I'm about to put a solar plant on my roof so this is very useful.
I also learned that it is good if you know the topic ;-)
Thank you so much! I have two small panels in parallel on my boat, and a very simple (and cheap) regulator. I tried an MPPV system, and it constantly drained 20mA from the battery for its processor, day or night, without providing any obvious benefit. I can't afford 20mA for 24h from my boat battery. I thought I must be doing something wrong. But now I feel I need not worry at all!
20mA indeed is not a lot for solar power systems where we try to produce MWh. But in your case, maybe it would be of benefit to find a space for an additional panel. This often is the better way.
Wow, the technology evolved more than expected in this field. Thank you!
You are welcome!
Now I understand what the shade management algorithm is doing - a more global optimization in the MPPT controller. Older inverters don't have this and older panels don't have they half-cut and bypass diodes, so things are improving!
Definitively they are improving! And it is important that we know what it means for us ;-)
Really interesting video ! I have PV solar on my roof for over 14 years now ! In those days, it was rather expensive. But anyway, I'm glad we did it. Now, thanks to your information, I'm preparing a second installation on the roof of my workshop. A work in progress... but defenately very satisfying !
Solar panels are like tattoos: You always want to get another one ;-)
@@AndreasSpiess Can't really tell, I have no tattoo ;-)
@@PhG1961 Me too. But I know lots of people who have a few of them ;-)
It's so interesting to see the progression in the PV world. I also started in 2006, with three Sharp 165Wp panels on a Mastervolt Soladin 600 inverter. That system would seem like a toy nowadays.
@@SolarWebsite Mastervolt is a really good brand. We have 10 panels for a total of 2.2 KW with an Infinity inverter. This brand does no longer exist, but it's quite good, so far, never had any issues.
Great video, thank you for taking the time to experiment and share the results with all of us!
My pleasure!
It would be nice if some research could be done into modelling roof situations to better decide if optimizers/microinverters would be a good solution for special situations.
I think the vast majority of PV installations don’t need optimizers, but there still are a some roofs that might get better yield with them. Whether they are cost effective is another discussion, but to get the best yield sometimes they are needed in my opinion.
I agree.
I'm so glad, you made this RUclips channel.
I don't need to know any of that, but can't stop watching :D
Thank you for your kind words.
This is exactly what I concluded before installing my panels... but I have roof-surface facing all sides so optimizers seemed a reasonable solution to have west/south/east in a single string.
An alternative would have been a three string inverter and connecting them up separately, but that also creates some problems in terms of wire mess (and these inverters are pretty expensive).
Another possibility would be micro-inverters to east and west, and a simple string south... but again wire mess and on top of that mixed solutions. The price differences for different solutions were not big enough to be the deciding factor.
My main worry is indeed that an optimizer might break, it will be easy to find it but hard to access. 25 years warranty does give some confidence, but the price of the optimizer is not my main concern obviously - loss of production and man-hours to fix it are not covered.
I agree, in some cases optimizers are good. I am not against them. I just wanted to show what is true and what not. Then my viewers can decide..
@@AndreasSpiess And your work is appreciated. So many options out there gives a lot of space for installers to 'peddle' expensive solutions - information is key!
@QPW Chris Indeed, the implementation of any warranty beyond 5 years is always a bit of a question mark.
My house is about 17 years old now and some things are starting to break (as expected)... good luck finding *new* parts to repair anything!
@@diatonicdelirium1743 Yup, older parts are very difficult to replace. It is always better to buy standardised unbranded/generic items which will remain standard for long time than buy fancy branded items which keep getting upgraded!
Thanks Andreas - another fantastic, clear video. I have optimisers fitted in my "professionally" installed system - seems like I fell for the bullshit! One good point about the optimisers (although possibly not making them worth while) - they do let you know the individual output of each panel. Like your brother, I also lost most of the summer's production due to slow delivery of the SolarEdge inverter. I wish I had considered my installation after you, and could have taken advantage of your superb research!!
At least you have a working solar energy factory now. That is the most important fact!
as someone new to solar and trying to get up to speed you have helped me a great deal. All our panels are on the same southern facing plane of our roof and therefore microinverters are unnecessary but yet the company doing the install has used them. We will be questioning their usage and likely asking for their removal since they are ineffective and expensive. And requesting the appropriate price reduction according to their cost. Our situation is something of a nightmare, a company that takes advantage of its customers by being greedy, and overly accepting of what the equipment providers ad copy says. I dont buy half of it, and am becoming increasingly antagonistic towards my installer for that very reason. If we don't end up in court it will be a minor miracle. Thank you again for your common sense approach to this. I am very grateful.
Many installers are also "cheated" by the suppliers. Keep in mind: This is a growing market with many new entrants without deep knowledge. I would inform your installer, not punish him.
Many thanks for the explanation. We are currently looking at installing 7 PV panels with micro inverters. Your information is extremely helpful.
Enjoy your installation. Solar is a lot of fun!
Great experiment, I had long thought "but what about the bypass diodes?" to these claims.
BTW, I have an Enphase setup, not because I was sucked in, but because my panels are at different angles and alignments, 3 vertical, 5 South ,6 East and 6 West, and that would mean 4 strings, and probably more than 1 inverter... Instead I have no big central inverter at all, which is big bonus to me.
