Since that middle pin is the ground, it's also connected to the ground plane, which acts as a heat sink. The solder doesn't stay hot long enough after you remove the iron for that sucker to work. Use solder wick instead and trash that sucker.
@@james10739 Works great for me most of the time, just need to use the correct sized wick with the correct sized tip for the job and a soldering iron that can keep up with how much extra heat the wick wicks away.
Hi Andy, loving your vlog. Thanks. It makes sense to have a diode across a motor to prevent reverse voltage from the inductance of the motor affecting the rest of the circuitry. I remember taking this precaution when powering an inductive load through a power transistor back in the day.
There wouldn't be any back EMF from that fan. Like PC cooling fans, they have a driver circuit built in, if you have ever played around with fan's, spinning the fan blades doesn't generate a current.. you would have to open the fan and bypass the driver circuit to get it to generate a current.. The diode appears to be a protective solution should you connect your sense wire's incorrectly...
@@PeterMilanovski So it’s odd the PCB may be labelled incorrectly. Maybe trying changing over those connectors? Though the fan would not be taking the current indicated on the display
@@michaelbouckley4455 I'm not sure what is going on with the device, it's possible that the fan connection is plugged into the sense socket and therefore giving the reading on the display? Or something else has failed which would need an oscilloscope and some datasheets to troubleshoot it....
@@PeterMilanovski I think you would you agree that for the diode to be a protective solution for the sense wires, the sense wires would need to be connected to the header with the diode. Ultimately, Andy did not connect it that way -- he connected the fan to the header with the diode, and the sense wires went to the header without the diode. My comment was in that context, meaning I was assuming he had them connected properly. I did not say that there WOULD be any back EMF from the fan -- I was only suggesting that it's a reason the designers may have put one on that header. Perhaps it's unnecessary. Or perhaps not every single fan in the world has a "driver circuit" as you assumed. We don't know without more info.
The midle pin is the ground plane, there is lots of cooper dissipating heat from the solder iron. I normally apply the heat in one side of the pcb and suck the solder from the other side. The diode is a zenner connected in reverse it becames a "voltage regulador"
I am surprised no one suggested using a round toothpick. Just heat the joint, push a toothpick in from the other side and then take the iron away. The toothpick will not absorb any solder and you can just pull it out and the hole will remain open. This method is also helpful if you have a lead broken off in the hole. You can use the toothpick to push out the broken lead.
You need to replace the chip that is getting hot. It is surely a single or dual OpAmp chip that drives the gate of the power MOSFET(s) that charges or discharges the load/battery (pin 1 of one of the big 3 pin transistors on the main heat sink). The OpAmp is powered by the +12V supply rail that you fixed and would be used to measure a low voltage DAC output analog set point voltage reference from a microcontroller, compare this reference to the current shunt voltage, then output an analog voltage in the 0 to 12V range to drive the gate of the power MOSFET such that the current shunt observes the same current that the microcontroller is requesting. Although you measured open circuit on the 7812 voltage regulator output, it evidently first failed by shorting the input to the output, thus delivering the full 30+V power supply voltage to the OpAmp and blowing it up too. The OpAmp then sucked down enough current to melt the silicon or bond wires in the 7812 such that it now appears open circuit. OpAmps are a common off the shelf component just like the 7812 is, and you may even find the exact same chip if you go around opening power supplies, your Victron charge controllers and inverter, or perhaps the other el-cheapo PWM charge controllers.
Option 1: stand the board up, place the sucker on the top side, touch the iron on the bottom side. Option 2: you can place the iron on the edge of the hole, place the nylon tip on the hole hit the button without pulling the iron away. You're to slow with the method you're using. The sucker tip is made to take the heat. Keep inside the sucker clean. To compound your problem is the lead free solder. It requires a higher temperature to melt and solidifies much faster.
Just keep your soldering iron on the solder pad while sucking with the pump! The pump's tip is made out of heat resistant plastic, so it doesn't melt. When ypu remove your soldering iron, the solder solidifies instantly especially on the middle pin, which is connected to the ground plane and has a huge thermal mass (acts like a heatsink sucking away your iron's heat immediately)
I would add few suggestion as you wanted :) 1. That diode is for protection of your driving circuit (fan) .it will not allow to push voltage in case you would start spinning that fan 2. it looks like normal board but there is huge ground plane and that will reduce your temperature 3. there are special tools that are used to suck solder on ebay (they look like plastic cheap pistol) ..they are not the best but work very well and and they have pump inside ..so yeah cheap option ..and worth some money ..if not go with Weller or similar brand 4. Probably some IC is shorted because your regulator blew up and allowed voltage to go above 12V or some spike with same results..and you might expect more IC went wrong same time.. 5. I would suggest to change those IC (if you will be able solder/desolder ) but i would suggest proper tools for that..solder, desolder , hot air gun
looks like the Op-Amp went short, causing extraneous readings on the microcontroller then the linear regulator (7812) blew up due to it overheating. You might be able to replace the Op-Amp but it might have broken some SMD resistors which that and the Op-Amp will be hard to replace with only a soldering iron. you need a hot air station! Its probably best to just get a new one, Good luck Andy!
Ok, my 2 cents. 1. Solder wick with liquid flux on the wick. 2. Flow the solder then push a round toothpick through the solder holes. (My favorite.) 3. Flow the solder and bang the board on the workbench. Laugh all you want but I've been a bench tech for over 30 years and those are my tried and true methods for solder removal. Also, when removing ICs, cut the legs and remove 1 at a time. Traces tend to get lifted otherwise. At least on cheap electronics they do.
