Solar Panel Diodes. I found the problem (for a moment)...

Finally some results about the panels. I can measure a reverse current across some of the diodes. Does that mean they are faulty? Yeah, well, I managed to find 2 panels where the diodes measure OK. Let's put them in series and to the test! I'm excited. For a moment...
Please subscribe and join me on my journey of the Off-Grid Garage!
For more information, please check out my website with links to all batteries, materials, devices and products: off-grid-garag...
New to AliExpress? Use my referral code below to register and earn US$19 store credit. What a great start! a.aliexpress.c...
The 'buy me a beer 🍻' link is here (scroll all the way down):
off-grid-garag... Thank you for your support!
**** These are the batteries you are looking for ****
🔋➡EVE LF280 LiFePO4 cells 3.2V/280Ah: off-grid-garag...
🌞➡Solar, Controllers, Inverters and accessories: off-grid-garag...
🔌➡Batteries, charger and accessories: off-grid-garag...
⚡➡Cables, inverters and electronics: off-grid-garag...
🔧➡Tools, analysers and testers: off-grid-garag...
Take a sneak peak at the Off-Grid-Garage in the Victron VRM World:
vrm.victronene...

[@diyernh]

There was a commercial solar array that uses a thermal camera. By feeding power into the panels at night, the shorted diodes actually heat the panels. They just look for the hotter panel and know the diodes are shot.

[@ykchowaa]

Good idea. Just apply the revise voltage at night. if the string has current then meaure or checked by thermal camera for that

[@raytry69]

Or just measure them at the Sunlight without load. If the diode is faulty, that will generate some amps and heat up the cells.

[@HansKeesom]

For the moment, test all the other panels one by one with the two panels so you can sort the stack of panels that you have into good and bad panels. Then install the good ones on the roof again and worry about the bad ones later. No need to open the good ones.

[@Useitorloseit1]

Good morning Andy, I know this video is 2 years old. But it's today for me and your Joy is infectious. Your your Joy and perseverance is infectious, thank you

[@nigelcharles511]

As the conductor areas are large inside the box and you know their layout you could just drill small holes in the cover in the correct places and push the cable probes through. Afterwards just seal up the holes if the panel is good. The other observation is that the panels themselves might allow reverse current. You would need to disconnect the diodes to check. Diodes don't often fail closed circuit.

[@teardowndan5364]

In my own experience repairing electronics, semiconductors almost always fail shorted unless the overload is large enough to blow them up.

[@leonclose7823]

I thought the same thing about drilling holes for testing. A dab of sealant afterwards.

[@sinusgolf]

Fantastic tip

[@sergioalves4204]

Lessons learned: Always watch Andy's videos until the end! :-) just love it!!

[@Gary-wh7ce]

Good content as always. I found a faulty diode when a 100w panel should provide 21v voc and I was getting about 9-10v dc. One panel had a faulty diode and the second panel has a lifted solder joint. In both cases this saved me $$ each time and the fix was less than 50 cents.

[@petercamusojr1545]

If you suspect a bad solder joint. You can buy a can of "cool spray for electronics". Spay on the solder joint while using the diode check meter to check for a bad intermittent solder connection.

[@julianfine29]

I am glad I am not the only one having problems with Solar, I have done a massive system at my home to take me off grid with 40 x 330 Watt panels, 2 x 250/100 victron M PPT, 1 x 15 Kw Victron Quatro, and originally 20 x 225 Amp AGM batteries. The batteries only lasted 1 year even though I only discharged to 50%. Replaced with 6 x 4.8Kw Lithium Phosphate packs. But I made one fatal mistake. The Lithium Packs came with Copper Lugs which I replaced with Tinned Lugs on 5 but forgot the 6th one. Struggled to get the 6th one to work until I found that the Copper connected to my 50mm x 50mm x 5mm Aluminium bus bars had corroded within 1 week and that battery was no longer in circuit, due to corrosion with dissimilar metals. Moral of the story DIY works but simple mistakes can be made. By the way I am in South Africa where we have rolling Massive blackouts almost every day. But I am now completely off grid :).