For those setups, micro inverters are a good solution. Particularly for short strings.
A fantastic video! The best solar optimizer is to take cash used to buy optimizers and buy 30% more solar panels! Being a fellow amateur radio operator, the last thing you want are a bunch of poorly made switching supplies on your roof connected by a long antenna.
Indeed, like with motors, displacement is best replaced with more displacement ;-)
@@AndreasSpiess The saying in the US is. 'There is no replacement for displacement' ... :->
Thank you! It's absolutely horrific how much money is being wasted in optimizers!
You are welcome!
Hi Andreas. When I did my research regarding my installation, I came accross an online huawei simulation tool. It allowed me to choose, panels, cables, etc. It also provided the option of adding the huawei optimizers. What I noticed in the simulation charts, is that the inverter started to work sooner in the winter with optimizers. As I understand, an inverter needs a treshold current before it starts? The optimizer would lower the voltage to increase current in order for the inverter to start. This would be interesting in overcast weather.
I asked solar installers for their experience, but no one could help me verify the simulation claim of huawei. I also didnt find anything online about this particular scenario.
Maybe your LTSpice simulation can verify this?
If my simulation charts where correct, the optimizer was a good investment over a couple of years.
Regards,Wim
Hi Wim,
What were the numbers they told you in the charts?
The inverters I know start at 100 mA or below, so that would be at a power of 50 W (at 500 V).
You would have the same amount of loss in the optimizers at 2500 W (2% loss).
It really depends on the inverter and specific optimizer, but it doesn't convince me...
I also haven't seen any optimizer manufacturer advertising this as a big advantage of their product (I work in the PV industry, also with optimizers and rapid shutdown devices)
Cheers
You might pick up a few minutes at the start and end of a day, where you are making maybe 3% of peak power.... you're still better off taking your cash and getting another solar panel.
The open voltage of the panels increases extremely fast and reaches the needed minimum voltage.
And, as James mentioned, I do not care about a few percent of a kW when it only costs a few cents. One panel more gives you much more bang for the buck…
My solaredge system was installed 7 years ago in a single string split orientation system. 10 west 10 east. As i have nothing to compare it with i cannot know if I fell for the sales pitch. One benefit of the optimizer is that you can see when a panel in a string is not working and get it replaced under warranty. Not sure I would have noticed that one pannel in 20 was only giving about 50% without being able to see the output of each individually. Good video as always.
I agree in your case. So I assume the panel was defective and not the inverter. What was the issue? Maybe you would have discovered that the string is no more up to date. But for sure you would have to find the panel manually.
I am also glad to read that your optimizers still work. An important information!
@@AndreasSpiess I don't know what was wrong with the panel, the supplier turned up with a new one and swapped it out. All panels look about the same still. The system output does not appear to be reducing over time, just fluctuates from year to year. Keep up the great work.
This is why I decided to get optimizers. 👍 🤷♂️
Thanks for that dear Andreas. I see Dishka is in good shape still. All the best from Croatia.
Yes, she is still happy…
I agree that optimizers aren't really worth it compared to bypass diodes. But I do like the safety aspect of micro inverters. High voltage DC can cause arcs that do not extinguish. It is one of the major culprits in residential solar fires.
If other criteria are important for you, your decision was perfect. This video was just one claim often made.
Thank you for this contribution to such a controversal topic. The YT channel MC Electrical has 2 vids "Fronius vs. Enphase" ca. one year ago with real measurements. Similar did "NRG Solar - National Renewable Group Australia" laso ca. one year ago.
Some percent in difference in practical cases.
I also simulated different parallel vs. serial configurations for older 200W modules using LT-spice. Results were also not soo straight forward.
My 36 PV modules are on the roof now, but I also miss the Fronius und the BYD battery. Maybe next year... we will see. And the switch box for backup island operation has no delivery date up to now. Here we call it DDR 2.0. And do not refer to RAMs. ;-)
Thank you for the hint about the Australian video. I thought I had it included in my comment, but I forgot it. Now it is there. And good luck with your inverter and battery...
What I have read, optimizer is a good choice if you want monitor status of every single of solar module. It also has safety feature that reduce DC voltage to touch-safe levels and arc fault detection.
You are right. Keep in mind that after a few weeks, most users only check the power every month or so. And only for the whole plant and a whole day... The risk of having electronics in the heat and humidity of a roof stays forever...
At my latitude (42N) I found that South facing and tilted at 60 to 65 degrees was best for winter production while having the least effect on the total yearly output.
You are right. If you optimize for winter, south and steep is best. I added a few of those, too.
My biggest "a-ha" moment in this video was the "shadow-management" local-vs.-global optimum-seeking explained around 13:25 onwards. Once again, thanks, Andreas!
You are welcome!
@@AndreasSpiess Dave from EEVblog did a video about a similar topic, ruclips.net/video/AbxHoQF4ADk/видео.html - i suspect his older inverter was not equipped with global-MPPT.
@@AdityaMehendale I saw this video, too. It must be something like that. I assume bypass diodes would not have kicked in with this small shadow.