Perhaps if you could identify the chip getting hot? As it seems to be voltage and current readings giving grief maybe it is a reference voltage that has gone south. Some a to d converters rely on an external reference voltage. If it went wrong maybe it also caused the regulator to overload as it failed. A good sign is the pat you are getting readings which implies that the a to d part of the processor is working so the problem is in the periferals and not a programmed part.
yea the op-amp amplifies the signal from the current shunt probably, so it going short made that reading go wild and pulled more current than the heatsink on the 7812 regulator could handle (though it should have overheat protection and shut down before that)
@@SuperBrainAK they do have overheat protection however, in reality, they are still quite prone to dying instead. I have replaced many that it could be argued shouldn’t have died.
@@philbrooke-little7082 yea, one day a long time ago I had a nice breadboard setup with 3 different linear regulators with a Teensy ++ that could turn them on and off and do fun things, Then an acquaintance decided it would be funny to mess with the wires and shorted the 5v rail of the always on regulator for the microcontroller. I came back and noticed the power supply was on but no lights were on at the breadboard, so I touched the heatsink to move the board and got burned by it. I finally found the short but the regulator was unable to deliver any current, it was surprisingly still able to regulate to 5v but couldnt even light an LED. RIP 7805
Use SOLDER WICK, also called "copper braid" or "desoldering braid". Put the wick on the cold dry solder and heat the solder through the wick. Alternatively, I sometimes use a wooden toothpick to "unclog" a PCB hole.
The fan diode is there to protect the fan power circuit from back feed current from the fan eg when you hit it with the air gun When you spin them up they can generate a lot of current
That would be true if the fan didn't have a driver circuit but in the case of this fan and all PC cooling fans, they don't produce current if you spin the fan blades, you would have to modify the fan, bypass the driver circuit to get it to generate a current...
@@awesomusmaximus3766 I have never been able to generate a current from one of those types of fan's by spinning the fan blades... They need to have the driver circuit bypassed in order to do so...
did you ever get to do a part 3? i have an ebc b20h which also shows overvoltage.. issue also came after a power cut.. which could have caused either charger input or mains input to spike.. i was hoping to salvage it with a few new components..
Tipp: use a tooth pick to get a hole in the solder pads when the sucker is not working! Or use desoldering brasswool to suck the solder out. And give the poor L7812 a bigger heatsink - it will get very hot with a 1 Amp load!
The diode on the fan motor output would be to protect against back-EMF when using a brushed DC motor which could destroy the fan control FET, shouldn't be needed for a brushless motor. As for what died, I would guess an op-amp from receiving whatever the power adapter surged to when it popped the linear regulator..
Andy, the LM358A operational amplifier was failed and heated up to 80 degree. I suggest, that both OpAmps LM358A should be replaced - with the adjacent one also. These OpAmps are cost really cheap. The main problem will be resoldering of these SMD chips - this need some experience. OpaAmps are sensitive to overvoltage and can easily fail during it.
Sometimes you need to blow on the hole instead of sucking it... 😅 Solder wick works too. Bad luck... 🥺, but for all it's worth we expected sparks and smoke... You never disappoint. 👍
Another viewer! I have been watching you on and off for the past 6 months. Thought that I should make a comment, saw someone already stated use solder wick I concur; but you solder melt problem is the heat transfer to the ground plane of the board. Just use a temperature-controlled solder station. Use your thermal meter to find the bad component on the board. I have an off-grid home in Arizona, but I am at my home in Wisconsin. I switched to LiPo assembled batteries about 18 months ago from China, 6 200ah 24-volt units the quality was questionable! After watching your videos with the due diligence, you put in your analysis of the charging and other things I decided to build 6 more of my own. After much research it looks like the best option for the BMS is overkill solar their attention to detail and the use of the same APP that you are currently using got my attention. I have all the batteries and BMS awaiting my return to AZ next month. I will let you know my thoughts on everything as soon as I have assembled and tested them. Really like your attention to detail and the good analysis using the equipment. A retired EE from the university of California.
Thank you very much, Charles and welcome to the channel. Thanks for sharing your story. Great to hear that you will give it another go and even get more batteries. What was wrong with the first batch you received? The Overkill Solar BMS is a great option for a good price. The only downside is the relatively low current (only 100A) for a 48V system which is a bit on the low side for me. Not that I will use that much current but I like some overhead with these devices. I probably stick with the QUCC BMS for now, especially after the 2nd battery bank goes live.
All the opamps probably sits on the 19V rail, so they are all dead. If you are unlucky the analog input on the micro processor is also damaged and will not read correctly.
Desoldering braid? Interesting component level troubleshooting but, Please explain why you don't execute warranty and how could we avoid breaking tester in 1st place? Thumbs up and subscribed as always!
I guess when the original voltage regulator broke it let the full input voltage out to some sensitive electronic components like semiconductors or IC which killed them. That happens typically when a voltage regulator fails. But i remember that you checked the output of the broken one and it showed zero. Normally they tend to go short from in- to output, maybe just for a short period after it completely broke by overload then...
That unit had it, that 12v may have sent a full voltage 19v onto the board and damaged most of the parts, mostly the ic's. They might replace it if it was under warranty?
There are desoldering irons available. In the company we have a solder / desoldering station from Weller, very good and very expensive. At home I have a desoldering iron which is part of my hot-air soldering station. But for holes I still prever the desoldering air pump as shown in the video. Just apply the soldering iron on the side of the pcb with the biggest copper plane (mostly the ground) and the air sucking desoldering pump on the other side of the pcb. So you can heat up and melt the solder and when it is melted you can suck it through the hole straight into the pump without any time delay. And of course you should clean the pump regulary and check if it can still make and hold a vacuum when switched. Apply some grease to the inside rubber sealing (after cleaning 😉) if the vacuum doesn't hold for a few seconds.
I always use braided wire call solder wick with flux. I bet that row pins pull out at both ends. The numbers climbing like that means the reading is measuring a capacitor.