[@OffGridGarageAustralia]

Thanks for sharing Julian. Great to hear your on lithium power too now. There are always these nasty little things in your project which can completely throw you off. And everything takes forever if you need to search for your own mistakes...

[@normtomlins3821]

OMG, the ending... your killing me!

[@popcoingaming5086]

This is the biggest cliffhanger video you have put out. I can't stop thinking "if it's not the diode, what else could it be?" Eagerly awaiting the next video Andy.

[@OffGridGarageAustralia]

Hahaha, sorry for that 😉 But thanks for watching until the end. Most here haven't...

[@zarbonida]

PID Potential Induced Degradation

[@electrojessy4270]

@@OffGridGarageAustralia is the next video here tomorrow? I cant wait lol

[@danielglang6312]

@@OffGridGarageAustralia I usually watch til the end but not always. I do however watch individual videos multiple times just to check the comments and reply if I might be helpful. I suspect that just clicking on the post will bump up KZbin's watch count and help boost income from them.

[@andrewsweet43]

Yeah, to have a definitine answer about the diode problem, one must unsolder one of the problematic diodes and check it off circuit... Judging from my previous experience with diodes, 99% of the time they fail shorted. What you are measuring there probbably comes from faulty cells which is a bad thing unfortunately... You probbably realised that ath the end of the video :) Either way, kudos for the effort! Oh and an I-V curve tracer instrument would be great to test individual panels for problems like that! You would see the problem immediately on the characteristic curve. I am building such an instrument right now as my Master's degree thesis!

[@OwnMyBit]

I'm glad I watched to the end because through it I was thinking "I don't believe it is the diodes, 200mV one way and 1.4V the other doesn't indicate faulty diodes!"

[@OffGridGarageAustralia]

You're probably one of the three people watching until the end so far 😉. Yes, that is very suspicious to start with, right.

[@zarbonida]

@@OffGridGarageAustralia it's a PID problem

[@ArthurDentZaphodBeeb]

Always a great moment when you figure out such annoying, confounding problems.

[@DavidBTB]

Oh no, I gave a celebratory thumbs up before , "it's not the diodes". What a rollercoaster of emotions. :)

[@georgev.]

Great Diode discovery Andy and April's fools, jajaja or not maybe cell damaged if your measuring voltage. Apply the Diode reverse voltage and see if it blocks the current.

[@MindMeetMaker]

Maybe try using a hot air gun on the back boxes to soften the glue before trying to open them it may help?🤔

[@TBL_stevennelson]

I think that sounds like a good idea

[@russkubes]

I'm interested to see the next video for what you find next. I'm wondering if you took some diodes off and tested directly and found they are actually working, but maybe the reverse voltage you measured is making its way through the solar cells somehow. Is that possible? Really thought you had something there, but you'll get it.

[@friendlyguy2542]

Please keep us posted on replacement diode source and cost. Great videos and quite entertaining. I hope you work out a maintenance plan to keep an eye on panels and find problems shortly after the occur.

[@karensar9304]

I am very glad that I found the reason well done Ser

[@williamfairley2502]

Congratulations Andy love your videos

[@AveRage_Joe]

Having a good Ole fashion Diode replacing Party!!!! Wait WHAAAA😮😏

[@OffGridGarageAustralia]

Thanks for watching until the end! 😳

[@AveRage_Joe]

@@OffGridGarageAustralia Always!!💥🤘

[@tomsutherland3438]

Unsolder them then check hehehe and yes i watched until the end too!!!!

[@RemScanProducts]

Great News !! Love watching and learning from your experiences. Thx

[@etubbin]

This channel is like the tv show Survivor but for geeks. I love it. Keep it up Andy.

[@rohbags]

If it's not the diodes, then it must be a faulty/shorted cell in the same string the apparent faulty diode is across. Did you de-solder a diode to test it at the component level, outside of the circuit? Can parallel a few common 1n1004's as a temporary diode substitute to test the difference on a single panel to rule it out.