Maybe we should have all panels in parallel and use very thick cable or bars to minimize loss and the mppt as close to the panels as possible. What a test, hats of to you Andreas, taking speculation away and making it a real life test with the correct knowledge and measurements.
Having them in parallel isn't perfect either. A shaded solar panel can become a current sink without a blocking diode. Although counterintuitive because a solar panel is a diode, it's only a very weak one and reverse current can cause it to fail. Moreover, the thick cables will cost a lot more than having a mini inverter behind the panel and using classic 230 cabling.
@@matthewmaxwell-burton4549 I agree and value the response! I have a mix of different 18V max mppt panels just for testing. I actually wire them up in parallel with a Schottky diode in series of each panel. I know not 100% efficient but there is a best approach which each situation.
With small panels, maybe more arrangements are possible. If you work with higher power you have to take the capabilities of the inverter into consideration. I like if my inverter and battery use high voltages because it generally reduces loss and reduces the needed copper.
Thanks for the thorough testing. So the claim "one panel shaded halts all output of the entire string" began with earlier panels and now won't die. Glad mine are the latest Longi Hi-Mo. The spec sheet states "Half-Cell" but doesn't detail the implications or discuss shading. Glad you explained it. Seems I lucked out in my craigslist purchase. I thought my panels didn't output much until 10 am in Summer since one panel in each string of 7 is mostly shaded before that, so exploring optimizers. After your tests, I suspect more likely just due to the sun angle since my panels at tilted slightly to the southwest.
The tilt indeed has a big influence!
messing with the industry telling you should spent your money ... I like it.
I thought it was necessary. We should at least know the reality before we decide...
@@AndreasSpiess yes! Good decision to make this video!
@@AndreasSpiess truth must be told, the true duty of the enlighten! Thank you for the video :-)
In my case I have 3 panels pointed east- (slightly north east), and 5 pointed west (slightly south west), all on a 45 degrees roof. 2 pointed west get an early shadow. For that reason I choose microinverters. A solar-edge system wouldn't have worked in my case, 3 panels wouldn't produce enough voltage for the inverter to work properly. They say Solaredge has a minimum stringlength of 8 panels or so. I only have 8 panels, but often only 3 of them get full sunlight. For that reason I choose microinverters. It made the installation a bit more expensive, but I'm still happy with it, especially now the energy prices went up so much.
The nice thing about mictoinverters or (solaredge)optimizers is that you have per panel monitoring.
For anyone interrested: the panel pointed east made 267kwh so-far this year, the ones pointed west 340 to 380.
I agree, for small systems without battery, microinverters might be a good choice, particularly if the panels point into different directions.
Thank you, what an eyeopener. I would only want to point out that with the optimizers or microinverters you can monitor the power output per panel and that is a nice feature if you have to do troubleshooting in the future. Only you need to know the layout of your panels and which is which or you still don''t have any usefulness. But on the other hand, you really have a good point about the active electronics in the harsh environment on the roof during the hot summer sun. What is more likely to break anyway...
Single panel monitoring seems to be important for many people. Professional installers know that panels do not fail often…
The monitoring is a nice feature but not needed. Even without it, the panels will function the same. Even if you manage to find out that 1 out of 15 panels is defective, the cost of optimiser would have been much more than the cost you saved by getting 1 panel replaced
This is exactly what I had in mind when I designed our setup, nice to see my theory put to the test!
It feels always good to see that an idea was right ;-)
for NEC 2020 and roof mounted panels you need rapid shutdown modules for each panel anyway. Optimizers etc have that built in, so factor that cost in.
What would I have to add to my string inverter (LUNA2000) to be compliant? Here, I had to add nothing.
edit: I just found information: It seems you have additional specifications in the US. So you are right. Fronius sells such a RAPID SHUTDOWN box for 160 dollars. And it is needed for each string.
@@AndreasSpiess I think sometimes it's each panel if roof mounted,if ground mounted then don't need RS for each panel. I think Tigo brand is about $30 per panel for each shutdown module.
When you completely block a panel, the bypass diodes go to work to prevent blockage of the current. (All panels have bypass diodes) The value of optimizers is that they improve current flow during partial blockage. I use Tigo optimizers and they improve my partially shaded arrays power output by about 25%.
Good to know. So I assume you the glad guy who has a possibility to do a parallel test without the optimizers?
@@AndreasSpiess I put the optimizers on since I had a lot of (partial) shading at different times of the day. My "comparison" was before/after installing the optimizers and there is a clear boost in power. I like the Tigo optimizers since you don't have to install them on every panel, just the ones with shading. The software gives you a reading on what they call "reclaimed power" which is their calculation of the extra power "reclaimed" by the optimizers. For me, it's about 25% extra on my 4.5 kW array.
@@MarkSpohr That 25% is a value which is give by the company which may not be accurate. The actual benefit of optimisers is less than 10%. Remember that the optimiser themselves consume 2-3% power. Also, the shaded panels will activate the diodes regardless of optimisers as diodes are inbuilt to the panels. So, your 4.5kW system would perform the same as a 5kW system without optimiser. Unless the cost of optimiser was less than what is needed for 0.5kW additional panels, it is not worth it
@@shrujanamsyama9940 Where is your data that shows "less than 10%"? My data shows 25%.
