Is my solar setup faulty?

[Dilbert]

New battery with the new connectors in pic below

Sorry to change the subject slightly, but why do 'professionals' think it is OK not to correctly protect wires? The area highlighted has single insulated wires that can easily chafe with vibration and short on on the metal terminal. There is no extra mechanical protection against vibration to be seen in this installation!

 

[monsteruk]

@mmi / @Phil_G,

Both - thanks for your feedback, the scale is for "not under load conditions" makes sense and allows me to better make sense of the numbers I am seeing. I have a fear that the new battery is discharging too quickly when I turn things on. So, the battery was 100% charged at tea time yesterday (did a 25 mile drive back home from converter which made sure it was topped right up). I left some lights and a phone charger on for 3 hours last night to run the battery down a little, just to see how the solar reacted the today.

Was cloudy till mid-morning, but was seeing between 5W and 15W from panel. Sun come out around mid-day, and saw 52W which seemed not bad, and by 13:30 (see below) it was back up to 100% and in absorption mode. There had been no load on the battery since the previous evening.



Note. I am assuming that 3.9A is what is going into the battery? And the 2.1A from the panel at 26V is resulting in the 3.9A going to battery as its 12v?

So, at 13:30 battery is 100% and thus must be at 95AH capacity I am assuming

At this time I turned on the spotlights (4 individual lights) for 1hr 45mins. I don't know the specific watt rating of those LED spot lights, but most of the ones I see on the web for sale seem to be about 3W each? So, 4x3 = 12 Watts, on a 12v battery, means (in my undedicated mind) an hour running would consume 1AH of the batteries capacity. Lets round up how long the lights were on to 2 hours, so we expect the battery to drop 2AH for the test. So, I assumed that the "open circuit" rating after a a couple of hours of battery rest, should be around 98.5% (i.e 12.9V or about 93.5AH). Is that theoretically correct? Assuming of course, that the solar is not topping anything up (and it did go cloudy again to be fair)!

So, I have just gone back out to the van (no load on battery since lights were turned off after the test 2 hours ago) and this is the current reading.



So, based on that scale AGM scale, I am reading 12.37V as 74% remaining - which I can only interpret as running 4 LED spot lights for 2 hours has used 26% of the batteries capacity, which would be 24AH. So each hour the 4 sport lights were on they used 12AH. Can that be right? I had estimated 2AH used in two hours, and it actually used 24AH in 2 hours! To get to 24amps in a 2 hour period I would assume the LEDs would be drawing 144 watts an hour (i.e. 12amp for an hour at 12 volts would be 144 watts?)

I dont want to get into a dispute with the converter, especially if my figures are crazy or my expectations of what to expect are too high. But if it is wrong, I do of course want him to fix it.

Thoughts? Are my calculations all wrong or do I need to go back to the converter again, or do some other tests?

Thanks in advance...

 

[mmi]

Note. I am assuming that 3.9A is what is going into the battery? And the 2.1A from the panel at 26V is resulting in the 3.9A going to battery as its 12v?

So, at 13:30 battery is 100% and thus must be at 95AH capacity I am assuming

Hmmm, the clock on your screenshot is 13:32 and the battery is still absorbing almost 4 Amps - which means it is definitely not 100%. I would say not even close yet. When the battery is full, it won't accept much current any more - at that voltage.

Please see attached fresh screenshots of mine being connected into charger. The voltage is a bit high as it's -10C outside thus temperature compensation has grabbed it high. Anyways, my point is that current into battery is approx. 0A.


Another one from garage - getting there...

 

[monsteruk]

Thanks @mmi,

First up, -10 degrees is just not natural! ;-) thought was cold here at -2!

So, I think my education is being added to again which is great! I had taken the scale is accurate if battery is "at rest" as being if nothing is being drawn from the battery. I had not considered current going into the battery. So, scale only valid if no input to battery and no drain from battery and battery has rested in that state for a while. So, in my 13:30 screenshot which shows 13v, I cant trust that as 4amps is still going in (so battery not at rest). That makes sense, thank you.

