Please please please redesign it and make it more accessible. Bringing the display to the 21st century would also be nice. Keep making these great videos!
You can find much simpler designs on eevblog forum to base off of. All you really need is: micro, two smart FET outputs, NTC cold junction, and a 24-bit ADC. Makes sense its overly complicated, because unisolders idea was to do everything.
I, too would love to see a simpler design! I built a Unisolder and it was a challenge for me as electronics are not something I know a lot about. I would like to have the capability to use tweezers.
Yes please redesign!! I have the skills and tools to build one. I did get some PCBs made but when I started the BOM list and saw the cost and the availability issues I tabled this project. Once again thanks for your work!!
It’s been 84 years (as Rose would say) So happy to finally see it finally up and running. I really really hope you make the redesign, or even better, we could do a shared collaboration design (you and us, the viewers). Still watching since 2016, Anton :)
I've looked at the Unisolder on many occasions and every time i just think "what a mess", there are design decisions throughout that make no sense and the entire thing is overly complicated, it definitely needs a proper going through. I think a lot of the complexity comes from the idea of cost reduction or being truly universal, unfortunately the cost reduction steps are misplaced, instead of thinking "what is the cheapest way to do x" the better thing would be "does it need x?" Or "can i combine everything to do x for cheaper". For example, replacing the PIC32 with a fair more beefy micro, like the STM32H743 will increase the cost of the micro, but the built in ADC and reference is orders of magnitude better than the PIC so no externals are needed and it also includes a sufficiently good DAC too. Or using a SEPIC topology input converter will get you buck-boost functionality over a very wide input voltage and can easily generate auxiliary rails eliminating a whole slew of components. Being truly universal is nice, but it means that if you want to use one iron you are paying for parts/configuration to use all irons, potentially put the actual iron interface stuff on a small separate board with the footprint/land pattern for the OEM connector, that way you don't have to modify your handpiece and you only need the components you need. Obviously this isn't a full design review kind of thing, it's just about the station itself but basically it comes down to this; the unisolder is expensive, complicated and not user friendly at this point, what advantages does it have over other stations? I can't think of any.
The author did have rationale behind the PIC32, which I generally agree with. He plans to update the next design to a Microchip PIC32MZ devices now which offers more performance for some of the floating point calculations that are going on. Generally speaking I find MPLAB easier to work with for many reasons. The built in PIC simulation for debugging and step through code without any hardware or PCB is really handy and makes code development very rapid. I don't think any of the other tools offer this so it makes switching to another ecosystem a bit more painful. I think the only advantage this design has over any other is that it is actually very good in terms of accuracy and control as a result of some of the high end components used.
@@sdgelectronics I can definitely agree with personal preferences for IDEs and development workflows, such as the very tightly integrated solution of MPLAB, I like that you spend a minimal amount of time "making" the tool, compared to the amount of time using it, unlike a lot of other solutions. That said, I think most of the other stuff is pretty moot, with the low cost and very wide availability of STs basic development boards I don't think the simulator is *that* big of a deal, RS or Farnell/Newark/e14 have pretty much all of the devboards in stock for next day delivery if you need to test it on your exact platform and looking at the general cost of development of the Unisolder project the cost of a nucleo is just a drop in the ocean, and yes, switching to another ecosystem can definitely be painful but it's one of those things you just learn to deal with, over the past year I have migrated a product to a different environment I was previously unfamiliar with and am currently in the process of doing a second, YAY Semiconductor shortage! Just stating I am aware this project began long before the release of the STM32H743, however my are points still valid as there are/were many devices with comparable feature sets at the time. My main point for suggesting against the PIC32MX had nothing to do with processing power (even though that is an area where the STM32H7 would trounce it), it was the quality of the analog and mixed signal peripherals, using them instead of the external devices removes 4(?) line items from the BOM leading to an overall component count reduction, improved performance and lower price. There are definite time tradeoffs for learning a new environment, but I think it works both ways, development time can be reduced because you are now using internal peripherals for some functions so you don't need to write and test external interface drivers for them, board assembly is now (slightly)quicker and hardware troubleshooting is similarly (slightly) quicker. I'm definitely on the fence about accuracy and control, as there is so much calibration and configuration left up the the user there is potential for better accuracy and control but i highly doubt that temperature calibrations and PID loop tuning is carried out to the extent needed to actually achieve it. On the other hand though is the need to carry out this tuning and calibration, it doesn't matter what components you use as long as they don't drift (which can be attributed to component selection or just design) and there is the actual practical accuracy/control, many tips have a significant amount of thermal resistance between the heating element and the working surface (looking at every single cartridge included with the aixun t3a, not just because of the 3rd party nature but also geometry of the tips) so the feedback you get from your temp sensor and your loop control is only really an indication of the actual tip temperature. As far as control, no super fancy components are really needed, just FETs so I don't see any huge benefits there. Also, there is one other thing that I can't really put in any technical words, the boards are just ugly, I don't know why, I can't put my finger on any one thing but it just looks bad. I don't mean to tear apart some random bloke on the internets design in some youtube comments but I don't particularly feel like getting lost in the thread(s) on dangerousprototypes or EEVblog, also I know this isn't your design (any of your boards look much cleaner/neater/nicer than this), I just thought I'd mention some things about this that bug me, if you were potentially to do your own controller/version it would possibly be worth while fixing them.
Never knew someone had this brilliant and awsome idea. Sure wish it had evolved already to the point that people like, with only basic knowledge of electronics, would be able to build it. If it was available as a commercial product product, this is what I would buy! Wonderfull!
Im pretty sure you can buy it assembled on tindie or the like Although it's not cheap, and you could probably get a good used professional soldering station for the price
@@alexstone691 I think the real appeal of the Unisolder is that it can run the most desired handles and accessories. 245, 210, Pace, and tweezers. That covers a lot of ground. I think the cheapest comparable JBC unit is the DIU at $570. And that's just the unit. No handles or accessories. Obviously the used market is different, but I doubt a used DIU would be $275, plus I can't even find a used unit for sale as of this post. The Unisolder is $150/128 Eur shipped for the assembled boards and you get a handle. You should definitely be able to get a transformer and casing for another $100, if not less. The Hammond 1182P12 (for US users) is $61, and the 230v ac INDEL TST 120/003 (found on TME) is even cheaper at 25 Euros. There are some great pictures of Unisolder casings people have built that didn't cost them anything. Scavenge, get creative, but it's definitely possible to build a Unisolder for around $200 US, or 150 Euro (before tax). That looks impossible to beat. Just don't buy the boards from Tindie, there are quality issues. Get them from the Dangerous Prototype web page.