Hello my friend! So how did you make out? Have you solved the problem? I am considering ordering this battery tester. What do you think about them? Should I look for a different brand or do you like it doors. Batteryies recycling place close by uses the same units. They have about 20 of them in operation. I believe they should be well-built.?
Yes, there are powered desoldering stations and tools. Or, you can use the tip of a #11 Xacto knife. Or copper wick that absorbs the solder. Or a dremel tool.
I didn't fix mine and just bought a new one. Lack of time it mainly was for me... The tester works without any problems and has hundreds of hours on full power on the clock. Great device.
Disconnect the fan and test again to eliminate it as the cause to the problem... I have read other comments suggesting that the diode is there for back EMF protection from the fan but the that fan won't produce back EMF because like PC cooling fans, it has a built-in driver circuit which doesn't allow current to output even if you spin the fan blades. The diode appears to be a reverse polarity protection on the sense wire's which is what I would do if I were to be designing a circuit where it is possible to reverse the polarity... This is good design practice 👍.
They do have vacuum solder suckers I want one but I'm not sure I can justify one yet they start at like $120 American from what I have found and go up I did buy basically one of those solder suckers like you are using but built into a crappy soldering iron so like the hollow tip is metal and it has gotten me thru what I have needed to do but it's not great
In the video I could not hear the fan kicking in after turning on the device. In my device there is always a short fan self test. Maybe this is also an indication of a malfunctioning IC board. You could also check the voltage on the two possible sockets shortly after turning on.
A easy way to remove the solder is apply heat to one side of the pcb and use the pump on the other. The overheating part maybe can be replaced i don't think is a eprom.
Funny... I had the same problem just yesterday replacing a Mosfet on a speed controller. This worked for me... Heat the board with a heat gun first, then heat with the iron and then blow hard into the hole and blow out the solder. I'm not having you on... try it (on camera) 😄
Use lots a flux and copper wick for desoldering & cleaning, not John Wick though :))). Watch northridge fix for soldering excellence. If previous regulator failed dead short then you need to check at least ICs and elcos downstream. Keep up the great work!
Datasheets. The devil is in the detail and each figure is usually given independent of it's constraining variables. For example, you correctly stated the maximum input voltage of 30V. The minimum is 14.5V though for proper regulation of 12V under load. However the limiting factor will also most certainly be the "Heat dissipation" of 15W. It's a linear regulator. All voltage is linearly dropped as heat. If you have 30V in and 12V out, that's 18V dropped. If you put 1 Amp through it, then 18W of heat will be dissipating. The package cannot handle that, heatsink or not. The maximum heat dissipation is junction to package, even if you put a square meter of heatsink on it, 18W will cook it eventually. However. Note that the package to air W/*C will probably exceed the thermal limit LONG before you get near 15W. A to-220 is probably barely capable if self disappating 2 or 3 W of heat. The ground plane on the tab might give you 5W. An actual small to220 heatsink might get you 10W. A large block of a heatsink might get you the full 15W, but you will need to do all the thermal transfer functions to work that out.... and yes... they really, really do expect you to do all those sums. It may be the case that the designed of the charge did not do all those sums. Another gotcha with regulators that tends to cook them into early death is "in rush" current. If you have a chunk of capacitance on your board behind the regaultor and you have no inrush protection, when you connect the external power the capacitance will appear as a direct short across the regulator. The caps will demand 100s of amps for a few milliseconds. Powering the 12V reg from 15V you *might* get away with it. Connect it to a 30V source and you probably wont. 100A 18V = 1800W through a 20W regulator = night night.
What is the IC number of that component? Probably a opamp. The SMD doesn't look that small have a look at some SMD soldering videos. Or you can send it down and I will have a crack. Looks promising that the main IC is ok probably as that is the irreparable part.
looks to me like a shorted mosfet on that board. That 8 pin soic that was blazing hot is probably the mosfet in question. I have seen 8 pin mosfet packages like that in laptop motherboards. I would have to know the number on that chip to verify it's a mosfet though.
Often the board has multiple layers of copper inside. Sometimes helpful to let the heat soak for 30 seconds or more ... or just hold the iron on the recalcitrant hole while inserting the new part like you finally did.
it also seems like the A/D converter may be bad, or check the voltage dividers between the shunt/sense/ and microcontroller. Im not sure how it was designed... The amps are sensed by checking voltage drop across the shunt(s) so follow those traces to the A/D or voltage divider. You may need to sit and create a schematic of that area. Remove the large transistors/IGBTs and do a diode test on them... one of them may be affected as well
yea i ordered the big one still waiting. they told me shipping by fedex ip but they shipped it china post ems. i have been waiting almost 2 month. China to thailand should have been 8 to 12 days. not really happy about that. they charged me 70 dollars for shipping which is crazy but i payed cause it was supposed to be fast.
Hello, dear Andy ! Did you tried to change the main MOSFET ? My EBC-A20 showing the same symptoms now when showing 21A current out of nothing and can't analyze anything anymore, not discharging at all , only charging with about 6A regardless what I set but calculating like it is 21A :) and no voltage sense in the software (although on display shows it real) That all happened after reverse polarity by accident and I found out the main MOSFET exploded... I see there is no part 3 of this saga so you never tested this tester ? I ordered the new MOSFET for it , I need to wait more than a month to receive it (I live in Vava'u, Tonga). I'll try and let you know (if anyone will notice this comment at all... not much hopes but writing just in case... )
I received 2 packs of MOSFETS (ordered from 2 different sellers, but exact same part number which is exploded in my EBC-A20) and replaced it. No luck anyway. Same 20A current displaying regardless of anything set or connected or no. Then I changed the MOSFET to another one. Anyway did not worked. Nor charging or discharging mode. But then I found out it is actually only discharging with current 18A according to clamp amp meter ! Immediately after connecting a battery the discharging current starting going. I check - there is no short between +/- crocodiles. only 130Kohm. I don't know may be the relay burned too and it's always connected to discharge mode. But why 18A then. So I gave up and going to buy EBC-A10H 30V 10A 150W ; it's max 10A discharge but EBC-A20 can discharge 20A but max 85W, while A10H can 10A/150W :) Charging is max 5A also but with max 150W , instead of 85W.