@@MarkSpohr The data on what basis? Is its a theoretical data or based on actual difference with & without optimiser? The 10% is based on empirical evidence on how the diode works. The difference between the diode & optimiser is minimal and is in the margins
great video, thanks for bring this topic up and make it clear. it is ridiculous how some people mislead others in this field
claiming incorrect result relying on unrealistic condition
I agree!
Thank you for not using an ugly typeface. Yes, I notice these things.
You are welcome!
Writing from the wet tropics of Australia: Yes I agree the use of optimizers is rarely worth the extra cost. Here we have issues with huge UV exposure and many systems get replaced after just a few years. Bypass diodes are a way to save infrastructure cost and our feed in credit is now so low it is hardly even worth having panels at all.
You are lucky you are far away from Europe. Here, energy currently is a very hot topic ;-)
I had a hunch, and I'm glad a channel I know did the same test. I've come across a few videos saying the same thing, that micro-inverters while they technically do something, the cost/benefit might not be as significant thanks to diodes (or good placement).
Microinverters are a good choice for small systems, I think.
Fantastic video -thank you for the useful insight! Some people have more money then sense - sadly some of these solar installers will believe the marketing hype and happy to accept the kick backs / commissions from these big vendors!
Most of them are not specialist. So they go with what they are told :-(
WOW! awesome video - I confirmed the exact same findings by trial and error when trouble shooting for a bad panel. In defense of the micro's and optimizers - older generations of solar panels probably did not have that pesky diode. But modern half cell solar panels just changed the game completely, they are amazing in dealing with shading. I wish there were more grass roots videos or communities like this, because most solar system installers have no clue about the intricacies of solar installations beyond wiring and city codes, hence the easy way out is to just simply offer a money making stop-gap measure. Good job, man!
Thank you! These diodes seem to be the standard for the last 10 years.
What could really help is a system that can inform you, on an app, what cells are not getting unlight. Without that, you have to either go outside and check, or just guess, and it may not be that obvious whether you have one cell a leaf blown over (easily fixed), one that a bird crapped on (a bit harder), or a cloud drifting overhead, which you can't do anything about.
I divide the power by the number of strings. Then I see if something is wrong (which seems to be very rare). But if you want this feature, optimizers or microinverters are your solution.
Hi Andreas,
I have a very important observation to the topic, where probably optimizers have an important role! I have installed a Huawei Sun2000-8KTL inverter with 2x9 Sharp 410 PV modules conencting to the 2 MPPT trackers. These PV panels are on a south directed 35° inclination roof in two parallel rows, where I have a tall tree in front. So like a solar clock, the shade wanders around. I was convinced by the installer that no optimizers are needed, along with an australian vlogger, who made tests on different mppt inverters. However! Today the sky is crystal clear, in Hungary, the heating is off, that it warms the house, but the 7.2kWpeak system gives only 0.6kW at 2x310V, compared to a foggy ugly diffuse day few days ago, when I had over 2kW. Also some other day, with clear sky, it happened, that in the morning I had 210V and 9 amps on one string, then jumping up to 310V and 1 A. So, my conclusion is, that if you totally block the light from given areas, then the diodes will bypass, and you get to an optimum, but if you have some partial light, casted by leaves on a tree, it will screw up the mppt, and give low currents at high voltages. I will do observations further, maybe you could test this feature, or we can discuss some experiments and data to see here more clear.
Strange behavior of the inverter. Maybe you contact the Huawei support. That is a strange behavior.
Thank you very much for this analysis.
I always asked myself how the cells behave in such a situation, because this year I am planning to built a small solar balcony power plant with storage, where I can only install two panels in very different angles. So I will have very different sources with relatively low voltage and have to figure out the best converter to charge a battery so I can use the power when I am home, which is mainly at dark hours. Schöne Grüße aus München!
In your situation, I would go with microinverters.
Dear Andreas,
Hello from Geneva!
I am so happy to have found your channel by accident, i am watching all your video and love them! Thank you for your work and very clear explanations. I really appreciate the fact they debunk the « sales speech » of certain manufacturers such as Enphase but also most video on RUclips still encouraging blindly to ise micro-inverters in any situations. (I noticed it especially in french video (in France they still encourage 2-3 kWp lower cost installation with expensive Enphase micro-inverters 😅)
I just made up my mind today to go for a 14kWp 2 strings setup, i will get the Huawei SUN2000 hybrid inverter and get the battery in few years hopefully cheaper. Initially i ordered 6 optimizers (chimney shade) but i will now probably cancel them.
Thank you!
Welcome aboard the channel! For small systems, microinverters might be a good solution because string inverters usually need a certain startup voltage.
For your large system, a string inverter is a good choice.
Great video! I am planning on installing solar power on my roof, I have about 100 square meters of the roof facing south that can have panels. This gave me valuable information in deciding what to get. Since I'm also into Home Assistant and dabble in electronics, you are my perfect video guru😁. Best wishes from Sweden
So enjoy your project! We love it even if it does no more cover our full needs because of the weather and the low sun.