@@Galactus314 Thank you very much for all your information. Will check it all out for sure. Not afraid to build. I am an old time RC airplane scratch builder. :-) I am able to handle most shop-tools small and large quite well. I will go looking for all the main parts to try and put one together. Thank you again !
Nice, and awesome that you're putting in your time to make this more accessible. Having source and buy extras of a part adds to the cost of a project very, very, fast. I ran into parts nightmare trying to source components to build a GPIB disk drive and a Pi Radioberry. Having to figure out replacements for the now unobtanium parts was a real pain. As to the Radioberry I already have the PCB and all the components except for the FPGA which Intel says they won't ship until December 2022 - yes, 2022 - it was Feb 2023.
We face even leadtimes of 115 weeks for some components. Most components are out of stock till next summer. My designs are altered with the availability with our supplier. Sometimes we pay 3-20x the price :P Goodluck with sourcing!
@@FrankenProjects That sucks. I'm on a kickstarter where the maker asked whether anyone who joined the campaign has excess of certain types of components they would be willing to part with at a reasonable price. I'm lucky because most of the stuff I do in my hobby is retro using components most people aren't interested in anymore. Hasn't stopped the sellers of obsolete parts jacking their prices up too.
I had the exact same thought about redesigning the PCB, however as an EE student i have yet to design a PCB. And taking on something like that seems to be a bit of a big bet for me, not only because of time… I am very much interested in how you would go about the design. Maybe you could even make a series out of it and explain your reasoning behind some decisions. That would be super interesting. I’d guess that Sparky might be interested in this too :).
I just had to check, I started designing PCBs in 2016, when i was 16 years old. I tried etching PCBs at home (with varying degrees of success) and started designing 2 Layer PCBs after that, ordered from JLCPCB. Some of my first PCBs i ever made include Arduino Dev Boards, simple audio amps with some passives and a LM386, PWM Load controllers, electronic loads and power supplies and more. In 2019 I got a job at a company that makes cnc routers and controllers, laying out PCBs for their controllers. It's never too late to start your own design. There are way enough learning resources available. And EasyEDA is, well, easy. Good Luck with your career! And good luck soldering!
@@tollertup Thanks for the kind words and the tips. Back in 2016 I didn't even know that I once would go study, let alone studying EE. I did have an interest in electronics back then, but that was more computer stuff than anything else. I even worked in a whole different field than electronics. But since then my interest in electronics grew and I ended up starting my EE studies. For now, I have to nail down my basics, then I can start with designing circuits and pcbs. But first I need to get to know the basic active and passive components, that's on the radar for the fall/spring semester. Luckily, I can say that the soldering part is definitely not a problem for me. A skill that I envied people a long time for, but now I have it down too.
@@Galileocrafter Keep it up, I'm sort of studying EE too, though it's not called that. Been designing PCBs and learning from others' work for quite a long time now. I'm wondering if I should try to redesign a driver from the ground up, make minor changes on the current design or just build this existing one that just works.
@@AlanDike The electrolytic capacitors in the power section are cheap chinese brands (so reliability/meltdown concerns), the mosfet driving the tip has no protection if it fails (which will ruin the tip, but possibly addressed in the newer versions), and while the unit itself is safe to solder SMD components, there is no ESD protection for the unit itself (ESD can damage the unit). Operator and component safety aren't a concern, it's safety for the unit itself. Check out the EEVblog thread on the T3A.
(Seems previous comment got deleted due to link? ill try again) I have been working on designing a multi-iron jbc station for a bit now. My plan for the stands was to buy replacement tool stands/holders for the compact line directly from JBC. They come in around 30Euros, and are the same piece that actually holds the iron in the standalone JBC stand. They are much cheaper than a whole JBC stand (which comes in at 150Euro+), and it should be fairly easily to mount them to something to create a very usable stand that is fairly cheap. Love the videos!
RUclips normally filters stuff out, especially links. Not sure how it decides what's spam or not. Good to know about the tool holder, thanks for the info.
@@sdgelectronics Should be pretty good, but i havent tried it. The same holder should work for all 210/245/470 style handles. They also sell a holder piece for am120 tweezers but annoyingly not for the AT420 tweezers which i was planning on using. They are not that expensive, so maybe i'll just bite the bullet and buy the am120 holder part and see if the AT420 will fit.
I'm not sure how much time you want to spend changing the design, but it would be really nice to have something like a 2.4'' touchscreen lcd for something like this. Other things that could be changed (besides the things you already stated) in my opinion: - Calibration in the menu, not with a trimmer. - Some better way to set the type of handle, maybe even in the menu. - Possibility to have 2 types of handles on there at the same time. So a T245 and T210 for example. - Possibility for preset temperatures
Love these type of DIY videos. Thanks and hope you can progress with it. I really like that you can make it work for tweezers. Hope to see that set up in one of your future videos. Big thumbs up
Would love to see a redesign as video. Seeing the way you progress through the entire process would be interesting as well as the finished product. Your content & format is excelent, keep up the great work!!
Also, if you could walk us through the process and explain the rationale behind the design choices it would also be extremely interesting. Could be an entire video of its own.
@@sdgelectronics it does not matter, you can base it upon experience as well, from what you have seen in the field. Would be interesting to hear your insight.
Any progress with the re-design? Sure, there are now more and more cheap JBC clones popping up - but cheap caps seems to be only one issue. An open custom design would make it also much more repairable. And it would be great to support all the original JBC handles.
My first impression is that the project is incredibly complicated for what it is. I did just watch Marco Reps video on his JBC controller, in contrast to this, his design is elegant and simple.
Great video and looks like a fun project. I have a fascination with soldering irons and this looks like it’s right up my alley. I just recently picked up the T3A with T245 handle on your recommendation and it looks like the Chinese tips that came with it my be off as far as temp goes. Wonder if my money would be better spent buying more tips for my Hakko FX-100 induction station and genuine JBC tips for the T3A instead of this.