Keep that soldering iron on the pad and the solder sucker at the same time and hit the button don't move the iron. I work at an Army Depot and worked on the Minuteman missile doing electronics, they work well, but you can buy a nice PACE machine. Maybe do a calibration?
A lot can be figured out or diagnosed with just a meter. At the very least you should check out the parts connected to heat sinks, since many are power FETs that like to fail as a short. I would say it is a better than even chance one is shorted.
I wish I was there to help reverse engineer that thing to determine the connector connections. From here the one with the diode is the fan to cut back EMF surge. The thin traces are probably the sense lines since they use less current than the fan. If the fan is 12v see if one of the 4 PCB pins connects to the IC you just replaced. I bet the + labeled end of the diode is connected to the 12v regulator output. That is definitely the fan connection.
Ground plane is sucking all the heat.... bigger soldering iron :D I guess that diode is freewheelling diode. Fan motor is basically a coil and turning current off from it will cause a voltage spike in it. That diode suppress it. 350mA sounds too much... maybe? Was fan running?
1/1000 comment here, @5:30 you probably just need a better tempereture controlled iron then crank it to get that ground hole clean. Other than that just make sure you press the solder suckerat the same time you remove the iron from the joint, not much thermal capacity in the joints so it cools fast.
i always use «Braid» for to remove solder, it’s more hard to learn but it’s better once learned.. what was happen to your factory/original 12v regulator 7812? ..if it was short, that explain the damage to your board.. those regulators have a over temperature protection and the only way to blow them is by over-voltage (over 35v.. continuously or with a spike!).. the chip under overheat look short and to confirm if he is alone, just remove it (using some Braid with the iron around 350c and a bit of lead on it.. capillarity effect!)..
Heya, that is to bad probeble a sort cap or risitor or even the chip you touched witch was hot. this is a late responce but if you still have it I know a youtuber how maybe able to repair it
If you just take the mantle out of an ordinary coax cable and soak it into a bit of flux, it WILL suck up the soldering stuff (tin/lead/silver). If needed a compresor could work too but then you have to find ALL the blasted conducting stuff around your soldering spot. Very luxury: a sucking soldering iron will be the best. I get less and less enthusiastic about this ZKE tester....
desoldering tips 1) don't pull the iron away, put them BOTH to the joint from opposite directions, iron angled low to minimize interference with desolder tip. 2) try solder braid!!! reason pin 2 was a bitch,.. it's the regulator ground pin so that via is connected to the ground plane, which quickly wicks heat away from the solder. Also, yes, there are vacuum desodler tools. IMHO, they all suck arse.
That tin remover suck's. I had few, now i only use desoldering wire with some flux. Yes there is electronic desoldering vacuum tools but somehow expensive.
![Blox Fruits Dragon Rework Update [Full Stream]](http://i.ytimg.com/vi/EqDAp8udhm0/mqdefault.jpg)
Since that middle pin is the ground, it's also connected to the ground plane, which acts as a heat sink. The solder doesn't stay hot long enough after you remove the iron for that sucker to work. Use solder wick instead and trash that sucker.
Solder wick. Yes, for at least 40 years there's been desoldering devices with hollow tips that suction from the center of the heated tips.
Yup, solder wick with extra flux if needed can be a lifesaver.
I was going to mention wick last episode, but I was excited to see the old solder sucker in action and forgot to mention it.
You stole my comment 😱👍
Solder wick doesn't work well for me most of the time
@@james10739 Works great for me most of the time, just need to use the correct sized wick with the correct sized tip for the job and a soldering iron that can keep up with how much extra heat the wick wicks away.
Hi Andy, loving your vlog. Thanks. It makes sense to have a diode across a motor to prevent reverse voltage from the inductance of the motor affecting the rest of the circuitry. I remember taking this precaution when powering an inductive load through a power transistor back in the day.
The Diode on the FAN is a flyback diode, it conducts the back EMF, when the FAN is switched off, back to the supply rail.
Right, thank you.
Sort of a Transient Voltage Suppression (TVS) diode.
Keep your head up Andy! Life can knock you down but you got the fight to keep going!
Fan = inductive load, diode is to avoid "back EMF."
There wouldn't be any back EMF from that fan. Like PC cooling fans, they have a driver circuit built in, if you have ever played around with fan's, spinning the fan blades doesn't generate a current.. you would have to open the fan and bypass the driver circuit to get it to generate a current..
The diode appears to be a protective solution should you connect your sense wire's incorrectly...
@@PeterMilanovski So it’s odd the PCB may be labelled incorrectly. Maybe trying changing over those connectors? Though the fan would not be taking the current indicated on the display
@@michaelbouckley4455 I'm not sure what is going on with the device, it's possible that the fan connection is plugged into the sense socket and therefore giving the reading on the display? Or something else has failed which would need an oscilloscope and some datasheets to troubleshoot it....
@@PeterMilanovski
I think you would you agree that for the diode to be a protective solution for the sense wires, the sense wires would need to be connected to the header with the diode. Ultimately, Andy did not connect it that way -- he connected the fan to the header with the diode, and the sense wires went to the header without the diode. My comment was in that context, meaning I was assuming he had them connected properly.
I did not say that there WOULD be any back EMF from the fan -- I was only suggesting that it's a reason the designers may have put one on that header. Perhaps it's unnecessary. Or perhaps not every single fan in the world has a "driver circuit" as you assumed. We don't know without more info.
