Great explanations. The pwm controller forces panels down to the battery voltage where the mppt lets the voltage hit the maximum power point. A damaged cell will cause no power on that third of a panel until the backward voltage is high enough. So a lower voltage will allow that damaged cell string to produce.
Yeah, I'm thinking of testing the panels individually with an MPPT controller and see if they reach the max power point or close... Just need a 12V battery.
Andy, one thing I do not understand, how can a diode which is bypassing 24 cells for reverse voltage and current become conducting. Aren't they just for reverse polarity protection of the panels (so to minimize the risk of heating and destroying your cells when you feed a plus of a charger on the minus of the cells) how can the diode ever become conducting when he is against the polarity which you draw from the cells. It looks like the same diode you put on a motor circuit to protect your setup from negative voltage to your charger/mosfets. Or am I completely wrong? I think on a dc motor it's called a flyback diode to my knowledge. If you have cracks in the cells between the minus and plus you should not see any voltage or should not get any current, a crack never passes any electricity. So if you see a high voltage and a short current of 5A then there should be connectivity, cells cannot be interrupted otherwise no current will flow. If a cell is 0.5 V having 72 in series will give you 35V, having a short of 5A will mean a good connection, those diodes will not conduct only in reverse polarity or if they are somehow voltage driven when in full load which would be strange. Just my thoughts, good luck and have a beer! Can you try your electronic load with wich you test your cells as it's not dependend on the voltage of the cells which might be to less when supplying power/under load on your charge controllers. Or have 4 12V lamps in series would show if the cells are able to deliver power.
The bypass diode is for when you put panels in series. Imagine you have 10 panels in series (like on my house). That's 30 strings of 24 cells in series, right. If one of the strings is shaded, the other 29 strings will deliver full power and just push through that shaded one. The current will always look for the path with lower resistance and finds the diode in parallel to the shaded string. And because the voltage from the other 29 strings is much higher, it keeps pushing amps but through the diode. The diodes are not so relevant in smaller installations I guess.
You have Victron 150/35 charge controllers. These have a maximum of 150v input, however the spec sheet states "MPPT range 60V to 130V". Your problem strings are at 145v so exceed this....so you are possibly outside the MPPT range and the controller may be pulling the panel voltage down.
Yes, someone commented this on a previous video. It’s not good to be at the max voltage that any equipment can use. When the panels are shaded he is seeing vmp. When panels get full sun he might be going over 130v. Causing the mppt to not operate properly
@@RJ-cc1fz1. battery soc. 2. What the entire strings produce BEFORE the charge controller 3. The settings of the charge controller. I get frustrated by a non methodical approach to fault diagnosis, especially when panels are carted up and down ladders adding to serious injury risk. It seems obvious to me that a simple test of what is being produced before the charge controller would yield good info and from that show the direction further testing needs to happen.
No, No, No. That's not the point. The panels are within the specs of the SCC and perfectly within the MPPT range. You never run the panels with 145V at max power. That's just not true. It's the Voc what you're referring to not the Vmp. I explained this in the video.
You can visually inspect for micro cracks with the help of a infrared camera and by giving the panel energy from a lab power source without the diodes on the back. Then look at the panel through the infrared camera and you will see the panel light up( yes it makes infrared light). Where the panel does not light up you have the problem. Keep up the good work! Awesome videos as usual!
@@OffGridGarageAustralia You can make one out of any digital camera. Just open it up and remove the IR filter from the cmos image sensor after you remove the lens.
Now calculate back the watts per m2...you will be surprised what the outcome is....old pv vs new pv, it will really surprise you, if the panels are good. Keep up with your series, you do a very good job, keep up the good jod!
Dear my friend , yes the crack do this problems but the PID still do too much problems for the old pv panels , these old pv panels without PID protection. The cracks and the PID , both doing same decay in power.
My 2 cents worth. These panels have a problem, they aren’t going to be the ones that will be the permanently on the west roof. You have plenty on other panels. Decided which of these you want on the west roof. Set them up on the ground in the configuration you want and test if they work. If so install them. Then you’ll have plenty of time to come back to them at a later date without compromising the charging of your system. And yes I’m as keen as everyone else to find the problem, but it just seems he’s missing out on full charges while doing all this testing. Anyway keep up the great content, and if you need a hand doing any of this get in touch, I’m just up the road at Buaraba ( between Gatton and Esk). I’m happy to help out.
Diodes will almost always fail as short circuit so if a diode was damaged you will have easily noticed in the open circuit voltage as that will have been reduced by 33% but that is not the case base on your earlier tests. The only explanation to reduced current is that some of the cells are cracked so they act as a smaller area cell thus lower current. Is just strange that all panels will have more than one cracked cell. Curious how this new panels you got perform since they are from a different installation but same type of panels. You only need a single cracked or shaded cell in a group of 24 series cells for that string to be bypassed by the diode assuming that the other unaffected 24 cell strings produce enough voltage for the battery or MPPT minimum require. When you do a short circuit test and that results in a proper current say 5 or 6A in full sun then all you can be sure about is that one of the 3 groups of 24 series cells is OK you have no idea if the other two are OK or not.
That is a good point. Thanks again for your comment on this. I probably connect each panel individually to an MPPT SCC and cool down the panels in full sunlight to see how they perform. Even if I find out they have one string not working correctly, there is nothing I can do about it. It affects the whole panel anyways and is not repairable.
Andy, Great explanation as always. Why not use an std extension cord and fashion some adapter ends to get more than 7 meters. That would allow easy setup and you to troubleshoot the problem on the ground. Keeps you off the ladder unless you are lifting a solution not a problem :)
I made panel by mono crystal line cells and some panel with crack cells too but they still gave good power unexpected ly they still working fine reverse polarity diodes are silicone type still working good 1N5408 three in parallel also used in your case Schottky is used?type.. please find if they are ok...thanks for nice information you are giving
For every panel to have the same random failure? Seems unlikely. I am at a loss as to what is going on, but if it was just down to cracks then you should expect to see variations in performance across the array.