I have done another test this eve (when dark, so not input to battery). I have switched some lights on for an hour and have taken a before reading, and will take an after after an hour or two after the end of the test. I am expecting this test "should be valid" as battery was properly "at rest" before and after the test. I cant publish the figures yet as I need the battery to settle for an hour or two first as only just finished the test.

Be interested in your thoughts tomorrow when I share the numbers if you have time.

Thanks again everyone for indulging me with my noobie questions!

 

[monsteruk]

@mmi / @Phil_G,

I have now done a test based upon the advice you have given, to check the new battery is working as it should.

Table below is a summary, detail and screen shots below. Percentage and AH values obtained by looking up the battery voltage on the AMG state of charge table.

Time	Description	volts	percentage charged	AH remaining
19:48	Before test. Battery in resting state - no charge / drain on battery for 3 hours before this reading.	12.35v	72.5%	68.9AH
19:49	Test - Kitchen LED strip lights turned on for exactly 1 hour. 4ft of LED strip lights under kitchen shelf, 3ft of LED strip lights at floor level (see picture below). Lights turned off at 20:49 - so were on for exactly 1 hour.
22:30	Battery back in resting state - no charge / drain on battery for 1 hour 40 mins before this reading.	12.29v	69%	65.55AH
07:24	Battery left to rest overnight (over 10 hours) with no other use.	12.25v	66%	62.7AH

Conclusions

So, running the LED lights (pic below) for 1 hour depleted the battery by either 3.3AH or 6.1AH depending upon whether you use the post test battery reading from later on the same night or the battery reading from the following morning.

Does that seem to be a reasonable outcome for running those LED strips for 1 hour?

Screen shots

LED strip lighting turned on for the 1 hour test (under shelf and floor level):


Pre test reading:



Post test reading later on eve



Post test reading from following morning


Battery state of charge, lookup table.

 

[t0mb0]

@mmi / @Phil_G,

I have now done a test based upon the advice you have given, to check the new battery is working as it should.

Table below is a summary, detail and screen shots below. Percentage and AH values obtained by looking up the battery voltage on the AMG state of charge table.

Time	Description	volts	percentage charged	AH remaining
19:48	Before test. Battery in resting state - no charge / drain on battery for 3 hours before this reading.	12.35v	72.5%	68.9AH
19:49	Test - Kitchen LED strip lights turned on for exactly 1 hour. 4ft of LED strip lights under kitchen shelf, 3ft of LED strip lights at floor level (see picture below). Lights turned off at 20:49 - so were on for exactly 1 hour.
22:30	Battery back in resting state - no charge / drain on battery for 1 hour 40 mins before this reading.	12.29v	69%	65.55AH
07:24	Battery left to rest overnight (over 10 hours) with no other use.	12.25v	66%	62.7AH

Conclusions

So, running the LED lights (pic below) for 1 hour depleted the battery by either 3.3AH or 6.1AH depending upon whether you use the post test battery reading from later on the same night or the battery reading from the following morning.

Does that seem to be a reasonable outcome for running those LED strips for 1 hour?

Screen shots

LED strip lighting turned on for the 1 hour test (under shelf and floor level):
View attachment 97747

Pre test reading:

View attachment 97741

Post test reading later on eve

View attachment 97742

Post test reading from following morning
View attachment 97743

Battery state of charge, lookup table.

View attachment 97748

I think all you can really read into this is that you ran your lights for an hour and the resting voltage dropped by 0.06v, which seems pretty reasonable. I really wouldn't read much into that resting voltage versus SoC table you've found, the general idea is right but I wouldn't trust those specific values for any given battery - if you search around the net you can find equivalent tables which are +/-0.3v at any given SoC. I have an AGM battery and my resting voltage at 100% SoC is definitely lower than 13v, and I'm pretty sure the resting voltage at 50% SoC is a fair chunk higher than 12.05v. I might measure it though if I've got time.

 

[t0mb0]

I'm not quite sure what issue you're trying to diagnose, but if you're trying to work out whether there is some unexpected current draw on your battery, you're really going to have to measure that directly rather than try and back it out from resting voltages and SoC tables. You have two options here:

1) Fit an actual shunt-based SoC monitor to the battery. This might be useful, but can be somewhat expensive - lots of people on the forum use victron bmv devices.
2) Just buy a £30 clamp meter which will enable you to instantaneously measure the current flowing out of the battery when you turn your lights on.