Interesting project! The overvoltage protection part on the schematic around the DC converter isn't quite obvious to me. Some annotations to the schematic would be really nice.
oh, your idea to redesign sounds really good :D I've build a similar solder station to the one MarcoReps build on his channel. And it does work nicely, although it's only pumping 50 to 70W into my T210 handpiece. And that takes a good time before reachting set temperature. If you were to redesign this project, i would certainly build it!
Soo I did try to make some changes to the design and software. But I halted it due to lack of clarity and support. Looking back at it, maybe I could try it again. But buying a JBC station was much easier at the time :P
That's partly why it took a bit longer to do this video than I'd expected. There's some basic steps that are not clarified anywhere, even in the UniSolder notes Github.
As a guy that does PCB design for a living, this PCB is definitely not optimized for any kind of assembly at all, and the schematic is hard to follow. Lots of stuff could have been designed simpler for a shorter and cheaper BOM, and much of the power supply and analog stuff looks unnecessarily complicated to me for what it is. In my opinion, it would also be better to divide the schematic into more pages and let each handle "one feature", such as voltage regulators, MOSFET output stage, and amplifier for temperature reading on separate pages.
In terms of performance form my experience the reason for a cartridge based handpiece is primarily to omit needing heater or control loop calibration when changing tips. Next would be handpiece temperature (keeping the handpiece cool), contol loop response, grip to tip minimum distance, and then easy tip replacement (essentially heater replacement if the tips last as advertised). So it is interesting that the hobbiest approach promotes the option of heater and sensor calibration. Which would go against the feature evolution from many soldering station cartridge based manufacturers. Also, in production the soldering personel are provided with equipment and training to accomplish manufacturing specific soldering tasks as per the production engineer supervisor. You probably already know the following considerations, but I just wanted to give you some more input. Your high thermal stress test soldering board with thermal vias is not really representative of what works in production, please see the following videos: ruclips.net/video/d9DeitwDRGM/видео.html at time 10:01 ruclips.net/video/77frEufFv74/видео.html at time 4:21 The option for thermal vias is to have many very small vias that are filled and tinned (which is expensive) or to seperate larger vias and the power tab connections, like in the footprints in the videos. Your fooprint for the D2PAK with not work in SMT reflow soldering because the power tabe solder will wick down the holes. Also, vias that are not used as possible wire attachment (usually options for prototype emergency jumpers (wire)) should be solder masked. Also, intermal planes have a large keepout around a thermal via grouping, meaning you will not see seperate holes in the planes for a thermal via. Internal planes are set back with a keep out. For the DIP packages any power plane or copper pour that can conduct heat will have thermal relief on the footprint pad.
Nice video! Redesigning and partial assembly seems like a great idea! I wonder if the Unisolder 5.2c is any better? It does have a way to attach a rotary encoder. Unisolder seems interesting at first but, IMO, it's far too complex. Sort of like trying to build an airplane in your garage from a set of plans. Virtually everyone has problems getting it to work. I'd like to see simple, separate control boards for some specific irons. Most people make this for JBC 245 iron. Conversely, almost no one is going to build this for a 936 or FX888 iron. Also, OLED displays are over-rated. They suffer from burn-in and fading with age.
I think my design would be for the JBC irons only, since we now know they are the best type to use using resistive heaters. I'll have to see if any microcontrollers have a TFT driver and display RAM built in
I'm really not sure why you'd need distinct AC and DC power inputs? Why wouldn't you just use only the AC input given there's a bridge rectifier in there? I suppose you have to think about the soldering iron tip "ground" reference for ESD (and maybe sensing in-the-holder purposes) vs a DC power supply?
seems like feature creep is precisely why this project is bad. Too many features, too many parts, too difficult to build, too much to go wrong, expensive, ... Like why on earth does it have usb isolation?? It just makes no sense. It seems like a good thing to be sold prebuilt and tested, though.
i have same offset with temperatures on thermometer lower 40 - 60 degrees from info on display , right themperatures only with calibration values 920 not 1000
Hi, great project. I recently finished building it and getting it up and running. Unfortunately, all the time the display shows me that it is running on 1/4 power, and the max that was reached is 63W. Is it possible to increase this somehow so that the soldering iron heats up faster? Could it be a matter of an old software version?
It had been a work in progress for too long. The situation with the passive components makes assembly a bit tedious, that's why I would like to update it and have JLCPCB do all the passives at least.
Always wanted to build one of these. But bought the dde instead a year ago. Still a fun challenge though. One thing I’m not sure your aware of is your kaisertech code doesn’t work. Not sure why. 👍🏻
@@sdgelectronics Not worded the original comment correctly, i would use a separate winding or decoupling caps and then use a linear reg as in my experience switching regs dont last that long usually cause of caps Also would make things a bit simpler, but yeah im used to avoiding smd stuff as they are usually impossible to get with local dostributers here
sane here i have the blank pcb but getting parts and putting it all together is a pain. it sits on my desk for ever now and i just dont have time to search around everywhere for all the parts, there is a china made version over on tindie but many says the quality is not great and they have had issues with it also the builder is using some fake parts etc. the newer 5.2c does use a 4 layer pcb as opposed to the 5.2 is only 2 layer normally. Id be in for a couple maybe three though
The PIC isn't a bad choice, it's about the same price as the equivalent STM32 and has better documentation etc. At least for me I was able to download and build the source code in just a few minutes without any hassle, so it seems to work OK.
The unifolder feels really overly complicated in some parts. A simpler version one could just order with some/most parts pre-soldered would be so nice. I am also not sure why it uses a PIC over a STM32 or even a basic AVR.
The PIC32 is pretty good and has the advantage that it's readily available and not subject to stock issues like we see with the other types. Some people have their preferences, but overall it makes little difference what device is used as it's all written in C anyway.
@@sdgelectronics It sure is about what we are used to, but unless you are talking e.g. digikey it feels more exotic - at least to me. Or my aliexpress search skill just suck :) It's always fun to try something new, but if you want to make it more approachable it might be good to pick something where people might already have the programmer. *shrug* From the code it certainly would not matter much - and as long as there is good cross platform OSS support (via platform io for example) I agree, it probably does not matter too much.
I'd really love a video where you test this station with some Hakko T12 tips. I'm curious if the pour T12 performance during your testing was the tips (Not being genuine, if i remember correctly. Please correct me if i'm wrong about this.), the station or just the fact that the T12 tips is a bit underpowered compared to other systems you've tested. My money is on the underpowered soldering system, but it would be an interesting test for sure. I really enjoy your soldering iron videos, they give me some ideas about the next soldering irons in the office.