More information is needed for diagnosis. Much more. If a car stops working, do you take the entire engine apart before checking the gas tank? Or see you have fuel, so condemn the gearbox? Fault diagnosis is a skilled job, and can only be done when you understand the whole system and how the different parts of that system interact with each other.
15 A für die Dioden ist gut dimensioniert. Aber ein Blitzschlag in der Nähe kann die Kennlinie versauen so das die Diode nicht ganz defekt ist. Deswegen schlage ich vor die Diode versuchtechnisch abzuklemmen wenn du einen Fehler bei den 3 Strängen innerhalb des Moduls findest. 15 A for the diodes is well dimensioned. But a lightning strike in the vicinity can ruin the characteristic curve so that the diode is not completely defective. Therefore I suggest trying to disconnect the diode technically if you find an error in the 3 strings within the module
I always cracks me up at how low current the bypass diodes are rated for. The panel will have 9-30A running through it. Yet the diodes are something like 3-5A ??? How the hell is that supposed to work. I have no answers, but some of the others here may know something about this?
We use an insulation tester to push a low current, high voltage injection throigh electrical equipment to test if it has a short or is down to earth. Perhaps you could find a way to utilise an insulation tester (aka megger meter) to push the higher voltage through the panels to see if the electrical pressure shows up with a short 🤷♂️ I’m gonna try this on a good and bad panel when I get home and see if it tells me anything.
If you have access to a thermography camera you can Diagnose damage panels directly in the roof. Just search for Practical guide Solar Panel Thermography.
Failure mechanisms of bypass diodes are put in two categories of catastrophic and wear out, but what is localized breakdown at the die level (which supposedly can eventually cause a complete failure of the panel or cell or device(Satpathy & Pamuru (2020), page 177, Solar PV Power: Design, Manufacturing and Applications from Sand to Systems))? Next, the electron model of current flow from the cathode, then the conventional model of current flow with the arrow as flow from the anode (which is a fictional convenience), mixed into aggravating flip-flopping, near-failing, bi-directional, physics of worn diodes add to the paradoxes.
Hi Andy, this is not a comment on the panel problem , but another different issue, i am going from agm batteries to a lithium pack with BMS, i have a major concern to what happens if the BMS switches off with a fault either over or undercharging? I have a midnight classic controller and victron phoenix inverter, if you suddenly disconect the neg line from both this seems like a real bad idea, especially regarding the charge controller as it would be left with maybee two kw of solar coming into the controller with nowhere to go, and how would the bms handle all tthis current if it comes back on line? I am begining to think all the equipment should be commiincating from the bms to shut down the inverter and charge controller safely, and not all the eqipment has a comms facility. Ps love all your fault finding, watch your vids every day.
@@lenfryers2839 Hay lens. Your comment has been published as an answer to mine so, I think, it can be not read by people that can help you. I don't have batteries, I'm just in-grid (for now). Publish it in the main thread and good look. Have a nice day
saw someone check if the cells were broken by switching on power on them (do not remember if it was correct or reverse polarity) and got defects to start shining but not much so it needs to be really dark but then it should be possible to see problems
Andy if you are getting a huge voltage drop when the charge controller is trying to draw current, sounds like you have a faulty panel somewhere! It's kinda like a week lead acid battery! Under heavy current draw, the voltage will sag! This is how they test lead acid batteries! The testing device is nothing more than a heating element with a voltage gauge! (Well the old battery testers are anyway which are still being sold in Bunnings! And all over eBay for around $25 - $40). I'm not sure if a short circuit test with the current clamp will show up the fault although I don't see any reason why it shouldn't, otherwise you might need a high powered load Resistor with MC4 connectors and an analogue voltage gauge because sometimes the deflection of the needle is more important than the voltage which a modern multimeter with an LCD screen is not very good at doing!
@@OffGridGarageAustralia Your tests seem to have confirmed that one or possibly two strings within each panel have a fault that effectively introduces a high impedance in the string. An open circuit voltage will not detect the high resistance cos your meter impedance is many times higher. Similarly, the bypass diode on the panel will turn on when you do the short-circuit test and bypass the faulty string, just as it should do. Expose the diodes and the string terminals and measure the voltage on each string when connected to the MPPT charge controller. This will show up the faulty string. You can then solder a wire across these terminals to produce a fully working panel of lower voltage, same current and hence lower power. If you look carefully at the MPPT controller readings at the moment the controller is hunting for a maximum power but is unable to find it. This is because the controller is designed to work with low impedance panels but yours are now high impedance. If you short out ths faulty string you will reduce the impedance back to normal and the controller will work again. You will find a different MPPT controller other than Victron will give different results; some will perform better with high impedance panels and some worse. Good luck!
Andy, I would install used panels on my roof but... I'm waiting to final diagnosis of yours problems! And simple solution: how to test used panels before installing it.
I have reviewed all the movies. I do not believe so this is a problem with cabling. I also do not believe so that is the problem with the broken panels - is possible to have one broken, or maybe more than one, two - but all nine? Not possible. Maybe three is one broken panel, each in one string - possible… you could change that nine panels between them - simply scramble it - one string should be better than other. Most interesting thing is in 3rd part in series. 2 panels on the ground working well with PWM charge controller but not working with MPPT Victron controller - WHY? It should not happened - what is the difference? Algorithm and…. battery voltage. Could you use Victron to charge 12V battery - what current than will be? Maybe, just maybe, you removed from the roof two good panels, which works correctly with PWM charger - but voltage on two panels was too low to power 52V battery! Look to the movie - from the string you received only 65V at that moment. Is this enough to charge 52V? IMHO checking panels for open Voltage and Circuit Current give information that all should be OK. (however checking diode is good idea) maybe that power is produced by panels but lost - for example to the ground - some static resistance or passing current under highest voltage - that’s why yours home installation signalling fault (grounding) but when you measure that with 1.5v mustimeter that is not showing.