The midle pin is the ground plane, there is lots of cooper dissipating heat from the solder iron. I normally apply the heat in one side of the pcb and suck the solder from the other side. The diode is a zenner connected in reverse it becames a "voltage regulador"
*copper *zener
Good old solder wick for those troublesome spots.
I am surprised no one suggested using a round toothpick. Just heat the joint, push a toothpick in from the other side and then take the iron away. The toothpick will not absorb any solder and you can just pull it out and the hole will remain open. This method is also helpful if you have a lead broken off in the hole. You can use the toothpick to push out the broken lead.
You need to replace the chip that is getting hot. It is surely a single or dual OpAmp chip that drives the gate of the power MOSFET(s) that charges or discharges the load/battery (pin 1 of one of the big 3 pin transistors on the main heat sink). The OpAmp is powered by the +12V supply rail that you fixed and would be used to measure a low voltage DAC output analog set point voltage reference from a microcontroller, compare this reference to the current shunt voltage, then output an analog voltage in the 0 to 12V range to drive the gate of the power MOSFET such that the current shunt observes the same current that the microcontroller is requesting.
Although you measured open circuit on the 7812 voltage regulator output, it evidently first failed by shorting the input to the output, thus delivering the full 30+V power supply voltage to the OpAmp and blowing it up too. The OpAmp then sucked down enough current to melt the silicon or bond wires in the 7812 such that it now appears open circuit.
OpAmps are a common off the shelf component just like the 7812 is, and you may even find the exact same chip if you go around opening power supplies, your Victron charge controllers and inverter, or perhaps the other el-cheapo PWM charge controllers.
Congratulations 👏👏👏 now it's my turn now I repairing my inverter
Hakko makes a nice desolder tool. I enjoy your work.
Option 1: stand the board up, place the sucker on the top side, touch the iron on the bottom side. Option 2: you can place the iron on the edge of the hole, place the nylon tip on the hole hit the button without pulling the iron away. You're to slow with the method you're using. The sucker tip is made to take the heat. Keep inside the sucker clean.
To compound your problem is the lead free solder. It requires a higher temperature to melt and solidifies much faster.
Just keep your soldering iron on the solder pad while sucking with the pump! The pump's tip is made out of heat resistant plastic, so it doesn't melt. When ypu remove your soldering iron, the solder solidifies instantly especially on the middle pin, which is connected to the ground plane and has a huge thermal mass (acts like a heatsink sucking away your iron's heat immediately)
Hello! Nice video and Nice channel. Thank you for sharing your experiences!
I would add few suggestion as you wanted :)
1. That diode is for protection of your driving circuit (fan) .it will not allow to push voltage in case you would start spinning that fan
2. it looks like normal board but there is huge ground plane and that will reduce your temperature
3. there are special tools that are used to suck solder on ebay (they look like plastic cheap pistol) ..they are not the best but work very well and and they have pump inside ..so yeah cheap option ..and worth some money ..if not go with Weller or similar brand
4. Probably some IC is shorted because your regulator blew up and allowed voltage to go above 12V or some spike with same results..and you might expect more IC went wrong same time..
5. I would suggest to change those IC (if you will be able solder/desolder ) but i would suggest proper tools for that..solder, desolder , hot air gun
looks like the Op-Amp went short, causing extraneous readings on the microcontroller then the linear regulator (7812) blew up due to it overheating. You might be able to replace the Op-Amp but it might have broken some SMD resistors which that and the Op-Amp will be hard to replace with only a soldering iron. you need a hot air station! Its probably best to just get a new one, Good luck Andy!
Ok, my 2 cents.
1. Solder wick with liquid flux on the wick.
2. Flow the solder then push a round toothpick through the solder holes. (My favorite.)
3. Flow the solder and bang the board on the workbench.
Laugh all you want but I've been a bench tech for over 30 years and those are my tried and true methods for solder removal.
Also, when removing ICs, cut the legs and remove 1 at a time. Traces tend to get lifted otherwise. At least on cheap electronics they do.
Hi Andy, thanks for videos showing Your experiences.
How can I dispose for a inflate Lifepo4 battery?
Perhaps if you could identify the chip getting hot? As it seems to be voltage and current readings giving grief maybe it is a reference voltage that has gone south. Some a to d converters rely on an external reference voltage. If it went wrong maybe it also caused the regulator to overload as it failed. A good sign is the pat you are getting readings which implies that the a to d part of the processor is working so the problem is in the periferals and not a programmed part.
yea the op-amp amplifies the signal from the current shunt probably, so it going short made that reading go wild and pulled more current than the heatsink on the 7812 regulator could handle (though it should have overheat protection and shut down before that)
@@SuperBrainAK they do have overheat protection however, in reality, they are still quite prone to dying instead. I have replaced many that it could be argued shouldn’t have died.
@@philbrooke-little7082 yea, one day a long time ago I had a nice breadboard setup with 3 different linear regulators with a Teensy ++ that could turn them on and off and do fun things, Then an acquaintance decided it would be funny to mess with the wires and shorted the 5v rail of the always on regulator for the microcontroller. I came back and noticed the power supply was on but no lights were on at the breadboard, so I touched the heatsink to move the board and got burned by it. I finally found the short but the regulator was unable to deliver any current, it was surprisingly still able to regulate to 5v but couldnt even light an LED. RIP 7805
Use SOLDER WICK, also called "copper braid" or "desoldering braid". Put the wick on the cold dry solder and heat the solder through the wick.
Alternatively, I sometimes use a wooden toothpick to "unclog" a PCB hole.
Second that! Much better on many pcb board. So easy to get frustrated and add too much heat to the board and destroy the lanes/the board.