@@alexstone691 Well the tip is just a resistive heater. Even when plugged straight to 24V it will draw as much (or as little) power as it can. So a 50W heater connected to a 150W PSU will still only draw 50W. My question about the tip/station is aimed more to test if the original Quicko station does not have a control loop problem or a PSU problem, where the power supply can not deliver as much power as the tip can handle. I believe that when he tested the Quicko station, he used his bench top power supply, so that point would be mute. Still a nice test of the control loop.
@@TheAmmoniacal I saw your unanswered question on the forums. If it's reading room temperature and constantly dumping power into the heater, check the Sense A and Sense B connections. Did the calibration with a 10 ohm resistor work when no handpiece was attached?
Have you calibrated your temperature measurement? I tried doing something similar with a thermocouple once and I found the thermal coupling was unreliable. Just a bit of oxidation made it read way low, and it required some pressure to get good heat transfer.
The genuine JBC ones are repeatable. I suspect the low cost tips do not have reliable thermocouple contact - it's well known that the T12 cartridges for example are very fussy in this respect too.
Maybe You can give some bords for some of your subscribers with the components . A content maybe olso this is a very good alternative for Jbc and is much cheaper .
The jc programmer auxin t3a 200W u have reviewed has a similar functionality of supporting multiple heaters/irons (t12, c245). How does unisolder compare with the t3a in terms for circuit protection measures, accuracy & precision & features?. I'm especially interested in the protection measures because chinese always skimp on protection measures.
I think more could be done in the firmware or even in the hardware to reduce the chance of a total burnout of your soldering tip. One benefit of the UniSolder is two FETs have to fail for a hardware failure to cause the heating element to be constantly powered.
@@sdgelectronics yes indeed we were discussing this matter on the eevblog forum on recent aixun t3a thread )- indeed having 2 mosfet in series seems the best option. as a compromise between the safety and the added cost, it's really very unlikely that both will fail closed circuit. this is one of the specific features which might be desirable from the unisolder design. (without going too much overboard! ahem)
BTW good job getting the unisolder working. and being able to demonstrate it's full performance. you have conquered the magical mythic beast well done for that. quite surprising really given how complicated it seems to be. everything but the kitchen sink
Unfortunately not. There's a lot of parts on here that are unnecessarily hard to get hold of. It really needs an optimisation, but I don't have time to re-do the unisolder PCB as it stands. I do plan to design my own though specifically for JBC handpieces.
@@sdgelectronics Thank you! Really love all these videos. Btw if you do take this project on and want to update/imporve it i would also put ether put a STM32 or ESP has the micro. There is a lot that needs to be addressed if this open source sodlering station were to go mainstream. If you want help make sure to ask your community!
Can please make a review on this soldering station please OSS T12-X Soldering Station electronic welding iron LED Digital display BGA Rework Station With Soldering Tips Welding tools From Aliexpress its very awesome 1
Yeah I immediately dismissed this project, it seems overly complex (for hobbyists) and is a pic micro meh.. I would love to see this made more modular or use more off the shelf modules for ps, display, and a matrix for the handles instead of having to reference all the resistor id mods etc.
Please please please redesign it and make it more accessible.
Bringing the display to the 21st century would also be nice.
Keep making these great videos!
I’d love to see a redesign! A tweezer-capable t245-only version would be awesome.
You can find much simpler designs on eevblog forum to base off of. All you really need is: micro, two smart FET outputs, NTC cold junction, and a 24-bit ADC. Makes sense its overly complicated, because unisolders idea was to do everything.
I agree with you. Maybe I should have said: “I’d love to watch Steve design one!”
I, too would love to see a simpler design! I built a Unisolder and it was a challenge for me as electronics are not something I know a lot about. I would like to have the capability to use tweezers.
Will be interesting to see how you design the case for it.
Yes please redesign!! I have the skills and tools to build one. I did get some PCBs made but when I started the BOM list and saw the cost and the availability issues I tabled this project. Once again thanks for your work!!
It’s been 84 years (as Rose would say)
So happy to finally see it finally up and running. I really really hope you make the redesign, or even better, we could do a shared collaboration design (you and us, the viewers).
Still watching since 2016,
Anton :)
Do you want me to send you the remaining components?
@@sdgelectronics yeah, why not. Let’s take it to the e-mails :)
I've looked at the Unisolder on many occasions and every time i just think "what a mess", there are design decisions throughout that make no sense and the entire thing is overly complicated, it definitely needs a proper going through.
I think a lot of the complexity comes from the idea of cost reduction or being truly universal, unfortunately the cost reduction steps are misplaced, instead of thinking "what is the cheapest way to do x" the better thing would be "does it need x?" Or "can i combine everything to do x for cheaper". For example, replacing the PIC32 with a fair more beefy micro, like the STM32H743 will increase the cost of the micro, but the built in ADC and reference is orders of magnitude better than the PIC so no externals are needed and it also includes a sufficiently good DAC too. Or using a SEPIC topology input converter will get you buck-boost functionality over a very wide input voltage and can easily generate auxiliary rails eliminating a whole slew of components.
Being truly universal is nice, but it means that if you want to use one iron you are paying for parts/configuration to use all irons, potentially put the actual iron interface stuff on a small separate board with the footprint/land pattern for the OEM connector, that way you don't have to modify your handpiece and you only need the components you need.
Obviously this isn't a full design review kind of thing, it's just about the station itself but basically it comes down to this; the unisolder is expensive, complicated and not user friendly at this point, what advantages does it have over other stations? I can't think of any.
The author did have rationale behind the PIC32, which I generally agree with. He plans to update the next design to a Microchip PIC32MZ devices now which offers more performance for some of the floating point calculations that are going on. Generally speaking I find MPLAB easier to work with for many reasons. The built in PIC simulation for debugging and step through code without any hardware or PCB is really handy and makes code development very rapid. I don't think any of the other tools offer this so it makes switching to another ecosystem a bit more painful.
I think the only advantage this design has over any other is that it is actually very good in terms of accuracy and control as a result of some of the high end components used.