That is a pretty big roof considering it is just one side! You can place 10kW panels easily on it Depending on latitude, climate and shading, you will get 30-50kWh/ day on average
Excellent empirical study. Thanks for sharing the results and mith bussting.
You are welcome!
Thanks for this! I was just contemplating optimizers. Dishka looks very content. :)
Glad it was helpful!
Great video and very informative, as always. Thanks for sharing. Happy to see Dishka at the end
You are welcome!
I've been running 12 panels with SolarEdge for 7 years now. Last year I had an optimiser that stopped working. The main advantage of this system is that you can track the performance of your panels. I called the company that installed my system and they replaced the optimiser free of charge (Warranty). 12 panels should give me 3420 Wp but they outperform every single year. Got at least 3800kWh every year minimum. The only thing I don't like is that the main invertor is passively cooled. So I added some PC fans on a timer to cool down the backplate during the hot summer months.
Thank you for sharing your experience. You were happy that they replaced the device free of charge! Here they usually are not allowed to go to a roof just like that ;-)
Andreas, thank you for your extremely well-structured and comprehensible video. It really helped my understanding. Michael👏
You are welcome!
You forgot two important advantages of optimizers: Safety and maintenance. Because we installed the system ourselves, we don‘t want to deal with 1000V on our cabling and risk light-arcs every time we unplug something. And we want to see defective panels without having to book a flight with a thermal imaging drone.
I don't know enough to understand how relevant your comments are. But it would be great if these points were addressed. Thank you for bringing them up.
My experience has been that the optimized systems also start producing earlier in the morning and keep going later in the evening, because the startup voltage of the system is lower.
He adressed maintenance. Don't put complicated electronics in a wet, cold and hot environment. The only thing you tend to monitor is a failing micro inverter, not a faulty panel. (solar panels are very robust, inverters are not).
I just switch my inverter off before I start the (very rare) work on the roof. Not a lot of work for me.
@@AndreasSpiess that does not help. The full DC voltage is present on the whole string as long as there is sun. This is dangerous. And optimizers completely eliminate this risk and let only 1V per panel out.
Although microinverters contain an optimizer, I believe the real benefit they offer is redundancy. They connect in parallel.
Thank you for the excellent video. I believe I have watched others of yours regarding antennas and vnas and thoroughly enjoyed those too.
Yes, microinverters contain an optimizer and they offer redundancy. The only thing I do not like is that they are exposed to extreme heat and humidity for 25 years. But this is a general remark...
@@AndreasSpiess As a relatively new enphase customer, the thing that bothers me the most is that I don't have full access to the very expensive equipment that I own. Some functionally is locked behind an installer login.
Not a big fan of their marketing, either. As you have exposed in this video.
@@saltyexxer8253 I have a "balcony power plant with two panels and two Enphase inverters for a test. I just added a simple Sonoff to know what they produce. Sufficient for me. If it is only half, one is dead ;-)
Hi Andreas and greetings from San Francisco. Thank you, you just saved me a bundle.
Cool! Sleep well. It must be late!
Yeah! A video about a real world engineering project! Thank you Andreas!
My pleasure!
The company I work for installed optimizers before. Now we don't except in very extreme cases. If we have a huge tree or different directions, we recommend it. On standard flat roof, not recommended.
Thank you for sharing! For extreme cases they are ok, I agree. With the tests of the video people should be able to see where these extreme situations are and where not.
Great video, as always! One thing that I do like about optimizers: in case of a fire, when you turn off output, the optimizers will drop the voltage on roof so there is less chance of electrical shock to firefighters.
I didn't know about bypass diodes, makes sense the manufacturers implemented that :)
Yeah its a good idea on paper, but an incorrect installation, a factory defect, or like in most cases a bad connector tend to start these fires. Connectors failing are one of the leading causes of fires and when adding optimizers you triple the amount of connectors in a system. Not to mention that not all systems are equipped with such devices and they can fail closed leaving the system still energized. So in the end you have to train for the worst case scenario and treat each as it is live. The real solution would be to integrate into the panel but I don't see module manufacturers wanting to do that.
I do not know how dangerous it is for firefighters. As always, you have to hold both ends of a "battery" in your hands to experience a shock. Reading through recommendations for firefighters in this case, they list various things. None of them related to be electrocuted by the panels.
Have you checked out 'ideal diodes'? MOSFET configured as a diode. Allows bypassing solar cells without the diode drop if the cell isn't producing as much as the rest. Small and inexpensive. Could be used on each panel individually for maximum efficiency.
I assume they have reasons for not using this circuits because they are well known to most engineers.
@@AndreasSpiess After reading the datasheet for MP6914 (an example ideal diode for solar panel bypass IC), I'm thinking the reluctance for using this circuitry is that it appears to require a separate power source. I was thinking they worked solely on the power provided by the solar cells (supposedly making them a simple drop in replacement for the bypass diode). This would complicate wiring if that is the case making them a pain to integrate.
They also show it used once per solar panel string/group rather than each panel...which reduces its overall benefit over just a single cheap low voltage drop bypass diode. If bypass diodes were being used per panel I could see more benefits of replacing those with ideal diodes (especially if many panels get regularly shaded from moving clouds or foliage).