The fan diode is there to protect the fan power circuit from back feed current from the fan eg when you hit it with the air gun
When you spin them up they can generate a lot of current
That would be true if the fan didn't have a driver circuit but in the case of this fan and all PC cooling fans, they don't produce current if you spin the fan blades, you would have to modify the fan, bypass the driver circuit to get it to generate a current...
@@PeterMilanovski Yes they do and it's easily provable
@@awesomusmaximus3766 I have never been able to generate a current from one of those types of fan's by spinning the fan blades... They need to have the driver circuit bypassed in order to do so...
@@PeterMilanovski I've seen over 15V
did you ever get to do a part 3? i have an ebc b20h which also shows overvoltage.. issue also came after a power cut.. which could have caused either charger input or mains input to spike.. i was hoping to salvage it with a few new components..
Tipp: use a tooth pick to get a hole in the solder pads when the sucker is not working! Or use desoldering brasswool to suck the solder out. And give the poor L7812 a bigger heatsink - it will get very hot with a 1 Amp load!
The Electrons are confused buddy. 😳. You have the right Idea, have a cold one.
The diode on the fan motor output would be to protect against back-EMF when using a brushed DC motor which could destroy the fan control FET, shouldn't be needed for a brushless motor. As for what died, I would guess an op-amp from receiving whatever the power adapter surged to when it popped the linear regulator..
Thank you.
Andy, the LM358A operational amplifier was failed and heated up to 80 degree. I suggest, that both OpAmps LM358A should be replaced - with the adjacent one also. These OpAmps are cost really cheap. The main problem will be resoldering of these SMD chips - this need some experience. OpaAmps are sensitive to overvoltage and can easily fail during it.
Sometimes you need to blow on the hole instead of sucking it... 😅
Solder wick works too.
Bad luck... 🥺, but for all it's worth we expected sparks and smoke... You never disappoint. 👍
Another viewer! I have been watching you on and off for the past 6 months. Thought that I should make a comment, saw someone already stated use solder wick I concur; but you solder melt problem is the heat transfer to the ground plane of the board. Just use a temperature-controlled solder station. Use your thermal meter to find the bad component on the board. I have an off-grid home in Arizona, but I am at my home in Wisconsin. I switched to LiPo assembled batteries about 18 months ago from China, 6 200ah 24-volt units the quality was questionable! After watching your videos with the due diligence, you put in your analysis of the charging and other things I decided to build 6 more of my own. After much research it looks like the best option for the BMS is overkill solar their attention to detail and the use of the same APP that you are currently using got my attention. I have all the batteries and BMS awaiting my return to AZ next month. I will let you know my thoughts on everything as soon as I have assembled and tested them. Really like your attention to detail and the good analysis using the equipment. A retired EE from the university of California.
Thank you very much, Charles and welcome to the channel. Thanks for sharing your story. Great to hear that you will give it another go and even get more batteries. What was wrong with the first batch you received? The Overkill Solar BMS is a great option for a good price. The only downside is the relatively low current (only 100A) for a 48V system which is a bit on the low side for me. Not that I will use that much current but I like some overhead with these devices. I probably stick with the QUCC BMS for now, especially after the 2nd battery bank goes live.
All the opamps probably sits on the 19V rail, so they are all dead. If you are unlucky the analog input on the micro processor is also damaged and will not read correctly.
Give me some good news!
Desoldering braid? Interesting component level troubleshooting but, Please explain why you don't execute warranty and how could we avoid breaking tester in 1st place? Thumbs up and subscribed as always!
The chip getting hot looked like it has a small spot where it blew up? Not sure, but looked like it in the video.
I guess when the original voltage regulator broke it let the full input voltage out to some sensitive electronic components like semiconductors or IC which killed them. That happens typically when a voltage regulator fails. But i remember that you checked the output of the broken one and it showed zero. Normally they tend to go short from in- to output, maybe just for a short period after it completely broke by overload then...
The diode is for back emf spike of the fan when you turn it off.
That unit had it, that 12v may have sent a full voltage 19v onto the board and damaged most of the parts, mostly the ic's. They might replace it if it was under warranty?
Why don't you use solder wick? Or make it yourself from the braided screen of a coax cable?
There are desoldering irons available. In the company we have a solder / desoldering station from Weller, very good and very expensive. At home I have a desoldering iron which is part of my hot-air soldering station. But for holes I still prever the desoldering air pump as shown in the video. Just apply the soldering iron on the side of the pcb with the biggest copper plane (mostly the ground) and the air sucking desoldering pump on the other side of the pcb. So you can heat up and melt the solder and when it is melted you can suck it through the hole straight into the pump without any time delay. And of course you should clean the pump regulary and check if it can still make and hold a vacuum when switched. Apply some grease to the inside rubber sealing (after cleaning 😉) if the vacuum doesn't hold for a few seconds.
Hi Bro, i have charging problem in my zke tech, dou you have any solution? should i change the ic Step down?
Can't find part 3, did you do one or did they just replace the unit for you?
Neither, I just bought a new one and never repaired the old unit. Still got it though...
I always use braided wire call solder wick with flux. I bet that row pins pull out at both ends. The numbers climbing like that means the reading is measuring a capacitor.
You need to have the solder sucker in place as the heat is being applied. If you occasionally melt its tip a little bit, you're doing it right.
Hello my friend! So how did you make out? Have you solved the problem? I am considering ordering this battery tester. What do you think about them? Should I look for a different brand or do you like it doors. Batteryies recycling place close by uses the same units. They have about 20 of them in operation. I believe they should be well-built.?
Yes, there are powered desoldering stations and tools. Or, you can use the tip of a #11 Xacto knife. Or copper wick that absorbs the solder. Or a dremel tool.
when you clearing hole i tend to heat from the back and suck from the other side keeping solder more molten but even then it can be a pia
Did this ever get fixed? I have one on the way ... did I make a mistake?