 

[monsteruk]

I'm not quite sure what issue you're trying to diagnose, but if you're trying to work out whether there is some unexpected current draw on your battery, you're really going to have to measure that directly rather than try and back it out from resting voltages and SoC tables. You have two options here:

1) Fit an actual shunt-based SoC monitor to the battery. This might be useful, but can be somewhat expensive - lots of people on the forum use victron bmv devices.
2) Just buy a £30 clamp meter which will enable you to instantaneously measure the current flowing out of the battery when you turn your lights on.

Hi @t0mb0,

thanks for taking the time to reply, very useful. Long story short on what I am try to do. Basically, I have had no end of issues with the battery / solar setup since the conversion was done back in March 2020. Its been widely acknowledged on earlier pages on this thread that it was not working correctly. The camper went back to the converter (for the second time) on 4th Jan 2021 - this time for him to fit a new panel and MPPT. However, he said he has managed to fix it all without changing the panel or the MPPT. What he said he has done is this:
1. Replaced the existing 110AH lead acid battery (which he said he had tested and was faulty), with an AGM 95AH battery
2. Replaced a cable going to the battery as he was not happy with the connector
3. Used better "bolt on" battery terminal connectors.

So, I am trying to check myself that the setup is now actually fixed, and working as it should. The pressure is on a little to test its all ok, as come Feb this year the panel and MPPT will be out of warranty. So, I want to prove to my own satisfaction that it is all working as it should before the warranties expire.

I am all ears about how I should do this. But my initial plan was to
1. test the battery independently first (forgetting about the solar), and make sure that is functioning correctly (tests above were aimed at this).
2. test that the solar panel is generating the power one would expect (for a 150W panel) in the current conditions, and charging the battery at an appropriate rate.

If I can prove the battery and solar are working ok before Feb, then I am done and I can forget about it. If I detect there is still an issue however, I need to find out quickly and get back to the converter whilst the parts are still under warranty.

Does that make sense? So I may have been going about the tests the wrong way, but thats what I am trying to achieve / prove.

On a different note. I like the idea of having something that shows the draw on the battery, I will look at your suggestions above. Something that showed it in the Victron app would be amazing. Currently I can see current going in, but cant see anything going out. Ideally, I would see something on the victron app when I turned lights on etc. What would be totally amazing, would be to see the "net amps position", i.e. if 3amps were going in from solar, but 2 were being drawn, it shows 1A charge. Not really sure whats possible. If achieving that is very expensive, then some form of clamp gauge so I can at least see load for testing would be great.

But my very short term need is to test that the battery and solar setup "post his fixes" are actually working as they should.

Many thanks.

 

[oldiebut goodie]

I can't see that the battery has been fully charged yet - unless I missed something? I would start with getting the battery fully charged with a charger then letting it rest for 24hrs before doing your tests. 12.37v is the highest voltage at rest that I can see above.

 

[t0mb0]

Hi @t0mb0,

thanks for taking the time to reply, very useful. Long story short on what I am try to do. Basically, I have had no end of issues with the battery / solar setup since the conversion was done back in March 2020. Its been widely acknowledged on earlier pages on this thread that it was not working correctly. The camper went back to the converter (for the second time) on 4th Jan 2021 - this time for him to fit a new panel and MPPT. However, he said he has managed to fix it all without changing the panel or the MPPT. What he said he has done is this:
1. Replaced the existing 110AH lead acid battery (which he said he had tested and was faulty), with an AGM 95AH battery
2. Replaced a cable going to the battery as he was not happy with the connector
3. Used better "bolt on" battery terminal connectors.

So, I am trying to check myself that the setup is now actually fixed, and working as it should. The pressure is on a little to test its all ok, as come Feb this year the panel and MPPT will be out of warranty. So, I want to prove to my own satisfaction that it is all working as it should before the warranties expire.