I'd like to see the difference between tree shading and cloud cover shade. I think that with tree shade where you have spots of sunlight on the panel, the shaded portion is actually similar to night-time, and therefore the overall resistance of the cells altogether is very high and will generate less than a uniform low light condition. I'm not sure why the overall output of the strings hasn't been measured before the charge controller, because with that one simple measurement, alot of info could be got. Also, I'm not sure why soc of the battery is not shown. Narrowing down the cause of a fault is a methodical process that involves all parameters because all parameters make up the system. Never make assumptions. There is so much information lacking, its impossible to make an informed diagnosis. The diodes were good to check, but after seeing reasonable production when each panel was tested under short circuit condition, they seem to be fine. Going back to the connected strings, and measuring before the cc, if you have expected output and the cc is ruled out due to being swapped for a known working one, the next check is the bms.
Somehow neither the voltage sense with an open circuit nor the current test with a short circuit make a lot of sense, because in both tests the pannel doesnt deliver power. I would measure the volatage of each module while they are connected in series and delivering power. It's quite unliekely that all the modules in one string have as same as many and as same as bad cracks. So if cracks are the reason for the low power, the modules should have different voltages.
Yes, that is correct. The best way would be to test the panels individually with an mppt charge controller and see if it can pull the full power in full sunlight. I can also cool down the panel to 25°C for such a test.
@@OffGridGarageAustralia That would be a lot of work. What I meant was measuring the volatge of each module, while they are still connected in series on the roof. Therte are fine measuring points (needels) for volt meters that can pinch throough the insulation of your solar cables. This way you could compare the voltage of all module sin a string wthout a lot of effort. If some of the modules show a lower voltage than the others, you can take only those ones down for further tests. For testing I would also prefer a PWM controller if there is one available. Those dont do any convertion of voltage and current, so show the current the modules really deliver.
You should recognize the white goo as common thermal contact paste, similar to zinc oxide sunblock, put in to transfer heat out of the diodes. Replace it.
If it's got micro cracks in the cells it should have snail trails starting to develop where the micro cracks are you will be able to see them in full sun or should be able to
Get enough power resistors to make 6 ohms 190 watts and check each panel for voltage and current. I don't know if 3x 60 watt light bulbs would work or not. Maybe 3 automobile headlights in series.
Taken in isolation, short circuit current isn't hugely helpful in diagnosis. Same with voc. But together they tell good info. Voc tells you all cells are connected in the string. If that measurement is in spec, then the current measurement (short circuit) will give good info. There are situations where no load measurement is in spec but as soon as a load is presented either current falls or voltage collapses. This indicates a poor connection within the panel. A bad connection will often have voltage but as soon as current flows, will go high resistance. The side effect is often heat at the point of the bad connection.
Andy good morning from Grass Valley California USA If you had 32 of your cells would you make 2 48volt packs then join them or 1 big 48volt pack? and would there be any pros or cons to one way or the other thanks Jimmy
That is a great question I'm asking myself for some time now. I have 16 more cells coming in and I'm tossing around some ideas in my head if I should build one large battery or two separate batteries. I will discuss this in one of the next videos...
@@OffGridGarageAustralia ok great can’t wait for the video I’m waiting for the rest of my parts then I’ll be building it. I’m building mine for when the Power company PG&E shuts the power off during hi fire risk times when it’s windy and dry Love your videos keep them coming
Hi Andy, you do not have or can rent an infrared camera? If yes so look through it at a panel under full sunlight and defective cells should be seen as hot surfaces ...
I haven't got one yet, no... Well in full sunlight, all panels are hot 50-60°. I doubt the faulty ones will be even hotter. As others suggested it might be better to supply DC to the panels and make them glow at night.
@@OffGridGarageAustralia Yes, it can work too. A simply defective cell should behave like an electrical load and should therefore heat up more than those that are OK.
I wouldn't move them down.. make it simple..make simple two T connectors and just check voltage drop on every panel. It should be about same ..if you see much smaller value than rest of the panels then you will know there might be something wrong with that one panel. It is not likely that all panels are not good...
You said lots of beer donations, easy with the beer, don’t drink and troubleshoot, you might get electrocuted or fall off the roof,remember you told us SAFETY FIRSTdid you recall!JUST DRINK ROOT BEER!
One thing you did is test all strings seperatly, so if one panel is faulty in each string the voltage would never be 110V unless one of the cells only delivers 2A, it's series so the weakest link is in force here. Will your diodes conduct when the other panels have a higher voltage when placed parallel, I don't think so. What I can imagine is in parallel you might feed the panels with current from the good string, so warming the cells instead of getting power from them, this you should be able to measure feom within your garage as the current direction should be seen with your meter. So one string feeds maybe 2 faulty strings. Sorry this is such a long story, the problem doesn't seem simple.
I probably have to test the panels individually with an MPPT charge controller and also cool them down to 25°. Just need a 12V battery for that... The diodes will not conduct as long as the produced voltage of the single string is high enough to deliver power to the overall panel.
@@OffGridGarageAustralia Overload is correct because voltmeter in diode tester has over 2V input voltage (with open contacts) on 2V range. In voltmeter mode it shows also OL - what "Open Loop" then... It's from old multimeters ;o)
Du weißt doch jetzt wo die Dioden und wo die kühlplatten sind. Warum bohrst du nicht 3 Löcher in das Gehäuse bis du auf das Metallplättchen kommst :) Dann kannst du darüber auch einfach messen! das würde dir eine Menge Arbeit und zerstörung sparen! :) Am Ende mit einem kleber silikon oder ähnliches wieder abdichten. Weiter so und ich hoffe deine Zellen laufen bald wieder in der Australischen Sonne! :)
@@OffGridGarageAustralia Welcome Andy! Good morning! I see a victron update is ready. i wonder if they're releasing the version you've been testing 🤔😉 Cheers n sunny day to ya bro 🌞☀🌤
I would quess the panels that had connectors cut off are the bad panels. Not lazy, instead preventing use of known bad panels.??? Most people cannot crimp an mc4 connector.