Optimizers are great for the company that installed your panels, if one of your panels get broken you can see that in de app / on line. So the install company will save time figuring out what is wrong.
Great investigation, and great demonstration and use of LTspice!
Thank you!
Thank you for this analysis, I’ve been sceptical of such micro inverters, they probably made more sense on older panel technology, now I now what to look at for my system, need to Google shadow management now on inverters 😊
Good luck with your system!
Sackstarch... Ich hoffe du lehrst an einer berufsschule oder fh eltech. Du erklärst sauber und extrem verständlich. Habe viel aus deinen früheren videos mitgenommen. Danke das du dein wissen teilst.
Vielen Dank für die Blumen ;-) Ja, ich habe früher an Unis und Fachhochschulen unterrichtet.
Answer before looking the video:
Complicated setup with different orientations in a string or lots of shading: yes
Otherwise: no
I agree!
The bye pass diodes will route the current around any panel not performing as well as the others this has been my take on the situation. What I did not see any reference too. Was the case of using mismatched panels ie units of diferant outputs where the use of the optimiser will allow these to all add Energey to the total . I personally use two panels connected in parallel to one optimiser each panel is 300 w one facing east and one facing west at quite a steep angle. I have 8 panels set up this way these are to catch morning and late evening sun. Especially in the winter when sun is lower in the Uk . During the mid day sun both panels are producing but well below there optimum so still ad to the total . The idea is to flatten and widen the hump and make the power more usable . I also find the feed back from each optimiser hand to see what each panel is producing. Again this shows that all panels point same direction but they do not all produce the same kw over the same time period . Again a lot of this could be done utilising the blocking diode but each diode that operates adds a volt drop in the feed that costs power also . Interesting vid
Connecting two panels with different directions to the same MPPT creates a difficult problem. The best voltage for one panel is very different from the other. So, which voltage is best? It can only chose one voltage if the panels are in parallel.
@@AndreasSpiess hi yes that is true . But in practice I find that as they are east and west facing panels. In the morning the the east panel is seeing the sun and provides power as the sun comes round the east panel starts to reduce its output while the west panel output starts to rise as the sun starts to shine from the west till sundown. At about midday I have current clamped both panels and they both are supplying some current for a short duration until the west panel output is rising and taking the lead . For me it just saves the cost off a optimiser on a panel that will spend 1/2 of the day out of direct sunlight .
I now have some more optimisers so out of interest I may just try adding an extra one and see if the return is a lot better
Installer and service guy here. SolarEdge breaks. ALL. THE. TIME. Save your $$ and get an SMA inverter and run your panels straight to each other. If you have issues and absolutely HAVE to have system optimization, get Enphase microinverters. If you're gonna spend the money on SolarEdge, just go all the way and get Enphase. I have about 12 service calls in queue here in central Texas, almost all SolarEdge.
Thank you for sharing your experience!
Very interesting! I wanted a low voltage DC system for a water heater and some 135W half size panels. After fiddling with designs I realised it was more efficient / simpler to have separate optimiser / boost converters on each pannel - with high current / low voltage losses mount up very quickly. It seems to work quite well actually.
Glad it works. I did not say, it does not work. I said it is not needed...
@@AndreasSpiess Yes sorry I didn't want to imply that! One thing occurred to me though.. for most applications isn't the selling point of the optimiser to convert directly to 230V AC at each panel? Not applicable in your own case, but I could then understand that it would shift and break up total losses from a central inverter to each optimiser so not quite as bad!
@@tolsen4893 These "optimizers" are called "Microinverters" (also covered in the video)
@@AndreasSpiess Ahhhhh! Sorry that makes sense!!
Andreas, great Video. Thank you.
As you pointed out, there are situations where optimizers might be worth it:
* Assume a situation where a panel is largely shaded. In this situation, the shottky diodes would bypass the whole panel. This woul not result in the whole string "tied down", but still, this one panel does not produce power anymore. Whith an optimizer, this Panel could still deliver it's small part to the system
* Solaredge Optimizers allow to build very long strings (up to ~11.5kW within a single string), and you kind of don't need to care about elecrical Specs (VOC of the String in cold conditions or the like) or direction of panels. This might facilitate cabling a lot - you need to factor this in for the investment decision
* Solaredge inverters tend to be cheaper than others as they don't need to perform the MPPT anymore - they operate at a constant voltage.
I would be interested in how you "controlled" the optimizer in your test setup? To my knwoledge, you need some Solaredge device to tell the optimizer what current it should produce.
Best regards
I did not want to condemn optimizers! I just wanted to show a false claim. In certain cases, they are good. However, if the bypass diodes kick-in, also the optimizers do not get energy from this part of the string...
I cannot control the optimizers. It is done by the inverter.
The diodes are inbuilt to the panel. If the panel is largely shaded, the diodes will be activated even before optimiser can do anything. Even if the optimiser manages to salvage some power, it will be in few tens of watts which is miniscule when comparing panels of 300-500W.
Keeping 10kW in a single string is a serious risk of safety and highly expensive due to high amount of current flowing. Even if you can do it, you must avoid doing it.