I am all ears about how I should do this. But my initial plan was to
1. test the battery independently first (forgetting about the solar), and make sure that is functioning correctly (tests above were aimed at this).
2. test that the solar panel is generating the power one would expect (for a 150W panel) in the current conditions, and charging the battery at an appropriate rate.

If I can prove the battery and solar are working ok before Feb, then I am done and I can forget about it. If I detect there is still an issue however, I need to find out quickly and get back to the converter whilst the parts are still under warranty.

Does that make sense? So I may have been going about the tests the wrong way, but thats what I am trying to achieve / prove.

On a different note. I like the idea of having something that shows the draw on the battery, I will look at your suggestions above. Something that showed it in the Victron app would be amazing. Currently I can see current going in, but cant see anything going out. Ideally, I would see something on the victron app when I turned lights on etc. What would be totally amazing, would be to see the "net amps position", i.e. if 3amps were going in from solar, but 2 were being drawn, it shows 1A charge. Not really sure whats possible. If achieving that is very expensive, then some form of clamp gauge so I can at least see load for testing would be great.

But my very short term need is to test that the battery and solar setup "post his fixes" are actually working as they should.

Many thanks.

If you want SoC monitoring with amps in/out to match your solar panel diagnostics, then you're really talking about a victron bmv device - you can get a version without the hardware screen, it just interfaces with the victron phone app.

In terms of diagnosing battery issues, I think giving it a decent workout and ensuring the fridge, lights etc all work as they should whilst keeping an eye on the voltages should be fine. Assuming they've redone the faulty connections, it really should be fine. You could use a multimeter to explicitly measure voltages at the battery and downstream at the lights/fridge etc if you really want to check for any undue voltage drop over the wiring.

In terms of testing the solar panel, you really need two things:

1) The battery discharged enough to accept whatever the solar panel should provide in terms of current.
2) A sunny day!

Obviously (2) is somewhat tricky in Feb but keep at eye on the forecast and discharge the battery a bit in advance of a decent forecast and you should get something.

As an aside, I really would charge your battery up to full at some point, I can't see any evidence in the thread that the battery has ever been up to 100% charge.

 

[monsteruk]

All,

I may have found something significant. I went out at lunchtime to see how charging was going. Plenty of sun about, similar to yesterday lunch when I saw 52W from the panel. I actually saw 38W, a little lower than yesterday but still a decent number.

On a visual inspection of the roof / panel it seemed a bit dirty (not washed it for a couple of months), so I got my bucket, warm water and soft sponge and gave it a gentle clean. That's all I did, never even opened the van. I expected to see the number go up a little from 38W as the panel was nice and clean.

Surprisingly, I saw the polar opposite. As soon as I washed the roof and panel, the number dropped immediately to 1W and a few seconds later to 0W.

So, slightly grubby panel pulling 35W, clean the panel, now pulling 1W and 0W seconds later.

So, I got my ladders to have a look at panel, specifically the area where the panel wires are... I just cant think of anything else that could be impacted by a gentle wash of the panel? And what I found was this



Where the cable goes into the panel, the seal is gone - so I assume water would just run into that? Would that cause the issue of it dropping to 0W straight after the clean, as water has got in?

Just for some context, the image below shows where this cable enters the panel:


For reference, here are the screen shots before and after wash....

Before wash:


After wash (just 5 mins after screen shot above):



Very shortly after, it dropped to zero, and has not come back up from zero since.



So, am I on to something? It dropped from 35W to 0W when the roof got wet. That seal on the cable into the solar is perished / not sealed, so I suspect water has got in. Would that cause this issue?

Would really appreciate any quick views, as this is driving me insane at this point! lol

Thanks all.

 

[t0mb0]

All,

I may have found something significant. I went out at lunchtime to see how charging was going. Plenty of sun about, similar to yesterday lunch when I saw 52W from the panel. I actually saw 38W, a little lower than yesterday but still a decent number.

On a visual inspection of the roof / panel it seemed a bit dirty (not washed it for a couple of months), so I got my bucket, warm water and soft sponge and gave it a gentle clean. That's all I did, never even opened the van. I expected to see the number go up a little from 38W as the panel was nice and clean.