@@sdgelectronics I can definitely agree with personal preferences for IDEs and development workflows, such as the very tightly integrated solution of MPLAB, I like that you spend a minimal amount of time "making" the tool, compared to the amount of time using it, unlike a lot of other solutions.
That said, I think most of the other stuff is pretty moot, with the low cost and very wide availability of STs basic development boards I don't think the simulator is *that* big of a deal, RS or Farnell/Newark/e14 have pretty much all of the devboards in stock for next day delivery if you need to test it on your exact platform and looking at the general cost of development of the Unisolder project the cost of a nucleo is just a drop in the ocean, and yes, switching to another ecosystem can definitely be painful but it's one of those things you just learn to deal with, over the past year I have migrated a product to a different environment I was previously unfamiliar with and am currently in the process of doing a second, YAY Semiconductor shortage!
Just stating I am aware this project began long before the release of the STM32H743, however my are points still valid as there are/were many devices with comparable feature sets at the time.
My main point for suggesting against the PIC32MX had nothing to do with processing power (even though that is an area where the STM32H7 would trounce it), it was the quality of the analog and mixed signal peripherals, using them instead of the external devices removes 4(?) line items from the BOM leading to an overall component count reduction, improved performance and lower price.
There are definite time tradeoffs for learning a new environment, but I think it works both ways, development time can be reduced because you are now using internal peripherals for some functions so you don't need to write and test external interface drivers for them, board assembly is now (slightly)quicker and hardware troubleshooting is similarly (slightly) quicker.
I'm definitely on the fence about accuracy and control, as there is so much calibration and configuration left up the the user there is potential for better accuracy and control but i highly doubt that temperature calibrations and PID loop tuning is carried out to the extent needed to actually achieve it.
On the other hand though is the need to carry out this tuning and calibration, it doesn't matter what components you use as long as they don't drift (which can be attributed to component selection or just design) and there is the actual practical accuracy/control, many tips have a significant amount of thermal resistance between the heating element and the working surface (looking at every single cartridge included with the aixun t3a, not just because of the 3rd party nature but also geometry of the tips) so the feedback you get from your temp sensor and your loop control is only really an indication of the actual tip temperature.
As far as control, no super fancy components are really needed, just FETs so I don't see any huge benefits there.
Also, there is one other thing that I can't really put in any technical words, the boards are just ugly, I don't know why, I can't put my finger on any one thing but it just looks bad.
I don't mean to tear apart some random bloke on the internets design in some youtube comments but I don't particularly feel like getting lost in the thread(s) on dangerousprototypes or EEVblog, also I know this isn't your design (any of your boards look much cleaner/neater/nicer than this), I just thought I'd mention some things about this that bug me, if you were potentially to do your own controller/version it would possibly be worth while fixing them.
It does seem like a project that was plagued by feature creep.
a cost optimized redesigned would be amazing
Never knew someone had this brilliant and awsome idea. Sure wish it had evolved already to the point that people like, with only basic knowledge of electronics, would be able to build it. If it was available as a commercial product product, this is what I would buy! Wonderfull!
Im pretty sure you can buy it assembled on tindie or the like
Although it's not cheap, and you could probably get a good used professional soldering station for the price
@@alexstone691 I think the real appeal of the Unisolder is that it can run the most desired handles and accessories. 245, 210, Pace, and tweezers. That covers a lot of ground. I think the cheapest comparable JBC unit is the DIU at $570. And that's just the unit. No handles or accessories. Obviously the used market is different, but I doubt a used DIU would be $275, plus I can't even find a used unit for sale as of this post. The Unisolder is $150/128 Eur shipped for the assembled boards and you get a handle. You should definitely be able to get a transformer and casing for another $100, if not less. The Hammond 1182P12 (for US users) is $61, and the 230v ac INDEL TST 120/003 (found on TME) is even cheaper at 25 Euros. There are some great pictures of Unisolder casings people have built that didn't cost them anything. Scavenge, get creative, but it's definitely possible to build a Unisolder for around $200 US, or 150 Euro (before tax). That looks impossible to beat. Just don't buy the boards from Tindie, there are quality issues. Get them from the Dangerous Prototype web page.
@@Galactus314 Thank you very much for all your information. Will check it all out for sure. Not afraid to build. I am an old time RC airplane scratch builder. :-) I am able to handle most shop-tools small and large quite well. I will go looking for all the main parts to try and put one together. Thank you again !
Nice, and awesome that you're putting in your time to make this more accessible. Having source and buy extras of a part adds to the cost of a project very, very, fast. I ran into parts nightmare trying to source components to build a GPIB disk drive and a Pi Radioberry. Having to figure out replacements for the now unobtanium parts was a real pain. As to the Radioberry I already have the PCB and all the components except for the FPGA which Intel says they won't ship until December 2022 - yes, 2022 - it was Feb 2023.
We face even leadtimes of 115 weeks for some components. Most components are out of stock till next summer. My designs are altered with the availability with our supplier. Sometimes we pay 3-20x the price :P
Goodluck with sourcing!
@@FrankenProjects That sucks. I'm on a kickstarter where the maker asked whether anyone who joined the campaign has excess of certain types of components they would be willing to part with at a reasonable price. I'm lucky because most of the stuff I do in my hobby is retro using components most people aren't interested in anymore. Hasn't stopped the sellers of obsolete parts jacking their prices up too.
I had the exact same thought about redesigning the PCB, however as an EE student i have yet to design a PCB. And taking on something like that seems to be a bit of a big bet for me, not only because of time…
I am very much interested in how you would go about the design. Maybe you could even make a series out of it and explain your reasoning behind some decisions. That would be super interesting.
I’d guess that Sparky might be interested in this too :).
Definitely would love a series
I just had to check, I started designing PCBs in 2016, when i was 16 years old. I tried etching PCBs at home (with varying degrees of success) and started designing 2 Layer PCBs after that, ordered from JLCPCB. Some of my first PCBs i ever made include Arduino Dev Boards, simple audio amps with some passives and a LM386, PWM Load controllers, electronic loads and power supplies and more. In 2019 I got a job at a company that makes cnc routers and controllers, laying out PCBs for their controllers. It's never too late to start your own design. There are way enough learning resources available. And EasyEDA is, well, easy. Good Luck with your career! And good luck soldering!