That's a conspiracy theory that damaged panels are routinely sold. The supplier would go out of business quick smart if that happens. Its a lazy theory.
@@yodab.at1746 I didn't think he bought from a dealer. I know they are used. I assumed he got them by picking them up from houses after retrofit. I don't do conspiracies. Just thinking out loud. It seems like a good fire safety practice to prevent bad panels being used.
@@yodab.at1746 I would expect a reputable dealer would never sale a panel without an mc4 connector nor a broken panel. I would not expect an average home owner of grid tied system to be able to test if the panel is good or bad. Buying used is often a roll of the dice
@@edc1569 Checked Amazon for "solar barrier diodes". They're all 12SQ040 series. Got a spec sheet. They drop .2 volts at zero amps and about .4 ish @ 5A. Thanx :)
Ground faults can give the exact problem you are enduring. Fitting second hand panels without meggering and recording all the results is a waste of time. Ask if you want an explanation. When commissioning a new system it is necessary to perform the above tests. Diodes that appear ok with diode check at 1.8 volts to 2.4 volts but will behave as a short at system voltage. I teach tradesmen this type of fault finding. I make simulators that check ok on diode check but are faulty. Diodes only fail open from mechanical damage. Electrical failure is a short circuit. Testing in the driveway with no frame earth will just confuse you. The panel you removed from your house because of ground fault is strongly implicated in your problem.
The discussion about short-circuit current is completely bogus. Short circuit current (Isc) and open circuit voltage (Voc) are the critical criteria. The only way a single cell working and the rest dead is if all the rest of the cells are shorted, and that would be obvious by the Voc.
Each cell should measure the same temperature, 45 dec C is the nominal, but could be higher. A colder cell is a nonconductive cell. An extremely hot cell indicates I ??? don't know. Since you are 27 deg south latitude, I am going to estimate a 10-20% loss of solar intensity for air mass thickness. The difference between STC temperatures and the nominal temp on the back sticker will cause another 10% loss. 190-(190 x .3)= 133 watts is around what a working panel should do with an mttp charge controller. With a PWM charge controller on a 12 volt battery the performance will be much much worse, because you will lose most of the voltage across the PWM. A faint shadow going across the panel will cause a significant loss of power. I can't estimate a value but is more than a faint shadow that goes along the long dimension up and down of panel. 72 cells *0.63 =45.4 volts So each substring voltage 24*0.63 =15.1 volts Go luck
@@petercamusojr1545 my panels, in the UK, produce 80-90% of the output specification written on the label. That happens in late spring when the sun has some strength but ambient temp is low.
@@yodab.at1746 alright what about in the heat of the summer/early fall Which Andy season is currently. I tested my 100 Watt panel with a mttp CC and got 72 watts, USA Va here wet with tree canopy, early spring. I went hunting for the loss. Will Prowse in the nevada desert gets 78 watts with an Mppt cc and 70 watts with a pwm cc he did a compassion of different cc was a good TH-cam. Then I went to the modeling and equations to predict the losses, currently working on a matlab scrip to solve for parameters. I have observed more than one person say less panel output do to summer heat.
It just gets better soon you can open a solar school lol. My voltage drags down in mppt action too but the amperage goes up you have explained voltage very well I like it now you can explain the amps. Good stuff now where's the beer drinking Frog. Frogtastic
Your videos accompany my morning coffee...
Good Morning then ;)
didn't you missunderstand this channel? Shouldn't that be your morning beer? ;-)
Very interresting ! I ran to my new panels to verify the voltage. All works fine.Thanks 👍🏻
Great explanations. The pwm controller forces panels down to the battery voltage where the mppt lets the voltage hit the maximum power point. A damaged cell will cause no power on that third of a panel until the backward voltage is high enough. So a lower voltage will allow that damaged cell string to produce.
Yeah, I'm thinking of testing the panels individually with an MPPT controller and see if they reach the max power point or close... Just need a 12V battery.
Very nice explanation of those diodes Sir 🙂👍
Thanks 👍
Looking forward to the next video.
Coming soon!
Andy, one thing I do not understand, how can a diode which is bypassing 24 cells for reverse voltage and current become conducting. Aren't they just for reverse polarity protection of the panels (so to minimize the risk of heating and destroying your cells when you feed a plus of a charger on the minus of the cells) how can the diode ever become conducting when he is against the polarity which you draw from the cells. It looks like the same diode you put on a motor circuit to protect your setup from negative voltage to your charger/mosfets. Or am I completely wrong? I think on a dc motor it's called a flyback diode to my knowledge. If you have cracks in the cells between the minus and plus you should not see any voltage or should not get any current, a crack never passes any electricity. So if you see a high voltage and a short current of 5A then there should be connectivity, cells cannot be interrupted otherwise no current will flow. If a cell is 0.5 V having 72 in series will give you 35V, having a short of 5A will mean a good connection, those diodes will not conduct only in reverse polarity or if they are somehow voltage driven when in full load which would be strange. Just my thoughts, good luck and have a beer! Can you try your electronic load with wich you test your cells as it's not dependend on the voltage of the cells which might be to less when supplying power/under load on your charge controllers. Or have 4 12V lamps in series would show if the cells are able to deliver power.
The bypass diode is for when you put panels in series. Imagine you have 10 panels in series (like on my house). That's 30 strings of 24 cells in series, right. If one of the strings is shaded, the other 29 strings will deliver full power and just push through that shaded one. The current will always look for the path with lower resistance and finds the diode in parallel to the shaded string. And because the voltage from the other 29 strings is much higher, it keeps pushing amps but through the diode.
The diodes are not so relevant in smaller installations I guess.
Hâte de voir la suite de cette analyse et recherche...
Merci beaucoup!
You have Victron 150/35 charge controllers.
These have a maximum of 150v input, however the spec sheet states "MPPT range 60V to 130V".
Your problem strings are at 145v so exceed this....so you are possibly outside the MPPT range and the controller may be pulling the panel voltage down.