[@peterevenhuis2663]

I am happy that finally you opened the connection box on the panels. So soon you will be back with full panel power.

[@allandeans3904]

Andy, 5 seconds into your video there is a clip of your panels on the roof. The left hand one shows some diagonal cracking in the cell at the bottom. I have experienced this myself and it generally leads to lower output depending on how many of the cells are cracked. Just thought I would mention it as it was just like the cracking I saw on one of ours (also in Brisbane area). Not suggesting this is your current problem but I think it will get worse and as you have spare panels maybe this one should be pensioned out. Cheers Allan

[@NaughtyGoatFarm]

Hi Andy. It would be great if there was some way to test the diodes without opening the connection box. Surely there is a way. What about a few strategically placed holes drilled in the box that can then be sealed after with silicon or tape.

[@OffGridGarageAustralia]

There is actually a diode tester out there for solar panels. Well, diodes is only one thing it measures, also does isolation and performance I think. Costs around $1600 though.

[@NaughtyGoatFarm]

@@OffGridGarageAustralia wow that's lot.

[@teardowndan5364]

A simple way to test diodes is to use a bench power supply: set the current limit to 3A, set the voltage limit to panel open-circuit rating or whatever your power supply can output if it does not go that high, then connect panels in the bypass direction while fully dark. If the diodes are good, the diodes' forward voltage should be around 2V (3 x 0.5-0.7V + wiring losses). If the voltage is much higher, one of the diodes is open-circuit and current is leaking through the array. If the voltage is much lower, at least one of the diodes is shorted.

[@adamelkoumy5141]

Thank you very much for sharing the knowledge.

[@65bigfalconable]

Nice work mate... panels that would of been taken to the bin .... now have a new life of making free energy....

[@polvincompoundsadmin5402]

Best way to test your diodes is under load, use a 12v battery and a car headlight bulb (50W) gives a better result.

[@RJ-cc1fz]

Awesome! Finally fixed! Good job. I’d absolutely help you fix the rest of them if you didn’t live across the ocean

[@panospapadimitriou3498]

wish i could go back in time and open more electrician books than plumber and car mechanics... !!!! i think i lke it!!!! with my character i could be completely bald due to stress but i love it some how!!!

[@davegeorge7094]

I would unsolder one leg on the diodes and re-measure. It could be a cell string short circuit and not the diodes but unlikely. This is new territory for me and been off grid for years.

[@OffGridGarageAustralia]

You're spot on Dave!

[@tomsutherland3438]

✓^^^^

[@garysmith2104]

Great accomplishment. Well done.

[@MrMalzbie]

Eine Achterbahn der Gefühle! Also ich würde ja Löcher in die Deckel bohren um die Dioden zu testen. Man weiß ja jetzt wo man messen kann. Aber jetzt bin ich erstmal gespannt was nun wieder is.... Mann! Du machst es spannend!

[@OffGridGarageAustralia]

Huch, wo ist denn dein Englisch hin? ☺️ Es ist alles Mist und geht nicht voran. Was einfach aussieht dauert alles ewig weil man ständig neue Probleme findet oder selber macht. Ich hatte einfach neue Panels kaufen sollen...

[@MrMalzbie]

@@OffGridGarageAustralia Och, es war einfach noch zu früh for english. Ja, diese Probleme sind schon blöd. Aber es ist auch super für deinen Kanal! Das hilft garantiert vielen Leuten. Ich schaus mir jedenfalls gerne an! :)

[@OffGridGarageAustralia]

@@MrMalzbie Du bist der Beste!

[@andrewkingsley7221]

Thank you for the series on the panels . You have answered my question for the panels here in my area as well. Now for the work to check and replace DIODES.

[@OffGridGarageAustralia]

You should watch until the very end!

[@zarbonida]

It's not the diodes

[@mikeman230]

Now you know the layout of the box make a drill template. Then check through small drill holes. Then seal with silicone

[@ranig2848]

Easy to diagnose - get one good panel, connect another panel which is under test and connect to solar charger. If it works well, panel under test is fine. If not... that other panel has an issue.