@@tollertup Thanks for the kind words and the tips. Back in 2016 I didn't even know that I once would go study, let alone studying EE. I did have an interest in electronics back then, but that was more computer stuff than anything else. I even worked in a whole different field than electronics. But since then my interest in electronics grew and I ended up starting my EE studies. For now, I have to nail down my basics, then I can start with designing circuits and pcbs. But first I need to get to know the basic active and passive components, that's on the radar for the fall/spring semester. Luckily, I can say that the soldering part is definitely not a problem for me. A skill that I envied people a long time for, but now I have it down too.
@@Galileocrafter Keep it up, I'm sort of studying EE too, though it's not called that. Been designing PCBs and learning from others' work for quite a long time now. I'm wondering if I should try to redesign a driver from the ground up, make minor changes on the current design or just build this existing one that just works.
Would love a redesign series on this!
The Aixun is looking better and better. Some safety concerns, yes. But the performance, pricepoint, and simplicity look great.
what safety concerns exist for the axiuns?
@@AlanDike The electrolytic capacitors in the power section are cheap chinese brands (so reliability/meltdown concerns), the mosfet driving the tip has no protection if it fails (which will ruin the tip, but possibly addressed in the newer versions), and while the unit itself is safe to solder SMD components, there is no ESD protection for the unit itself (ESD can damage the unit). Operator and component safety aren't a concern, it's safety for the unit itself. Check out the EEVblog thread on the T3A.
Would be great giving it a redesign mainly so we can learn from your experience correcting the mistakes.
Great video. I've been waiting for him for so long.
The boss music starts playing when you unpack those PCBs.
(Seems previous comment got deleted due to link? ill try again)
I have been working on designing a multi-iron jbc station for a bit now.
My plan for the stands was to buy replacement tool stands/holders for the compact line directly from JBC. They come in around 30Euros, and are the same piece that actually holds the iron in the standalone JBC stand.
They are much cheaper than a whole JBC stand (which comes in at 150Euro+), and it should be fairly easily to mount them to something to create a very usable stand that is fairly cheap.
Love the videos!
RUclips normally filters stuff out, especially links. Not sure how it decides what's spam or not. Good to know about the tool holder, thanks for the info.
@@sdgelectronics Should be pretty good, but i havent tried it. The same holder should work for all 210/245/470 style handles.
They also sell a holder piece for am120 tweezers but annoyingly not for the AT420 tweezers which i was planning on using.
They are not that expensive, so maybe i'll just bite the bullet and buy the am120 holder part and see if the AT420 will fit.
I'm not sure how much time you want to spend changing the design, but it would be really nice to have something like a 2.4'' touchscreen lcd for something like this. Other things that could be changed (besides the things you already stated) in my opinion:
- Calibration in the menu, not with a trimmer.
- Some better way to set the type of handle, maybe even in the menu.
- Possibility to have 2 types of handles on there at the same time. So a T245 and T210 for example.
- Possibility for preset temperatures
I would love to see a redesign capable of 2 t245 handles or one handle and one tweezer (with a larger screen).
Yeah I agree being able to do T245 and a set of tweezers or T245 and a T210 etc set up this thing would be a one stop shop.
Yes, with open source code. I have wanted this for a long time, but I have too little knowledge about circuit design. But I can help coding :)
It’s ironic that you need a fairly good soldering iron to build the board to try to make a fairly good soldering iron.
But you can never have too many good soldering irons.
It’s solder-ironic
Love these type of DIY videos. Thanks and hope you can progress with it. I really like that you can make it work for tweezers. Hope to see that set up in one of your future videos. Big thumbs up
I’d love to see an induction-based station design. We need an open project with Metcal performance!
Would love to see a redesign as video. Seeing the way you progress through the entire process would be interesting as well as the finished product.
Your content & format is excelent, keep up the great work!!
This is becoming the "Solderings Dam Good" channel ;-)
A redesign would be interesting to see. Maybe with another microcontroller like a stm32?
Don't bother. Aixun T3A/B is currently the best bang for the buck.
Would love to see you take a crack at a properly designed 13.56 Mhz station. Especially since Metcal profit ratios seem at least as large as JBC.
Also, if you could walk us through the process and explain the rationale behind the design choices it would also be extremely interesting. Could be an entire video of its own.
To be honest, I don't understand all of the rationale, but then there are thousands of ways to do the same thing.
@@sdgelectronics it does not matter, you can base it upon experience as well, from what you have seen in the field. Would be interesting to hear your insight.
Any progress with the re-design? Sure, there are now more and more cheap JBC clones popping up - but cheap caps seems to be only one issue. An open custom design would make it also much more repairable. And it would be great to support all the original JBC handles.
My first impression is that the project is incredibly complicated for what it is. I did just watch Marco Reps video on his JBC controller, in contrast to this, his design is elegant and simple.
Great video and looks like a fun project. I have a fascination with soldering irons and this looks like it’s right up my alley. I just recently picked up the T3A with T245 handle on your recommendation and it looks like the Chinese tips that came with it my be off as far as temp goes. Wonder if my money would be better spent buying more tips for my Hakko FX-100 induction station and genuine JBC tips for the T3A instead of this.
Nice soldering skills!
Impressive smd soldering skills 🗿💯
I'd love to see it able to run a hot air station in addition to an iron.
Great video....cheers.
Make sure nothing blows up, he says, while using the intrinsically safe fluke. xD
Очень интересный проект. Спасибо за подробное видео.
great video as always
Interesting project! The overvoltage protection part on the schematic around the DC converter isn't quite obvious to me. Some annotations to the schematic would be really nice.
they also released a new axiun t3b station now for 115 and 210 tips but its only 96w max according to the company.
oh, your idea to redesign sounds really good :D
I've build a similar solder station to the one MarcoReps build on his channel. And it does work nicely, although it's only pumping 50 to 70W into my T210 handpiece. And that takes a good time before reachting set temperature.
If you were to redesign this project, i would certainly build it!
If it could drive T470 cartridges, it would be the bees knees.
I wonder where this project's at today, it's been a few years. :)
Thank you for the video!
hmm interesting i might just have to give this a go. I think you should fork it and fix it. It seems a few of us were watching this had the same idea
You could just use the Axion stand (or something similiar) for the cradle.
Can you run two irons? An iron and a desolder iron? I’d love to see an open-source desoldering station with JBC tips!