Yes, someone commented this on a previous video. It’s not good to be at the max voltage that any equipment can use. When the panels are shaded he is seeing vmp. When panels get full sun he might be going over 130v. Causing the mppt to not operate properly
Too much information is missing. Or ignored.
@@yodab.at1746 that’s not really helping. What info is he ignoring and or missing
@@RJ-cc1fz1. battery soc.
2. What the entire strings produce BEFORE the charge controller
3. The settings of the charge controller.
I get frustrated by a non methodical approach to fault diagnosis, especially when panels are carted up and down ladders adding to serious injury risk.
It seems obvious to me that a simple test of what is being produced before the charge controller would yield good info and from that show the direction further testing needs to happen.
No, No, No. That's not the point. The panels are within the specs of the SCC and perfectly within the MPPT range.
You never run the panels with 145V at max power. That's just not true. It's the Voc what you're referring to not the Vmp. I explained this in the video.
You can visually inspect for micro cracks with the help of a infrared camera and by giving the panel energy from a lab power source without the diodes on the back. Then look at the panel through the infrared camera and you will see the panel light up( yes it makes infrared light). Where the panel does not light up you have the problem. Keep up the good work! Awesome videos as usual!
Thank you. I always wanted an IR camera!
@@OffGridGarageAustralia You can make one out of any digital camera. Just open it up and remove the IR filter from the cmos image sensor after you remove the lens.
Now calculate back the watts per m2...you will be surprised what the outcome is....old pv vs new pv, it will really surprise you, if the panels are good.
Keep up with your series, you do a very good job, keep up the good jod!
Dear my friend , yes the crack do this problems but the PID still do too much problems for the old pv panels , these old pv panels without PID protection.
The cracks and the PID , both doing same decay in power.
Thanks for the PV vids Andy. I have picked up 12 200W suntec panels for $50 and expect to do some troubleshooting shortly :)
The good old Suntec. Nothing than trouble for me... hope yours are much better.
Doing a great job Andy!
Cheers Andy, very helpful 👍
My 2 cents worth.
These panels have a problem, they aren’t going to be the ones that will be the permanently on the west roof.
You have plenty on other panels. Decided which of these you want on the west roof.
Set them up on the ground in the configuration you want and test if they work. If so install them.
Then you’ll have plenty of time to come back to them at a later date without compromising the charging of your system. And yes I’m as keen as everyone else to find the problem, but it just seems he’s missing out on full charges while doing all this testing.
Anyway keep up the great content, and if you need a hand doing any of this get in touch, I’m just up the road at Buaraba ( between Gatton and Esk). I’m happy to help out.
Thanks, Dave. That's probably good advice. I could also test all panels individually with a MPPT SCC and a 12V battery.
Diodes will almost always fail as short circuit so if a diode was damaged you will have easily noticed in the open circuit voltage as that will have been reduced by 33% but that is not the case base on your earlier tests. The only explanation to reduced current is that some of the cells are cracked so they act as a smaller area cell thus lower current. Is just strange that all panels will have more than one cracked cell. Curious how this new panels you got perform since they are from a different installation but same type of panels.
You only need a single cracked or shaded cell in a group of 24 series cells for that string to be bypassed by the diode assuming that the other unaffected 24 cell strings produce enough voltage for the battery or MPPT minimum require. When you do a short circuit test and that results in a proper current say 5 or 6A in full sun then all you can be sure about is that one of the 3 groups of 24 series cells is OK you have no idea if the other two are OK or not.
Very good analogy Dacian, good to see you here 👍
That is a good point. Thanks again for your comment on this. I probably connect each panel individually to an MPPT SCC and cool down the panels in full sunlight to see how they perform. Even if I find out they have one string not working correctly, there is nothing I can do about it. It affects the whole panel anyways and is not repairable.
Andy, Great explanation as always. Why not use an std extension cord and fashion some adapter ends to get more than 7 meters. That would allow easy setup and you to troubleshoot the problem on the ground. Keeps you off the ladder unless you are lifting a solution not a problem :)
I made panel by mono crystal line cells and some panel with crack cells too but they still gave good power unexpected ly they still working fine reverse polarity diodes are silicone type still working good 1N5408 three in parallel also used in your case Schottky is used?type.. please find if they are ok...thanks for nice information you are giving
Diodes are for shaded sections to be by-passed, sometimes current splits to both diodes and cells.
For every panel to have the same random failure? Seems unlikely. I am at a loss as to what is going on, but if it was just down to cracks then you should expect to see variations in performance across the array.
More information is needed for diagnosis. Much more.
If a car stops working, do you take the entire engine apart before checking the gas tank? Or see you have fuel, so condemn the gearbox?
Fault diagnosis is a skilled job, and can only be done when you understand the whole system and how the different parts of that system interact with each other.
15 A für die Dioden ist gut dimensioniert.
Aber ein Blitzschlag in der Nähe kann die Kennlinie versauen so das die Diode nicht ganz defekt ist.
Deswegen schlage ich vor die Diode versuchtechnisch abzuklemmen wenn du einen Fehler bei den 3 Strängen innerhalb des Moduls findest.
15 A for the diodes is well dimensioned.
But a lightning strike in the vicinity can ruin the characteristic curve so that the diode is not completely defective.
Therefore I suggest trying to disconnect the diode technically if you find an error in the 3 strings within the module
Heya, oke watched the 1st 100 vlog still have 362 vlog's to watch thank you for showing everything till now I learned a lot from you vlog's
I always cracks me up at how low current the bypass diodes are rated for. The panel will have 9-30A running through it. Yet the diodes are something like 3-5A ??? How the hell is that supposed to work. I have no answers, but some of the others here may know something about this?
I know someone has probably mentioned it, but... maybe substitute a panel on the other roof and see if the problem follows.
We use an insulation tester to push a low current, high voltage injection throigh electrical equipment to test if it has a short or is down to earth.
Perhaps you could find a way to utilise an insulation tester (aka megger meter) to push the higher voltage through the panels to see if the electrical pressure shows up with a short 🤷♂️
I’m gonna try this on a good and bad panel when I get home and see if it tells me anything.