I didn't fix mine and just bought a new one. Lack of time it mainly was for me...
The tester works without any problems and has hundreds of hours on full power on the clock. Great device.
That's good to know ... thanks Andy 👍@@OffGridGarageAustralia
Disconnect the fan and test again to eliminate it as the cause to the problem...
I have read other comments suggesting that the diode is there for back EMF protection from the fan but the that fan won't produce back EMF because like PC cooling fans, it has a built-in driver circuit which doesn't allow current to output even if you spin the fan blades.
The diode appears to be a reverse polarity protection on the sense wire's which is what I would do if I were to be designing a circuit where it is possible to reverse the polarity... This is good design practice 👍.
They do have vacuum solder suckers I want one but I'm not sure I can justify one yet they start at like $120 American from what I have found and go up I did buy basically one of those solder suckers like you are using but built into a crappy soldering iron so like the hollow tip is metal and it has gotten me thru what I have needed to do but it's not great
Add a little bit bigger Heatsink for ic 7812 or connect a 12v mini cooling fan ;) 👍
I wonder if your EBC-A40L has similar wiring? 🤔
In the video I could not hear the fan kicking in after turning on the device. In my device there is always a short fan self test. Maybe this is also an indication of a malfunctioning IC board. You could also check the voltage on the two possible sockets shortly after turning on.
A easy way to remove the solder is apply heat to one side of the pcb and use the pump on the other.
The overheating part maybe can be replaced i don't think is a eprom.
If you heat the solder then sharply bash the PCB on the desk the shock often throws the solder out of the hole.
yup, I do that too
Thanks Andy
Andy, did you ever get this one fixed?
Funny... I had the same problem just yesterday replacing a Mosfet on a speed controller. This worked for me... Heat the board with a heat gun first, then heat with the iron and then blow hard into the hole and blow out the solder. I'm not having you on... try it (on camera) 😄
Diode on the fan motor (essentially a coil) is a flyback snubber just like you have on a relay coil.
Yes Andy! That way! It looks right! No wait! Swap it! Yep. That should be good!
Use lots a flux and copper wick for desoldering & cleaning, not John Wick though :))). Watch northridge fix for soldering excellence. If previous regulator failed dead short then you need to check at least ICs and elcos downstream. Keep up the great work!
Datasheets. The devil is in the detail and each figure is usually given independent of it's constraining variables.
For example, you correctly stated the maximum input voltage of 30V. The minimum is 14.5V though for proper regulation of 12V under load.
However the limiting factor will also most certainly be the "Heat dissipation" of 15W.
It's a linear regulator. All voltage is linearly dropped as heat. If you have 30V in and 12V out, that's 18V dropped. If you put 1 Amp through it, then 18W of heat will be dissipating. The package cannot handle that, heatsink or not. The maximum heat dissipation is junction to package, even if you put a square meter of heatsink on it, 18W will cook it eventually.
However. Note that the package to air W/*C will probably exceed the thermal limit LONG before you get near 15W. A to-220 is probably barely capable if self disappating 2 or 3 W of heat. The ground plane on the tab might give you 5W. An actual small to220 heatsink might get you 10W. A large block of a heatsink might get you the full 15W, but you will need to do all the thermal transfer functions to work that out.... and yes... they really, really do expect you to do all those sums.
It may be the case that the designed of the charge did not do all those sums.
Another gotcha with regulators that tends to cook them into early death is "in rush" current. If you have a chunk of capacitance on your board behind the regaultor and you have no inrush protection, when you connect the external power the capacitance will appear as a direct short across the regulator. The caps will demand 100s of amps for a few milliseconds. Powering the 12V reg from 15V you *might* get away with it. Connect it to a 30V source and you probably wont. 100A 18V = 1800W through a 20W regulator = night night.
What is the IC number of that component? Probably a opamp. The SMD doesn't look that small have a look at some SMD soldering videos. Or you can send it down and I will have a crack. Looks promising that the main IC is ok probably as that is the irreparable part.
looks to me like a shorted mosfet on that board. That 8 pin soic that was blazing hot is probably the mosfet in question. I have seen 8 pin mosfet packages like that in laptop motherboards. I would have to know the number on that chip to verify it's a mosfet though.
Ground is more masive that's the reason is hard to clear the hole
Often the board has multiple layers of copper inside. Sometimes helpful to let the heat soak for 30 seconds or more ... or just hold the iron on the recalcitrant hole while inserting the new part like you finally did.
the diode may for blocking noise from the brushless motor....
it also seems like the A/D converter may be bad, or check the voltage dividers between the shunt/sense/ and microcontroller. Im not sure how it was designed...
The amps are sensed by checking voltage drop across the shunt(s) so follow those traces to the A/D or voltage divider. You may need to sit and create a schematic of that area.
Remove the large transistors/IGBTs and do a diode test on them... one of them may be affected as well
I just bought a used EBC-A10H and getting phantom current of about 8A always showing with nothing connected. :-(
One of the FETs was blown. Replaced it and have been merrily using it ever since
yea i ordered the big one still waiting. they told me shipping by fedex ip but they shipped it china post ems. i have been waiting almost 2 month. China to thailand should have been 8 to 12 days. not really happy about that. they charged me 70 dollars for shipping which is crazy but i payed cause it was supposed to be fast.