Surprisingly, I saw the polar opposite. As soon as I washed the roof and panel, the number dropped immediately to 1W and a few seconds later to 0W.

So, slightly grubby panel pulling 35W, clean the panel, now pulling 1W and 0W seconds later.

So, I got my ladders to have a look at panel, specifically the area where the panel wires are... I just cant think of anything else that could be impacted by a gentle wash of the panel? And what I found was this

View attachment 97763

Where the cable goes into the panel, the seal is gone - so I assume water would just run into that? Would that cause the issue of it dropping to 0W straight after the clean?

Just for some context, the image below shows where this cable enters the panel:

View attachment 97764
For reference, here are the screen shots before and after wash....

Before wash:
View attachment 97765

After wash (just 5 mins after screen shot above):

View attachment 97767

Very shortly after, it dropped to zero, and has not come back up from zero since.

View attachment 97768

So, am I on to something? It dropped from 35W to 0W when the roof got wet. That seal on the cable into the solar is perished / not sealed, so I suspect water has got in. Would that cause this issue?

Would really appreciate any quick views, as this is driving me insane at this point! lol

Thanks all.

View attachment 97766

That does look like a potential problem, seems odd that it dropped off a cliff after cleaning - although I assume that it wasn't simply that the panel was covered in water?

The perished sealant certainly isn't ideal, as you don't want water getting in but, unless there are some actual bare wires in there somewhere, it getting wet shouldn't actually cause an issue. If there's some cable joint in there though, that isn't waterproof then that certainly could be an issue.

The other thing is that I guess there could be some dodgy connection either in the panel, or in the connections and by washing it you either jogged it or (if you used warm water) raised the temperature to a point where the connection was broken by expansion. I seem to remember your original issue occurred more in actual sunshine, is that right? If so, the heating leading to a bad joint could explain both.

I'd wait for the panel to sit a while to see if it "comes back" and then try and trigger it again. Lightly jog the connection box and cabling to see if you can induce the fault. Then chuck cold water over it, followed by warm to see if that induces the fault.

 

[monsteruk]

That does look like a potential problem, seems odd that it dropped off a cliff after cleaning - although I assume that it wasn't simply that the panel was covered in water?

The perished sealant certainly isn't ideal, as you don't want water getting in but, unless there are some actual bare wires in there somewhere, it getting wet shouldn't actually cause an issue. If there's some cable joint in there though, that isn't waterproof then that certainly could be an issue.

The other thing is that I guess there could be some dodgy connection either in the panel, or in the connections and by washing it you either jogged it or (if you used warm water) raised the temperature to a point where the connection was broken by expansion. I seem to remember your original issue occurred more in actual sunshine, is that right? If so, the heating leading to a bad joint could explain both.

I'd wait for the panel to sit a while to see if it "comes back" and then try and trigger it again. Lightly jog the connection box and cabling to see if you can induce the fault. Then chuck cold water over it, followed by warm to see if that induces the fault.

Thanks @t0mb0,

So, the panel went to zero and it was wet. I then got a soft, micro-fibre towel and dried it off. No difference, was still 0W.

Based on you comment, I have just been out again, and cloudier now, but it has "come back to life", showing 5W.



I have tried tapping the junction box / wires and this did not induce the issue. I will now go and try the cold / warm water.

Quick question... The wire where the seal has perished... Is that wire doing into the solar panel itself? Or would that wire be the one that went down into the van and eventually on to the MPPT?

Will report back on water test!

Thanks

 

[t0mb0]

Thanks @t0mb0,

So, the panel went to zero and it was wet. I then got a soft, micro-fibre towel and dried it off. No difference, was still 0W.

Based on you comment, I have just been out again, and cloudier now, but it has "come back to life", showing 5W.

View attachment 97769

I have tried tapping the junction box / wires and this did not induce the issue. I will now go and try the cold / warm water.

Quick question... The wire where the seal has perished... Is that wire doing into the solar panel itself? Or would that wire be the one that went down into the van and eventually on to the MPPT?

Will report back on water test!