Interesting. And as is often the case: keeping things simple gives the best value for money.
I agree!
Extremely interesting this post of yours. I guess that common mortals like myself trust it when they say "You have to have it" but now I ask ... do I really? Thanks to, at least, planted the bug and make me research this more deeply.
Research and open eyes are always good ;-)
Another great video by you! Thank you for all this relevant information that helps me now know what to expect when this situation arrises and now what to worry about!
You are welcome!
The optim. very good, if You have to increase your panels, but you cannot get same watt . If your grid has 410 and next year you cannot buy just 450.. in case is very usefull.
Also if you have a difficult shape of roof, with shading.. optimisers can win 10% of production.. and panel life as well.
A simple roof, simple and nice situation you dont need.
I think we agree...
Thanks for the video Andreas. I giggled at the 15 o'clock and my cat ran off when she saw your cat.
Sorry for scaring your cat;-)
Another reason towards micro inverters is, that the system gets slightly more reliable. When something like a Strom destroys parts of your array, your prodution isn't gone completely.
Also a failing inverter is not as much of a problem with bigger setups. And the components are less expensive, as the requirements are lower ;)
I agree if you think that microinverters are extremely reliable even if they have to work in harsh and hot environments. Time will tell…
I have had Enphase for a while and one thing I noticed is that with more microinverters the slower the response time on loads. When I only had 10 it would respond fairly quickly when the A/C turned on/off. When I upgraded to 24 there was a noticeable lag on both turn off and on.
In my last video I showed a bit how inverters work. They are PID controllers. If you connect many in parallel they probably influence themselves.
@@AndreasSpiess The power line control interface also interferes with my old X10 controlled lights, is slightly annoying. But I hear there are solutions.
Interesting video! I recently had my PV installed with a SolarEdge inverter and 2 sets of 9x 370W solar panels with optimizers.
I certainly did have to wait quite a while a while until the inverter for SolarEdge was available. I choose optimizers because 1 set is on the roof of my house and the other is on the roof of my "shed"/parking spot. I also thought I would need the optimizers due to potential shadows of trees a few moments during the day, I guess that was less true than I thought.
Don't know exactly how much I paid for each component, but it wasn't necessarily cheap and let it install by professionals (they already came back to sort out an issue).
I would not look back because your PV works.
This is for people who have to decide and I wanted that the are not cheated.
Thanks for the info. The cat was a nice touch at the end.
You are welcome!
Correct, there is only few scenarios where optimizers pay itself. And yes, optimizers are more often point of failure.
:-)
Run DC at home: Kettle, TV, computer and anything with universal motors or switchmode power supplies works just fine on DC You might only need to throw away your AC clock radio.
Interesting concept. I would not want to go so far. AC has its advantages, particularly for a married man ;-)
Thanks for the video. I hadn't previously considered that MPPT may not track to optimum due to shading. Perhaps because I just expect them to work as expected!
Those bypass diodes are certainly worth their weight when used and sized properly.
the inverter tracks to the maximum global mppt, but obviously can't step up or down single panel voltage
If you look at the U-I diagrams of the panels you see, that the panels adjust voltage according the current. This effect is used by the MPPT.
There is one thing to consider, and that is partial shading of a single cell. My panels have 3 anti-parallel diodes, one diode for every 24 cells. If one cell gets partially shaded, it is reverse biased by the other cells in that series. That can lead to localized heating(often referred to as "hot spotting") of the surface of the cell still exposed to the sun. That permanently degrades the performance of that cell. A MPPT that does not globally scan, means that the string only produces the current of the weakest cell(the one partially shaded) and you lose a lot of output, but, you don't damage the cell/panel. Clearing the leaf, or the bird deposit is all that's necessary to restore full rated power of that panel. The ideal solution, would be a short barrier Schottky diode anti-parallel to each cell. That one cell gets by-passed by its diode, and you lose only ~5 watts of power in that whole string, and the cell is protected, until the panel gets cleaned.
I agree 100%. If all the panels in a string are mounted on the same plane, the losses of the DC optimizers is greater than the differences in the outputs of the panels. Even as they age, I don't think the differences will be anywhere near that 2% used up by the extra electronic modules. They cannot compensate for the damage of the hot-spotted cell caused by the partial shading. the only hope is, to limit the current load on the panel, protecting the cell, but, instead of the whole string being de-rated, you only lose the rating of the one affected panel.
Thank you for sharing your thoughts!
Optimizers have a safety advantage , like safe DC from solaredge where it cuts the voltage to 1volt when the string is broken / disconnected. In comparison to a string situation where the current/voltage keeps flowing.
My cables are insulated. So I do not see a big safety advantage. To be electrocuted you need a defective plus and a defective minus pole and touch both at the same time. Still, it is DC which is much less harmful than AC.
What Attix15 means i think is that when there is a damaged cable internal or exteranal there forms a ark because of the high voltage and a optimizor stops this. And at high voltage DC i would not say it is much less harmfull then AC.
@@kwynton8272 Most modern inverters have arcing prevention. And I did not read of a lot of accidents because of arcing. AC is more harmful to your heart because it interacts with nerve signals and can make your heart stop. Otherwise, the effects are the same.