If you have access to a thermography camera you can Diagnose damage panels directly in the roof. Just search for Practical guide Solar Panel Thermography.
Failure mechanisms of bypass diodes are put in two categories of catastrophic and wear out, but what is localized breakdown at the die level (which supposedly can eventually cause a complete failure of the panel or cell or device(Satpathy & Pamuru (2020), page 177, Solar PV Power: Design, Manufacturing and Applications from Sand to Systems))? Next, the electron model of current flow from the cathode, then the conventional model of current flow with the arrow as flow from the anode (which is a fictional convenience), mixed into aggravating flip-flopping, near-failing, bi-directional, physics of worn diodes add to the paradoxes.
Good look, but I think the problem is beers are not cold enough. 😂
Hi Andy, this is not a comment on the panel problem , but another different issue, i am going from agm batteries to a lithium pack with BMS, i have a major concern to what happens if the BMS switches off with a fault either over or undercharging? I have a midnight classic controller and victron phoenix inverter, if you suddenly disconect the neg line from both this seems like a real bad idea, especially regarding the charge controller as it would be left with maybee two kw of solar coming into the controller with nowhere to go, and how would the bms handle all tthis current if it comes back on line? I am begining to think all the equipment should be commiincating from the bms to shut down the inverter and charge controller safely, and not all the eqipment has a comms facility. Ps love all your fault finding, watch your vids every day.
@@lenfryers2839 Hay lens. Your comment has been published as an answer to mine so, I think, it can be not read by people that can help you. I don't have batteries, I'm just in-grid (for now). Publish it in the main thread and good look. Have a nice day
Great video!! Thanks for sharing😊
Thanks for watching!
Very informative video, thanks:)
Thank you.
saw someone check if the cells were broken by switching on power on them (do not remember if it was correct or reverse polarity) and got defects to start shining but not much so it needs to be really dark but then it should be possible to see problems
Andy if you are getting a huge voltage drop when the charge controller is trying to draw current, sounds like you have a faulty panel somewhere!
It's kinda like a week lead acid battery! Under heavy current draw, the voltage will sag! This is how they test lead acid batteries! The testing device is nothing more than a heating element with a voltage gauge! (Well the old battery testers are anyway which are still being sold in Bunnings! And all over eBay for around $25 - $40). I'm not sure if a short circuit test with the current clamp will show up the fault although I don't see any reason why it shouldn't, otherwise you might need a high powered load Resistor with MC4 connectors and an analogue voltage gauge because sometimes the deflection of the needle is more important than the voltage which a modern multimeter with an LCD screen is not very good at doing!
Great class professor! Thank you..
You're very welcome! 😂
Instead of opening boxes, put a variable load on each panel. You will see bad ones at 35V make 2A bump to 3.5A at 25V.
Yeah, I could do this too. That's a good idea.
Maybe you could use that battery capacity tester on it?
micro cracks might be real problem... hope you will find them :)
If so, nothing I can do about it, right? The panel is either not fully performing or toast altogether.
@@OffGridGarageAustralia Your tests seem to have confirmed that one or possibly two strings within each panel have a fault that effectively introduces a high impedance in the string.
An open circuit voltage will not detect the high resistance cos your meter impedance is many times higher.
Similarly, the bypass diode on the panel will turn on when you do the short-circuit test and bypass the faulty string, just as it should do.
Expose the diodes and the string terminals and measure the voltage on each string when connected to the MPPT charge controller. This will show up the faulty string.
You can then solder a wire across these terminals to produce a fully working panel of lower voltage, same current and hence lower power.
If you look carefully at the MPPT controller readings at the moment the controller is hunting for a maximum power but is unable to find it. This is because the controller is designed to work with low impedance panels but yours are now high impedance. If you short out ths faulty string you will reduce the impedance back to normal and the controller will work again.
You will find a different MPPT controller other than Victron will give different results; some will perform better with high impedance panels and some worse.
Good luck!
Andy, I would install used panels on my roof but... I'm waiting to final diagnosis of yours problems! And simple solution: how to test used panels before installing it.
I have reviewed all the movies. I do not believe so this is a problem with cabling. I also do not believe so that is the problem with the broken panels - is possible to have one broken, or maybe more than one, two - but all nine? Not possible. Maybe three is one broken panel, each in one string - possible… you could change that nine panels between them - simply scramble it - one string should be better than other.
Most interesting thing is in 3rd part in series. 2 panels on the ground working well with PWM charge controller but not working with MPPT Victron controller - WHY? It should not happened - what is the difference? Algorithm and…. battery voltage. Could you use Victron to charge 12V battery - what current than will be?
Maybe, just maybe, you removed from the roof two good panels, which works correctly with PWM charger - but voltage on two panels was too low to power 52V battery! Look to the movie - from the string you received only 65V at that moment. Is this enough to charge 52V?
IMHO checking panels for open Voltage and Circuit Current give information that all should be OK. (however checking diode is good idea) maybe that power is produced by panels but lost - for example to the ground - some static resistance or passing current under highest voltage - that’s why yours home installation signalling fault (grounding) but when you measure that with 1.5v mustimeter that is not showing.
This is perfect bro. Thank you 😊
Looking forward to this next video. Great stuff. The ir camera or even IR temp gun on front of panel, can we see an difference?
I'd like to see the difference between tree shading and cloud cover shade. I think that with tree shade where you have spots of sunlight on the panel, the shaded portion is actually similar to night-time, and therefore the overall resistance of the cells altogether is very high and will generate less than a uniform low light condition.
I'm not sure why the overall output of the strings hasn't been measured before the charge controller, because with that one simple measurement, alot of info could be got. Also, I'm not sure why soc of the battery is not shown.
Narrowing down the cause of a fault is a methodical process that involves all parameters because all parameters make up the system. Never make assumptions. There is so much information lacking, its impossible to make an informed diagnosis.
The diodes were good to check, but after seeing reasonable production when each panel was tested under short circuit condition, they seem to be fine.