[@henvan8737]

As you will change a lot of diodes try to get some with higher current and voltage rating, otherwise the same thing will happen again.

[@mikenewman4078]

Could have been lightning in its past life.

[@brunovg]

Just finished to watch all your videos 🤪 What a journey Andy 😅 Keep the good working and solar info. Greetings from Portugal 👍🏻

[@karoneykemboi5480]

Andy: We've got him! We've got him! Uh-oh, false alarm. What a bummer

[@andrewradford3953]

Meatless BBQ sounds good, same beer that I drink - great, nice small panels - count me in! Would be a fun trip in the Leaf from the North side. Difficulty is organising farm sitters for us. Can't wait to see you solve this mystery.

[@RikHaanen]

Congratulations Andy, great to learn all of this from you! Thanks

[@RikHaanen]

But wait a minute... its not???

[@howardadams4072]

Andy, drill 4 little holes in the cover to put the test probes in for testing. Then if OK, you can just silicon the holes.

[@farqend]

I'm cracking one with you!

[@USA-GreedyMenOfNoIntegrity]

Another in Kentucky with you guys. Mine is bourbon.

[@robertolira8331]

You are an electronic wizard! Your DAKINE!

[@Tony-Elliott]

Brilliant video again Andy looking forward to the next video to see what else you've found

[@thorsten_b]

People not watching the video til the end: yay let's crack a beer. All the others: oh no, what a cliff hanger... greetings from Germany.

[@QUADBOYification]

I actually removed the 10A rated diodes from my two 50 Watt panels (in series). That alone got me from 55 Volts to 72 Volts (2 x 36v). The shading is only relevant at certain hours of the day and is already low efficiency. Mine are stacked on top of each other in portret mode and I discovered that panning them on the Y-axis (horizontal angle to the sun like a door) produces a lot more also. It seems that some panels have difficulty absorbing direct sunlight shining exactly straight on them making them hotter reducing voltage, that is what I discovered.

[@thomasmoore4576]

I'm so happy for you you fixed your panels I bet you that's what's wrong with mine

[@DutchAussieProductions]

You made my day again Andy. I haven't seen anyone this happy for a long time. Have a great day. "Its not the diodes" ?

[@vickipps6821]

Well done

[@kipdennis3796]

Good work Andy,Solar is addictive, they always think am not well upstairs always thinking and doing solar system every now and then

[@bobtobb9292]

Four small holes drilled in the cover would allow testing and save cover repair if diodes are good.

[@USA-GreedyMenOfNoIntegrity]

Good job. You’ve taught us a bunch! Thank you.

[@Luke-san]

Now the question is ... why did they fail? Could it be a production fault or maybe static electricity due to lightning. New sub and started watching your progress. Edit... yes I watched till the end of the end now. You can't measure a diode in a circuit One side has to be disconnected. However they should all act the same on the panels since the panels are all the same type and manufacturer. My to go test would be to disconnect the diodes and solder 2 new test cables and connectors. ... One panel by one panel.

[@cristof48]

Bonjour, if it's not the diode, may be a corrosive contact between line or busbar. in this case, the diode is working well on direct direction, but the résistance is bad on reverse direction.

[@OffGridGarageAustralia]

It's something like this. My bet is on micro cracks in cells across these panels.

[@rudikaatz317]

Gratulation aus Brandenburg! (es lag doch nicht an zu wenig gekühlten Bier :) Congratulations from Brandenburg! (It was not because the beer was not cooled enough :)

[@bobonestone1807]

Heat and overvoltage would be my guess I would actually go with a little bit higher voltage on the replacement

[@moinmoin3678]

Best cliffhanger ever !!!!!!!!

[@gladegordon3624]

Great to see you got to the bottom of it! Your sharing and documenting this adventure should help in testing the 22kw used solar panels I've accumulated;)

[@ltcmdrdata4611]

Well done Andy but, with all this (well working) panels on the roof you´ll need a bigger or 2nd battery

[@OffGridGarageAustralia]

incoming...