Thanks

The seal you've highlighted is the one going into the van, this would only cause an electrical issue if the water was going down the cable as far as a non-waterproof joint. A dodgy seal in the junction box would definitely cause an issue too - chuck water over this to test. These could be the case but my hunch would be a dodgy connection in the panel itself which breaks connection when heated. Keep an eye on the actual voltage as you heat the panel, in that last screenshot it jumped back up to 22.68v from the 14.43v which is presumably close to the correct operating voltage for the panel. If warm water does trigger it, then it's time to get the hair dryer out and work out which bit of the panel has the issue....!

 

[t0mb0]

Oh hang on, looking at your pictures, they've drilled the holes to get the solar panel output wires through the solar panel itself!!

They obviously haven't drilled the cells themselves, but who knows how close those holes are to the internal connections of the cells to the junction box. Even if not that close, the vibration of drilling could have caused an issue. If it comes down to a dodgy connection, I'd apply a hair dryer to the area around those holes and see if that triggers it.

 

[oldiebut goodie]

As this is a new conversion why is there a failed seal already? It looks like the cowboys have been at work again! Back to them to sort out methinks, you shouldn't be having to do their work again straight away like this.

 

[t0mb0]

As this is a new conversion why is there a failed seal already? It looks like the cowboys have been at work again! Back to them to sort out methinks, you shouldn't be having to do their work again straight away like this.

That might be true in theory, but in practice it's a good idea to have a clear idea what's wrong before getting them to fix it so you don't have to keep going back!

 

[oldiebut goodie]

They have made so many bodges already and are supposed to have sorted this that I can see no alternative as there will most probably be many more problems to come. I wouldn't be happy with what they have done so far that has come to light.

 

[monsteruk]

Oh hang on, looking at your pictures, they've drilled the holes to get the solar panel output wires through the solar panel itself!!

They obviously haven't drilled the cells themselves, but who knows how close those holes are to the internal connections of the cells to the junction box. Even if not that close, the vibration of drilling could have caused an issue. If it comes down to a dodgy connection, I'd apply a hair dryer to the area around those holes and see if that triggers it.

Thanks @t0mb0 / @oldiebut goodie

I have just tried to recreate the issue (warm, not hot water didnt seem to do it), I then just did what I did earlier which was wash it again, rinse it off and dry it.

By the time I did that, most of the sun had gone, but it did show 1W and a better voltage. This is what it was when I finished, it maintained 1W, but bounced between 19v and 22v (screen shots below). So, I am assuming the 1W is due to the sun getting low, as opposed to me successfully reproducing the issue.

I have run out of time for additional tests now as the sun is going down.

Thinking about it you know, when I first noticed the drop to 1 and zero, I am not sure I had applied the warm water at that point, I think I had given it a very light spray with the hose (cold water) and then wiped it down with a micro fibre towel) - before I then cleaned properly with warm water and soft sponge - so I may have misled about the warm water.

So, assuming the sun comes up tomorrow and its another decent bright day, am I best to try and do what I did today again? Or try the hair-dryer near the cable entry points? I guess I could try hair-dryer first and then do what I did today if that doesnt work? I would like to reproduce it a second time if I can, as it would make the conversation with the converter easier!

Screen shots after attempting (but I think failing) to reproduce the issue for the second time.





Thanks

 

[Phil_G]

I think @t0mb0 might be right with the idea of heat or movement disturbing a bad joint or a fracture. Water won't cause an electrical problem between the panel & controller, its not that good a conductor at these low voltages, but it will cause corrosion/oxidisation of any joints it gets into, and would play havoc if it got into the controller itself. But electrically, other than the corrosion/oxidisation of joints, it wont directly affect your solar, it certainly wont 'short' the panel or battery.
It is a concern that they've drilled the panel - quite apart from possible damage to the internal interconnects, it breaks the layer seals and I would imagine would void the manufacturers warranty. You'd have made a better job fitting it yourself

As an aside, nothing to do with your problem, it is possible for a clean panel to produce less than a dusty one, especially at acute angles. The dirt can scatter the light so it impacts the cells at a more efficient angle, and also tends to reduce reflection - some panels have an intentionally coarse finish for this reason. A highly-polished panel surface could reflect almost all the energy from a low sun.