Going back to the connected strings, and measuring before the cc, if you have expected output and the cc is ruled out due to being swapped for a known working one, the next check is the bms.
We have a cloudy morning today and the panels perform better than with the tree shading as the light is more consistent through the clouds.
Somehow neither the voltage sense with an open circuit nor the current test with a short circuit make a lot of sense, because in both tests the pannel doesnt deliver power. I would measure the volatage of each module while they are connected in series and delivering power. It's quite unliekely that all the modules in one string have as same as many and as same as bad cracks. So if cracks are the reason for the low power, the modules should have different voltages.
Yes, that is correct. The best way would be to test the panels individually with an mppt charge controller and see if it can pull the full power in full sunlight. I can also cool down the panel to 25°C for such a test.
@@OffGridGarageAustralia That would be a lot of work. What I meant was measuring the volatge of each module, while they are still connected in series on the roof. Therte are fine measuring points (needels) for volt meters that can pinch throough the insulation of your solar cables. This way you could compare the voltage of all module sin a string wthout a lot of effort. If some of the modules show a lower voltage than the others, you can take only those ones down for further tests.
For testing I would also prefer a PWM controller if there is one available. Those dont do any convertion of voltage and current, so show the current the modules really deliver.
You should recognize the white goo as common thermal contact paste, similar to zinc oxide sunblock, put in to transfer heat out of the diodes. Replace it.
It's actually silicon, soft and squishy!
" Silastic " it's a silicone elastomer glue commonly use to secure electronics.
If it's got micro cracks in the cells it should have snail trails starting to develop where the micro cracks are you will be able to see them in full sun or should be able to
Get enough power resistors to make 6 ohms 190 watts and check each panel for voltage and current. I don't know if 3x 60 watt light bulbs would work or not. Maybe 3 automobile headlights in series.
Taken in isolation, short circuit current isn't hugely helpful in diagnosis. Same with voc. But together they tell good info. Voc tells you all cells are connected in the string. If that measurement is in spec, then the current measurement (short circuit) will give good info.
There are situations where no load measurement is in spec but as soon as a load is presented either current falls or voltage collapses. This indicates a poor connection within the panel. A bad connection will often have voltage but as soon as current flows, will go high resistance. The side effect is often heat at the point of the bad connection.
Yep, that's exactly right.
Andy good morning from Grass Valley California USA If you had 32 of your cells would you make 2 48volt packs then join them or 1 big 48volt pack? and would there be any pros or cons to one way or the other thanks Jimmy
That is a great question I'm asking myself for some time now. I have 16 more cells coming in and I'm tossing around some ideas in my head if I should build one large battery or two separate batteries. I will discuss this in one of the next videos...
@@OffGridGarageAustralia ok great can’t wait for the video I’m waiting for the rest of my parts then I’ll be building it. I’m building mine for when the Power company PG&E shuts the power off during hi fire risk times when it’s windy and dry Love your videos keep them coming
Probably got an order coming for ALL the diodes...
You just opened up a can of worms.
I know. Actually 6 cans at the moment... 😒
Hi Andy, you do not have or can rent an infrared camera? If yes so look through it at a panel under full sunlight and defective cells should be seen as hot surfaces ...
I haven't got one yet, no... Well in full sunlight, all panels are hot 50-60°. I doubt the faulty ones will be even hotter.
As others suggested it might be better to supply DC to the panels and make them glow at night.
@@OffGridGarageAustralia Yes, it can work too. A simply defective cell should behave like an electrical load and should therefore heat up more than those that are OK.
Verify with amp on the output of the controller when its say 2a verify output controller to the battery probably as more amp
Can we test diode when panel is outside but not connected to the system?
I wouldn't move them down.. make it simple..make simple two T connectors and just check voltage drop on every panel. It should be about same ..if you see much smaller value than rest of the panels then you will know there might be something wrong with that one panel. It is not likely that all panels are not good...
Risk assessment while trying to diagnose is very important. Lugging panels up and down ladders to fault find is stupid and dangerous.
You said lots of beer donations, easy with the beer, don’t drink and troubleshoot, you might get electrocuted or fall off the roof,remember you told us SAFETY FIRSTdid you recall!JUST DRINK ROOT BEER!
“OHMS LAW “YOU MADE ME REFRESH MY BASIC ELECTRONIC DAYS!
E = I multiply R
R = Voltage divided by current
I = voltage divided by resistance
One thing you did is test all strings seperatly, so if one panel is faulty in each string the voltage would never be 110V unless one of the cells only delivers 2A, it's series so the weakest link is in force here. Will your diodes conduct when the other panels have a higher voltage when placed parallel, I don't think so. What I can imagine is in parallel you might feed the panels with current from the good string, so warming the cells instead of getting power from them, this you should be able to measure feom within your garage as the current direction should be seen with your meter. So one string feeds maybe 2 faulty strings. Sorry this is such a long story, the problem doesn't seem simple.
I probably have to test the panels individually with an MPPT charge controller and also cool them down to 25°. Just need a 12V battery for that...
The diodes will not conduct as long as the produced voltage of the single string is high enough to deliver power to the overall panel.
Thanks Andy, very systematic and open minded trouble shooting. Much appreciated. Cheers Kevin
Thanks Kevin.
Me panel maybe broken or multimeter series connect but up vats and amper 2 woow
What model clamp meter are you using? And do you recommend it?
Ich glaub ich muss mal dringend aufs Dach...🤔 ich hab da ne Vorahnung
OL means "open line", never knew this untill this week. Saw on it another channel.
When I was in engineering school (*cough* years ago) it was "open loop"
I always have Overload in my head as older multimeters shows this when it goes out of range.
" Oh Lord" is what I have always said . lol
@@OffGridGarageAustralia Overload is correct because voltmeter in diode tester has over 2V input voltage (with open contacts) on 2V range.
In voltmeter mode it shows also OL - what "Open Loop" then...
It's from old multimeters ;o)
Du weißt doch jetzt wo die Dioden und wo die kühlplatten sind.