Nice video sir kindly request check out i2c soldering station from china the soldering station operate with all kind of jbc tip
Hi sir i have a question kindly explain how does t3a soldering station identify c245 or t12 is inserted in the station
Soo I did try to make some changes to the design and software. But I halted it due to lack of clarity and support.
Looking back at it, maybe I could try it again. But buying a JBC station was much easier at the time :P
That's partly why it took a bit longer to do this video than I'd expected. There's some basic steps that are not clarified anywhere, even in the UniSolder notes Github.
As a guy that does PCB design for a living, this PCB is definitely not optimized for any kind of assembly at all, and the schematic is hard to follow. Lots of stuff could have been designed simpler for a shorter and cheaper BOM, and much of the power supply and analog stuff looks unnecessarily complicated to me for what it is. In my opinion, it would also be better to divide the schematic into more pages and let each handle "one feature", such as voltage regulators, MOSFET output stage, and amplifier for temperature reading on separate pages.
Nice video
In terms of performance form my experience the reason for a cartridge based handpiece is primarily to omit needing heater or control loop calibration when changing tips. Next would be handpiece temperature (keeping the handpiece cool), contol loop response, grip to tip minimum distance, and then easy tip replacement (essentially heater replacement if the tips last as advertised). So it is interesting that the hobbiest approach promotes the option of heater and sensor calibration. Which would go against the feature evolution from many soldering station cartridge based manufacturers. Also, in production the soldering personel are provided with equipment and training to accomplish manufacturing specific soldering tasks as per the production engineer supervisor.
You probably already know the following considerations, but I just wanted to give you some more input.
Your high thermal stress test soldering board with thermal vias is not really representative of what works in production, please see the following videos:
ruclips.net/video/d9DeitwDRGM/видео.html at time 10:01
ruclips.net/video/77frEufFv74/видео.html at time 4:21
The option for thermal vias is to have many very small vias that are filled and tinned (which is expensive) or to seperate larger vias and the power tab connections, like in the footprints in the videos. Your fooprint for the D2PAK with not work in SMT reflow soldering because the power tabe solder will wick down the holes. Also, vias that are not used as possible wire attachment (usually options for prototype emergency jumpers (wire)) should be solder masked. Also, intermal planes have a large keepout around a thermal via grouping, meaning you will not see seperate holes in the planes for a thermal via. Internal planes are set back with a keep out. For the DIP packages any power plane or copper pour that can conduct heat will have thermal relief on the footprint pad.
Have you heard of AxxSolder? It looks intresting, but also a a pain to build and source all the parts without a kit.
Nice video! Redesigning and partial assembly seems like a great idea! I wonder if the Unisolder 5.2c is any better? It does have a way to attach a rotary encoder.
Unisolder seems interesting at first but, IMO, it's far too complex. Sort of like trying to build an airplane in your garage from a set of plans. Virtually everyone has problems getting it to work. I'd like to see simple, separate control boards for some specific irons. Most people make this for JBC 245 iron. Conversely, almost no one is going to build this for a 936 or FX888 iron. Also, OLED displays are over-rated. They suffer from burn-in and fading with age.
I think my design would be for the JBC irons only, since we now know they are the best type to use using resistive heaters. I'll have to see if any microcontrollers have a TFT driver and display RAM built in
Am i missing something or D4 would create a short in case of revered polarity dc? Also bypasses the fuse
Yes you're right!
@@sdgelectronics what is the tine stamp?
@@therealb888 1:45
I'm really not sure why you'd need distinct AC and DC power inputs? Why wouldn't you just use only the AC input given there's a bridge rectifier in there? I suppose you have to think about the soldering iron tip "ground" reference for ESD (and maybe sensing in-the-holder purposes) vs a DC power supply?
@@lmamakos the DC input is good for testing the board like was done in this video.
Support for the C470 (42V) tips would also be nice if you're doing a redesign
seems like feature creep is precisely why this project is bad. Too many features, too many parts, too difficult to build, too much to go wrong, expensive, ... Like why on earth does it have usb isolation?? It just makes no sense.
It seems like a good thing to be sold prebuilt and tested, though.
Nice video, thanks :)
Two years ago I successfully compiled it with X32 version 2.10 no need to use the older V1.30
i have same offset with temperatures on thermometer lower 40 - 60 degrees from info on display , right themperatures only with calibration values 920 not 1000
Hi, great project. I recently finished building it and getting it up and running. Unfortunately, all the time the display shows me that it is running on 1/4 power, and the max that was reached is 63W. Is it possible to increase this somehow so that the soldering iron heats up faster? Could it be a matter of an old software version?
I’m surprised there is a separate DC input, you can feed DC into a AC input as it will pass through the diodes just fine.
If you pass through the diodes with DC you are wasting power
Just messed up building materials and having it correctly markings on the PCB kills that project for me too much work
It had been a work in progress for too long. The situation with the passive components makes assembly a bit tedious, that's why I would like to update it and have JLCPCB do all the passives at least.
thanks for share ! ;-)
Always wanted to build one of these. But bought the dde instead a year ago. Still a fun challenge though. One thing I’m not sure your aware of is your kaisertech code doesn’t work. Not sure why. 👍🏻
Steve, can you tell something about value of type sense resitor
I would personally prefer to use the transformer for the logic power also
Would make it a bit simpler and possibly more robust
It does use the transformer as the supply for everything. No additional power supply needed.
@@sdgelectronics Not worded the original comment correctly, i would use a separate winding or decoupling caps and then use a linear reg as in my experience switching regs dont last that long usually cause of caps
Also would make things a bit simpler, but yeah im used to avoiding smd stuff as they are usually impossible to get with local dostributers here
sane here i have the blank pcb but getting parts and putting it all together is a pain. it sits on my desk for ever now and i just dont have time to search around everywhere for all the parts, there is a china made version over on tindie but many says the quality is not great and they have had issues with it also the builder is using some fake parts etc. the newer 5.2c does use a 4 layer pcb as opposed to the 5.2 is only 2 layer normally. Id be in for a couple maybe three though
I heard some people had fake components on those boards too, so it may not be the best choice.
@@sdgelectronics i was reff to your versions. id be in for a couple at least for both my setups
Need ultrasonic add-on, so we aren't paying $7,000 for a Sunbonder, Japan-Unix, etc. handpiece driver-box! We all wanna solder to glass!
I don't know about PIC anymore, seems like going STM32 is the go these days, ie, easier.