[@walterweese2527]

@@OffGridGarageAustralia of course you must have enough wage , to pay the bill and tax ;o) greetings to silke ;o)

[@MaxTunage]

Hey bud, if you use an analog meter that takes a 9 volt battery, you should be able to check the ohms in each direction of the panel without having to take the cover off the integrated junction box. Place the Analog Ohm Meter in the high ohms position, you should see the full 9 volts across the Analog Meter terminals using a DMM. This will make it much easier to check your panels while they are mounted, however, you'd have to cover them fully to prevent sunlight, or perhaps rename yourself, Midnight Solar. LoL! I hope you find this to be helpful. Cheers 'n' beerz!

[@philipwinkelhorst934]

Very good found this is big for the secondhand solar community. ! Love your vids! Plant some new trees and get ths big one a bit down. then your shed is not any more in the shadows. Congrats man!

[@mjp0815]

Can be the diodes or could be the cells. To tell for sure you need to de-solder the "faulty" diodes and test them in isolation.

[@apz9022]

Congrats!!

[@jmaus2k]

Amazing job. Totally impressed. You should be able to test without disassembling junction boxes by shading each section covered by a diode while a load is applied. Voltage should change drastically depending on if a diode has failed or not. I bet you could also put a reverse voltage on the panel(or even use meter diode checker) and get a consistent 1.4V per section that diodes work.

[@1981dasimpson]

you dont have to crack them open but drill a small hole so you could get some longer probs in this way they could be sealed again if all is good

[@blindliquor1]

Great Job... love the humor..lol

[@jaminthevanuk296]

Put diods outside under panel for easy maintainance make your own boxes

[@robsouto4998]

When is this time be continued?! I spent 16 minutes of this video so happy that you had found the problem, then you pulled the rug out at the end 😂

[@jasondevine6014]

Well done. I should check my panels more thoroughly after this.

[@nurofenusa1]

Thank you Andy, I now have something to test :)

[@phillipchen7749]

That is so great that you found the problem but better yet I have also learnt something thanks!

[@nigeldunn4467]

Well done Andy! Every panel in your west array must have most diodes faulty! So there is a fair chance they may have hotspots as well? So you still need to test each panel before you put it up on the roof even after repairing the diodes. If the panels are also faulty you would need to de-solder the suspect diodes and check them separately and then check the current through each string of cells on the panel in both directions. You probably should do this test after removing the diodes for replacement and repair.

[@frankammerman3296]

Life is good!

[@thomasmoore4576]

Thank you so much for doing the video ☺️

[@jws3925]

Andy, this is something I don't believe I have ever seen addressed. I have solar panels up myself and have watched a ton of solar enthusiast videos like yourself and don't recall any of them concerned about or even mentioned the diodes. How prevalent do think this failure is? This was a great "catch" on your part but now we have a number of people, including me, asking themselves if the panels being used have good diodes or not. HA, HA....after reading comments I watched the very end (I clicked off when talked about having a beer). IT IS NOT THE DIODE??? What?

[@muddy11111]

Good that it works with two panels, but connecting the three panel strings was always the problem. You're running test software on the charge controller because the original 150/35 was locking up at 132v every few days. A 3 string can exceed this, 2 will not. I would be looking at the controller and be asking how Victron resolved the lock up issue.

[@zemadeiran]

How about applying power directly to the panels from your lab power supply to test the diodes? You would not need to crack open the diode case If all the diodes are working you should see the full voltage with a minimal drop

[@OffGridGarageAustralia]

I could try that...

[@luc_libv_verhaegen]

Afaict, diode mode on a multimeter just applies a small voltage and then measures the voltage drop. Fundamentally not different from measuring resistance with a multimeter. You should be able to just measure resistance on still pristine panels without opening up the junction boxes and figure out which are faulty. You already created a dataset with the panels you already opened and diagnosed, so you can use those to verify this method and the values gotten.

[@OffGridGarageAustralia]

I'll show that in the next video.

[@USA-GreedyMenOfNoIntegrity]

You remind me of the back to the future movie.

[@OffGridGarageAustralia]

How comes?