Warum bohrst du nicht 3 Löcher in das Gehäuse bis du auf das Metallplättchen kommst :)
Dann kannst du darüber auch einfach messen!
das würde dir eine Menge Arbeit und zerstörung sparen! :)
Am Ende mit einem kleber silikon oder ähnliches wieder abdichten.
Weiter so und ich hoffe deine Zellen laufen bald wieder in der Australischen Sonne! :)
Keine Gewalt ist auch keine Lösung 😆
⭐️
Can't you take a small CC to the roof and make your measurements there?
Why? I took the panels down on my driveway and measured there.
Mc-3 or mc-4 connector issues…?
Spray some copper paste ….. solves three issues…..displace moisture….improves conductance….circumvents poor clearances…..
I told you about the cracks in an earlier vídeo hehehe
Ight here we go.. 🍿 🍷
Good show my good man! Ready for the next.. 🌞 ⛅ ☀
Thanks David!
@@OffGridGarageAustralia Welcome Andy! Good morning! I see a victron update is ready. i wonder if they're releasing the version you've been testing 🤔😉 Cheers n sunny day to ya bro 🌞☀🌤
I would quess the panels that had connectors cut off are the bad panels. Not lazy, instead preventing use of known bad panels.???
Most people cannot crimp an mc4 connector.
That's a conspiracy theory that damaged panels are routinely sold. The supplier would go out of business quick smart if that happens. Its a lazy theory.
@@yodab.at1746 I didn't think he bought from a dealer. I know they are used. I assumed he got them by picking them up from houses after retrofit. I don't do conspiracies. Just thinking out loud.
It seems like a good fire safety practice to prevent bad panels being used.
@@petercamusojr1545 there are many dealers selling used panels. Bimble Solar in the UK being just one (with a very good rep)
How would the installer who is on-site to install a new system know if the existing panels are faulty or not? They don't test them.
@@yodab.at1746 I would expect a reputable dealer would never sale a panel without an mc4 connector nor a broken panel. I would not expect an average home owner of grid tied system to be able to test if the panel is good or bad. Buying used is often a roll of the dice
1000 watts per sq.meter is what solar power reaches the earth on a clear day with the sun directly overhead.
Your problem is the charge controller shunt is giving you problems ✌🏼✌🏼✌🏼✌🏼
.2 volts?! What are they, Schottky diodes? :) Can you read a part number?
yup
@@edc1569 Checked Amazon for "solar barrier diodes". They're all 12SQ040 series. Got a spec sheet. They drop .2 volts at zero amps and about .4 ish @ 5A. Thanx :)
I think all solar diodes are Schottky diodes. These are the B1445G diodes. 15A max and 45V blocking voltage.
I like your videos but they’re mine boggling
The star frog was not in the next video!!!! Ha ha
I didn't say that. It will in *one of the next videos* 😉
Who else had to pause and watch 3x times until it was clear, that it was a lizard at 12:57 ?
Its a simple circuit. My god how much simpler can it get ?
Ground faults can give the exact problem you are enduring. Fitting second hand panels without meggering and recording all the results is a waste of time.
Ask if you want an explanation.
When commissioning a new system it is necessary to perform the above tests.
Diodes that appear ok with diode check at 1.8 volts to 2.4 volts but will behave as a short at system voltage.
I teach tradesmen this type of fault finding. I make simulators that check ok on diode check but are faulty.
Diodes only fail open from mechanical damage. Electrical failure is a short circuit.
Testing in the driveway with no frame earth will just confuse you.
The panel you removed from your house because of ground fault is strongly implicated in your problem.
The discussion about short-circuit current is completely bogus.
Short circuit current (Isc) and open circuit voltage (Voc) are the critical criteria.
The only way a single cell working and the rest dead is if all the rest of the cells are shorted, and that would be obvious by the Voc.
Each cell should measure the same temperature, 45 dec C is the nominal, but could be higher. A colder cell is a nonconductive cell. An extremely hot cell indicates I ??? don't know.
Since you are 27 deg south latitude, I am going to estimate a 10-20% loss of solar intensity for air mass thickness.
The difference between STC temperatures and the nominal temp on the back sticker will cause another 10% loss.
190-(190 x .3)= 133 watts is around what a working panel should do with an mttp charge controller. With a PWM charge controller on a 12 volt battery the performance will be much much worse, because you will lose most of the voltage across the PWM.
A faint shadow going across the panel will cause a significant loss of power. I can't estimate a value but is more than a faint shadow that goes along the long dimension up and down of panel.
72 cells *0.63 =45.4 volts
So each substring voltage
24*0.63 =15.1 volts
Go luck
I believe the type of shading is the clue as to what is going on.
@@yodab.at1746 I really do expect a 20 to 30% difference in output in the real world compared to STC.
@@petercamusojr1545 my panels, in the UK, produce 80-90% of the output specification written on the label. That happens in late spring when the sun has some strength but ambient temp is low.
@@yodab.at1746 alright what about in the heat of the summer/early fall Which Andy season is currently.
I tested my 100 Watt panel with a mttp CC and got 72 watts, USA Va here wet with tree canopy, early spring. I went hunting for the loss. Will Prowse in the nevada desert gets 78 watts with an Mppt cc and 70 watts with a pwm cc he did a compassion of different cc was a good TH-cam.
Then I went to the modeling and equations to predict the losses, currently working on a matlab scrip to solve for parameters.
I have observed more than one person say less panel output do to summer heat.
😊😊👍👍😊😊👍👍👌
🤦♂️🤦♂️🤦♂️🤦♂️🤦♂️🤦♂️🤦♂️🤦♂️
I think I was first maybe second this time.
You are first Buddy. 👍
🤦♂️
Second :D
It just gets better soon you can open a solar school lol. My voltage drags down in mppt action too but the amperage goes up you have explained voltage very well I like it now you can explain the amps. Good stuff now where's the beer drinking Frog. Frogtastic
Thanks, my man! 🍺🐸