@@AndreasSpiess Andreas how is it with the Firefighters in Switzerland? Here they have "issues" with extinguishing something with more then 400V and will not permit above 800V.
If you cut the cable between two solar panels and held both ends of the cable, then you’ll have a problem. But consider a few things:
1. You haven’t shutdown the solaredge inverter first, so there’s still a circuit voltage for a few seconds after a circuit interruption (400V capacitors inside the inverter).
2. If you isolate first at a string inverter, while your solar panels are still generating voltage, there’s no current because the circuit is interrupted on both poles. So you can’t get a shock if you insert your body into the circuit between two panels.
3. Solar Panels are generally not earthed, so there’s no return path for any circuit current via earth back to the inverter if your insert your body into the earth fault circuit. Dry rooftops at least don’t offer a good path to earth either so chances of suffering a shock due to your body being on the path of an earth fault are minimal.
3. Opening a connection via an MC4 connector is very safe. They are very well designed connectors leaving no naked conductors that can be reached with your fingers.
The most important intervention to make before going onto the roof to address your solar panels is to isolate the dc string first.
Thanks a lot for this series, it’s nice to get a professional and knowledgeable experience.
You are welcome!
Very interesting and thanks for Sharing your results. Bypass diode is the magic word here. In combination with an advanced Inverter (shadow Management). I worked in the PV industry some years back and experienced the rise of these companies. We had cooperations with these companies and also investigated what you showed. We had large Installations where we did what you investigated. I agree that it is not worth the money for a private small PV construction. It is more important to work with an expert to plan and Install the PV modules. Optimize to get the most of it is not trivial and needs good expert. Microinverters have their niche but is not the general solution like you said and showed.
I currently play around with microinverters as a "balcony power station". They are fun for one panel and very easy to use.
Been looking at getting solar installed, all the panels are going to be same side facing can get a SMA 3600 string inverter for good price and 8x550w ish panels for over sizing as I am in the UK I don't expect the inverter to clip often even with 4.4kw total solar panels (they likely to be doing less then 400w per panel as it's the uk, so I am considing 10 panels as that brings the peak voltage the solar panels can make to 440v DC witch is maximum mpp but still way below 540v max DC)
was looking at solar edge or enphase but adds £700+1300 inverter to £1500 (enphase) but as it seems it just add per panel monitoring really
I could get tigo TS4 power optimizers at a later date if I wanted monitoring, as they work with any normal string inverter systems as they add monitoring only cost £50 each or £600-700 with cloud connect WiFi thing (seems wasteful really for monitoring really)
we can do around 3600w > 16A@240v in the UK with only needing to notify the electrical provider that's it's now live, above that we require pre permission before switch on (g99 or g100+ 16A limiting device I believe if they won't allow g99)
I do my panel monitoring by dividing the string power by the number of panels. Good enough for the girls I go out with ;-)
One thing I might add as a owner of a Solar edge installation with optimizers is that it lets me check the output of the individual panels.
And not only that. If 1 panel stops working, the rest will still remain in service. Provided that the Optimizer is not defective. Then the system is out. But that is also the case with a regular string. What's better now? That extra cost? For me, yes.
And what do you change for the better when you have those values? Data without action has no value in my world. And often it steels the sleep ;-)
More panels or less shadow would change something, for example.
But I know, we are nerds and love our gadgets. So no problem with your decision.
@@AndreasSpiess You can identify a faulty panel. I did not decide for installation of these optimizers. They were included by our consultant for the solar installation. Back then I did not have the knowledge I have now and I might decide against the installation of these. Tho I did not have a problem so far and even if I would have one, the replacement of these modules is very easy, at least in my installation.
@@Iceteavanill The chance of defective panel is 1in 20. The cost of optimiser is same as a 335W poly panel. It is not worth it
@@shrujanamsyama9940 You can also detect degrading of the panel and determine the influence of obstacles on your roof. Idk there are pros and cons. If you want em install, em. If you dont, dont....
I came here for cat video's. That took a while in this particular video, but I got what I wanted. So, I like! 🙂
:-))
Thank you for this great video and your tests. This is exactly my situation and answers my questions! Greetings from Vorarlberg!
Glad the video was helpful!
Very informative, great approach. What we need now is a bigger sample set. As you say, the optimiser may have a 2% energy cost and you've robustly debunked the necessity of them in a string. However, could they potentially have enough benefit to counteract the energy cost? There are of course other benefits, not least the per-panel monitoring.
The reality is, they aren't as prohibitively expensive as made out. Our PV quotes ranged from £14,000-16,000 - the additional cost of optimisers, perhaps £500-600 is significant but almost in the noise. We ended up picking a middle ground option, that was SolarEdge. No regrets, a very efficient inverter, and as I say, amazing to be able to monitor each panel's performance.
So your choice was good! After more than a year of using our panels, I am no more interested in the details, more on the overall situation.
for long time i was looking some one compare it! and yes i decide do not use those! thank you Andreas!
You are welcome!
Wow. So helpful! Love your narration and information. Thank you for your work!
My pleasure!