[@KC-vp7wn]

Look for hotspots on the solar panel cells. If those cells are shorted you should see higher temperatures in areas on the panel with your IR temperature gun!

[@OffGridGarageAustralia]

Should you still be able to measure the max short current in such a case?

[@KC-vp7wn]

@@OffGridGarageAustralia Anytime you push current back into that panel, you see an increase in heat! It's got to dissipate into heat!

[@guywhoknows]

Without looking at the diode datasheet. I would assume they over heated. With the solar panels heating up and roof heating, it can get mighty hot in a little black box. I would assume the temp max I about 75oC. I'd go for different diode. Over the twin ones as they always get hot. I'd expect if you have an array of them they can't handle the voltage and amps within the string, as well as the heat. Least you know now.

[@peterwalker7869]

I am thinking you could test the panels with a restive load rather than cracking open all the junction boxes. Some light globes connected in series/parallel for the required load. If the panel fails that test then crack open the box. As for ordering and replacing the diodes, do you really need them? I can see why they are there, but if they create more problems than the solve, just chuck em in the can. But if you do replace them I would look for an equivalent, but not the same part.

[@sarathai2876]

Way to go! I feel your happiness.

[@sinusgolf]

Süper😁,semi conductors don't like heat,I think Australian climate is to hot,and it is bad if the panels are too close to the roof

[@gregb1714]

Instead of breaking open all the panel connection boxes, you could just test each panel individually to see if it produces 'normal' power (same as a panel with 3 good diodes). If so, it is good. The panels with 1 or more bad diodes will produce less power.

[@MarcusPocus]

so happy for you and i have learned something new!!! :)

[@mikenewman4078]

Find all the ground faults before committing to refitting the panels. The old one from the house from your description likely has a ground fault.

[@Sylvan_dB]

Instead of opening the junction box, use a 9v battery and measure the voltage when you pass it thru the panel. Reverse the polarity of the battery and measure again. meter minus --- panel minus [panel] panel plus --- 9v minus [9v bat] 9v plus --- meter plus That gives you the ability to measure up to nearly 9v drop across diodes. Panels almost always use 45v diodes (as you discovered) so I would not hesitate to put 2 or 3 9v batteries in series (up to 27v) if I thought 9v wasn't enough to show the max voltage drop across the diodes. You could also dispense with the meter and use an appropriate resistor and LED that will light when connected to the battery correctly. Of course, all this diode testing will be meaningless if light is hitting the panel as you observed.

[@jaminthevanuk296]

Great work son

[@Jim-Bob.]

Lov a good ending !! Can you show what a diode looks like as I fear I have a similar problem !! Yeah !!

[@Nobody_Famous]

7:11 shows the diodes

[@josipkubasa4944]

You not must the open junction box. If you connect in serial resistor (R=U/I) then voltage must be 36V. If you have less then is one string on panel in short. So.. 36/5 is arround 7.2 ohm.

[@kevinroberts781]

I knew you would figure it out. I can't help but think, if you weren't on the metric system you would have found it sooner!😘 We were wrong. Nice

[@DanBurgaud]

MBRB1545CT Spec: Vf ranges from 0.57 ~ 0.84V If pushing 15A through, you get 7.5W~12W of heat. It is enough to cook the diode considering there is no ventilation. This can occur when most of the panels within the string receives good sun (generates a lot of current), and this particular panel was shaded bypassing all the current through the diode instead of the panel.

[@ykchowaa]

O I check the B1545G the DC breakdown voltage is only 45V. and the purpose of the diode is let the power flow in one direction if the cell does not provide positive voltage. so that in case the panel or part of panel under shade. but other panel may be parallel connection it has some power build up in revise order. in that case the diode need to block the voltage. but just 45V... it would break down the PN junction.

[@alanhollister9122]

Well he didn’t show in this video but I heard him between 11.05 and 11:40 !!! Keep up the good work! Fried diodes.

[@jakstak]

Average Joe brought me here, I always watch till the very end, Great content and fault finding, Keep up the good work Andy :)