Hello, dear Andy ! Did you tried to change the main MOSFET ? My EBC-A20 showing the same symptoms now when showing 21A current out of nothing and can't analyze anything anymore, not discharging at all , only charging with about 6A regardless what I set but calculating like it is 21A :) and no voltage sense in the software (although on display shows it real) That all happened after reverse polarity by accident and I found out the main MOSFET exploded... I see there is no part 3 of this saga so you never tested this tester ? I ordered the new MOSFET for it , I need to wait more than a month to receive it (I live in Vava'u, Tonga). I'll try and let you know (if anyone will notice this comment at all... not much hopes but writing just in case... )
I received 2 packs of MOSFETS (ordered from 2 different sellers, but exact same part number which is exploded in my EBC-A20) and replaced it. No luck anyway. Same 20A current displaying regardless of anything set or connected or no. Then I changed the MOSFET to another one. Anyway did not worked. Nor charging or discharging mode. But then I found out it is actually only discharging with current 18A according to clamp amp meter ! Immediately after connecting a battery the discharging current starting going. I check - there is no short between +/- crocodiles. only 130Kohm. I don't know may be the relay burned too and it's always connected to discharge mode. But why 18A then. So I gave up and going to buy EBC-A10H 30V 10A 150W ; it's max 10A discharge but EBC-A20 can discharge 20A but max 85W, while A10H can 10A/150W :) Charging is max 5A also but with max 150W , instead of 85W.
Keep that soldering iron on the pad and the solder sucker at the same time and hit the button don't move the iron. I work at an Army Depot and worked on the Minuteman missile doing electronics, they work well, but you can buy a nice PACE machine. Maybe do a calibration?
Oh the pain of desoldering ground pins on pcbs...
A lot can be figured out or diagnosed with just a meter. At the very least you should check out the parts connected to heat sinks, since many are power FETs that like to fail as a short. I would say it is a better than even chance one is shorted.
I wish I was there to help reverse engineer that thing to determine the connector connections. From here the one with the diode is the fan to cut back EMF surge. The thin traces are probably the sense lines since they use less current than the fan. If the fan is 12v see if one of the 4 PCB pins connects to the IC you just replaced.
I bet the + labeled end of the diode is connected to the 12v regulator output. That is definitely the fan connection.
Thanks for your nice videos. Oettinger, bist Du deutscher? Ich hatte schon öfter gedacht das Du einen deutschen Akzent hast
Ich hab doch keinen Akzent! Was bist du den fuer einer 😂
@@OffGridGarageAustralia Cool, mit Akzent meinte ich das ich dein englisch echt gut verstehe! Gibt es Oettinger Bier in Australien?
@@marcusschleicher Ja, das gibt's hier. So wie fast alle anderen deutschen Biersorten. Alles eine Frage des Preises ;)
I would think the broken ic could be a ADC, analog digital converter, the part, that makes the voltages digital
It's an OpAmp LM 358A and it's toast!
Ground plane is sucking all the heat.... bigger soldering iron :D
I guess that diode is freewheelling diode. Fan motor is basically a coil and turning current off from it will cause a voltage spike in it. That diode suppress it.
350mA sounds too much... maybe? Was fan running?
I remove as much solder as I can with solder wick, then use a wooden toothpick to push thru the molten solder in the hole and let it cool.
1/1000 comment here, @5:30 you probably just need a better tempereture controlled iron then crank it to get that ground hole clean. Other than that just make sure you press the solder suckerat the same time you remove the iron from the joint, not much thermal capacity in the joints so it cools fast.
Check out the Weller WDD81X Digital Desoldering Station.
i always use «Braid» for to remove solder, it’s more hard to learn but it’s better once learned.. what was happen to your factory/original 12v regulator 7812? ..if it was short, that explain the damage to your board.. those regulators have a over temperature protection and the only way to blow them is by over-voltage (over 35v.. continuously or with a spike!).. the chip under overheat look short and to confirm if he is alone, just remove it (using some Braid with the iron around 350c and a bit of lead on it.. capillarity effect!)..
as others have stated use the wick. but if that doesn't work i have used compressed air (blow nozzle) to clear the problem.
Fan diode to absorb spike of reverse emf when the fan is de-energized.
Maybe is better to have another tester ready to use and repair this in second time? Battery testers are never enough!
Heya, that is to bad probeble a sort cap or risitor or even the chip you touched witch was hot. this is a late responce but if you still have it I know a youtuber how maybe able to repair it
solder wick, copper braid pulls solder right up when heated.
cheers andy !
It could be the mosfet that failed short. Measure ohms across the mosfet Drain and Source. It should be open circuit.
you have a short that's why the mosfets are heating up. you have a bad mosfet or a bad capacitor or a diode :)
Hakko FR-301 Desoldering Gun
If you just take the mantle out of an ordinary coax cable and soak it into a bit of flux, it WILL suck up the soldering stuff (tin/lead/silver).
If needed a compresor could work too but then you have to find ALL the blasted conducting stuff around your soldering spot.
Very luxury: a sucking soldering iron will be the best.
I get less and less enthusiastic about this ZKE tester....
Any one knows how to fix this ?
Heat with iron under board and solder sucker on top.
I think i can see a hole in the IC that gets Hot. I have seen that many times when Electronics let out there Blue Magic smoke.
I need help figuring out how to pick a daly BMS, the more I read the more confused I get
desoldering tips
1) don't pull the iron away, put them BOTH to the joint from opposite directions, iron angled low to minimize interference with desolder tip.
2) try solder braid!!!
reason pin 2 was a bitch,.. it's the regulator ground pin so that via is connected to the ground plane, which quickly wicks heat away from the solder.
Also, yes, there are vacuum desodler tools. IMHO, they all suck arse.
Always take pics before disassembly, by the way red is + , black -
Hi , versuch doch mal die gute alte Entlötlitze 😉😉
DeSolder braid and less time with the iron on the board
OpAmp goo brrr 😂
Change the OpAmp on the controller board
I did and it's still bonkers... The controller shows wrong figures on the display, shows amps even there is nothing connected.
That tin remover suck's. I had few, now i only use desoldering wire with some flux. Yes there is electronic desoldering vacuum tools but somehow expensive.
3:20 use solder wick/brade
Maybe it is free wiling diode