Shame that guy didn't go iBOM, it's much nicer.
The PIC isn't a bad choice, it's about the same price as the equivalent STM32 and has better documentation etc. At least for me I was able to download and build the source code in just a few minutes without any hassle, so it seems to work OK.
Do you know if the Aixun can be setup with JBC hot tweezers? Thank you :)
The unifolder feels really overly complicated in some parts. A simpler version one could just order with some/most parts pre-soldered would be so nice. I am also not sure why it uses a PIC over a STM32 or even a basic AVR.
The PIC32 is pretty good and has the advantage that it's readily available and not subject to stock issues like we see with the other types. Some people have their preferences, but overall it makes little difference what device is used as it's all written in C anyway.
@@sdgelectronics It sure is about what we are used to, but unless you are talking e.g. digikey it feels more exotic - at least to me. Or my aliexpress search skill just suck :)
It's always fun to try something new, but if you want to make it more approachable it might be good to pick something where people might already have the programmer. *shrug*
From the code it certainly would not matter much - and as long as there is good cross platform OSS support (via platform io for example) I agree, it probably does not matter too much.
Salve sto acquistando i materiali per unisolder ..
una informazione il display dove lo posso acquistare
I think the first step is to admit you have a problem Steve 😉
could this run a hakko fm2024-02 desolder gun?
I'd really love a video where you test this station with some Hakko T12 tips. I'm curious if the pour T12 performance during your testing was the tips (Not being genuine, if i remember correctly. Please correct me if i'm wrong about this.), the station or just the fact that the T12 tips is a bit underpowered compared to other systems you've tested.
My money is on the underpowered soldering system, but it would be an interesting test for sure.
I really enjoy your soldering iron videos, they give me some ideas about the next soldering irons in the office.
Im wondering if running the tips harder at idk 100w would not shorten their lifespan so much that they are still worth it at $2 a piece on aliexpress
@@alexstone691 Well the tip is just a resistive heater. Even when plugged straight to 24V it will draw as much (or as little) power as it can. So a 50W heater connected to a 150W PSU will still only draw 50W.
My question about the tip/station is aimed more to test if the original Quicko station does not have a control loop problem or a PSU problem, where the power supply can not deliver as much power as the tip can handle.
I believe that when he tested the Quicko station, he used his bench top power supply, so that point would be mute. Still a nice test of the control loop.
Ok, I'll order a connector to test it. I think the results will be the same as with the other systems though.
@@sdgelectronics Thank you. I also expect similar results, but it will be a nice test.
Keep up the good content :)
@@Nik930714 You increase the voltage and thhat increases the power
While i like the idea the pcb seems quite dense and wouldnt wish debugging that beast on anyone
I have one in my drawer that I debugged for a month before giving up. Might actually take another look at it now.
@@TheAmmoniacal I saw your unanswered question on the forums. If it's reading room temperature and constantly dumping power into the heater, check the Sense A and Sense B connections. Did the calibration with a 10 ohm resistor work when no handpiece was attached?
Have you calibrated your temperature measurement? I tried doing something similar with a thermocouple once and I found the thermal coupling was unreliable. Just a bit of oxidation made it read way low, and it required some pressure to get good heat transfer.
The genuine JBC ones are repeatable. I suspect the low cost tips do not have reliable thermocouple contact - it's well known that the T12 cartridges for example are very fussy in this respect too.
still waiting for next video
Maybe You can give some bords for some of your subscribers with the components . A content maybe olso this is a very good alternative for Jbc and is much cheaper .
If I manage to redesign it, I'll get a bunch of them made. I only have enough parts for this one sadly, they were all ordered about 2 years ago.
@@sdgelectronics ooo it took some time but in the end the rezolts are very impresive.
If you are in Australia, I have couple sets PCBs, which i can send if anybody is willing to cover the postage.
The jc programmer auxin t3a 200W u have reviewed has a similar functionality of supporting multiple heaters/irons (t12, c245).
How does unisolder compare with the t3a in terms for circuit protection measures, accuracy & precision & features?. I'm especially interested in the protection measures because chinese always skimp on protection measures.
I think more could be done in the firmware or even in the hardware to reduce the chance of a total burnout of your soldering tip. One benefit of the UniSolder is two FETs have to fail for a hardware failure to cause the heating element to be constantly powered.
@@sdgelectronics yes indeed we were discussing this matter on the eevblog forum on recent aixun t3a thread )- indeed having 2 mosfet in series seems the best option. as a compromise between the safety and the added cost, it's really very unlikely that both will fail closed circuit. this is one of the specific features which might be desirable from the unisolder design. (without going too much overboard! ahem)
BTW good job getting the unisolder working. and being able to demonstrate it's full performance. you have conquered the magical mythic beast well done for that. quite surprising really given how complicated it seems to be. everything but the kitchen sink
Wonder if JLC has all the SMD parts in their catalog.
Unfortunately not. There's a lot of parts on here that are unnecessarily hard to get hold of. It really needs an optimisation, but I don't have time to re-do the unisolder PCB as it stands. I do plan to design my own though specifically for JBC handpieces.
@@sdgelectronics The JLC thing is a boomer. Otherwise I hope you can come with a nice design. Thanks for your answer.
How much did you spend on the build so far?
Not sure actually, probably £150 with the transformer
Could you update your playlists so i can send people all the soldering irons you reviews you have done
Good plan, I've added a few new playlists
@@sdgelectronics Thank you! Really love all these videos. Btw if you do take this project on and want to update/imporve it i would also put ether put a STM32 or ESP has the micro. There is a lot that needs to be addressed if this open source sodlering station were to go mainstream. If you want help make sure to ask your community!
have u redesigned the PCB?
I haven't. There was a significant shortage of suitable components which stopped it being viable at the time.
@@sdgelectronics We look forward for your redesign when it become viable again, Thanks a lot
Can please make a review on this soldering station please OSS T12-X Soldering Station electronic welding iron LED Digital display BGA Rework Station With Soldering Tips Welding tools
From Aliexpress its very awesome 1
Yeah I immediately dismissed this project, it seems overly complex (for hobbyists) and is a pic micro meh.. I would love to see this made more modular or use more off the shelf modules for ps, display, and a matrix for the handles instead of having to reference all the resistor id mods etc.
The PCB design is all over the place, sadly