Solar problems

[Wh05apk]

For a long time I have been unhappy with the output of my panels, factory fitted is 120w, and the dealers added 2x150w zamp panels and a zamp pwm controller.
Recently I tried covering each panel and disconnecting, one of the panels is proving no power at all, although testing is proving 24v, so seems alive but why no power?
I have replaced the controller with a victron 100/30 unit, which is ahowing the output at 14-15amps, but the vans battery meter only shows about 8-9amps (I ran the battery down to about 70%) seems like about 40%loss? Initially I had assumed it was going through the vans charger which was limiting the charge, but it is connected direct to the battery.

Finally the attached photo shows how the panel is fixed to the fixings any idea what these are called so I can get a bit, rather than drill out, if I need to remove to access the underside?

 

[autorouter]

have replaced the controller with a victron 100/30 unit, which is ahowing the output at 14-15amps, but the vans battery meter only shows about 8-9amps (I ran the battery down to about 70%) seems like about 40%loss? Initially I had assumed it was going through the vans charger which was limiting the charge, but it is connected direct to the battery.

If the 100/30 output is in fact 14-15A, and the van battery meter is showing 8-9A, then the difference must be going somewhere. Is there a device still connected that might be taking 6A or so? The other possibility is the meter is incorrect, but that's a massive error if it is. More likely there's some pathway for the amps from the solar controller or battery that bypasses the van battery meter.

I think those screws might be 'Tri-wing' type, there are different sizes.

TRI WING SECURITY BITS SIZES 1 2 3 & 4 25MM 1/4"HEX CHROME VANADIUM BITS | eBay

Find many great new & used options and get the best deals for TRI WING SECURITY BITS SIZES 1 2 3 & 4 25MM 1/4"HEX CHROME VANADIUM BITS at the best online prices at eBay! Free delivery for many products.

www.ebay.co.uk

The panel problem might simply be a bad connection in the wiring somewhere. Can you try the panels one at a time, to see if you get an output?

 

[Wh05apk]

All very quiet here?

If I was looking to replace the non working panel, which is connected in parallel, does it need to be the same make/power?

 

[Wh05apk]

Thanks autoroute, I have disconnected one panel at a time and one panel makes no difference despite still showing a voltage.

At night with no solar and everything off the drain is only 1/2a or so, for some reason it won't seem to put in more than 9-10a even when showing 15+ a and discharged battery.

Thanks those tri wings look good, will see if I can get some to remove the panel and check the connections.

 

[donkey]

If you get a piece of blu-tac or similar you could press it into the screw head and make a mould of the driver required. Then take it 'carefully' along to your local tool shop as a pattern.

Geoff

^{_{Subscribers  do not see these advertisements}}

 

[Raul]

What’s the data of the panels? And when you say that no difference in amps when you disconnect one, what volts at that time? Does the voltage agh Ange even 1-2v?
You can replace with other make but similar spec, in parallel needs to have same amps or higher, and volts within 5%.

 

[autorouter]

The voltage more or less depends on the number of cells in the panel. Typically 9 x 4 = 36 cells, or 10 x 6 = 60 for the larger panels. So look for one with the same number of cells as a start. The amps depends on the area, so bigger cells means more amps. You'll probably find that a panel with the same number of cells, of about the same area, will match OK.

 

[Wh05apk]

What’s the data of the panels? And when you say that no difference in amps when you disconnect one, what volts at that time? Does the voltage agh Ange even 1-2v?
The voltage was remaining constant, I can't understand how if there is a voltage, surely the connections are OK. So there should be some current as well?

 

[Raul]

The reason I asked about the voltage is: when you have two matched sources ( two solar panels) the mppt locks on a certain power curve with a certain voltage for that time. When one of the sources disappears (one panel off), the power curve changes so the mppt sweeps the array again for new values. When it sweeps and samples best point, the voltage changes 2-3times then locks on a diferent voltage you had before. The voltage change when you power off one of the parallel panel, it’s a sign that both panels work/ contribute. If no change in voltage, then is very likely that one panel does not contribute.

 

[autorouter]

Thanks those tri wings look good, will see if I can get some to remove the panel and check the connections.

The ones I use came as part of a set of 'security bits', a bit like this one.
Amazon product ASIN B01GZWTHFK

^{_{Subscribers  do not see these advertisements}}

 

[Wh05apk]

The reason I asked about the voltage is: when you have two matched sources ( two solar panels) the mppt locks on a certain power curve with a certain voltage for that time. When one of the sources disappears (one panel off), the power curve changes so the mppt sweeps the array again for new values. When it sweeps and samples best point, the voltage changes 2-3times then locks on a diferent voltage you had before. The voltage change when you power off one of the parallel panel, it’s a sign that both panels work/ contribute. If no change in voltage, then is very likely that one panel does not contribute.

Previously I was using the zamp pwm controller and the voltage made no difference when I disconnected a panel. The output doesn't change when I cover the suspect panel, but does if I cover the other parallel panel.

 

[Wh05apk]

The ones I use came as part of a set of 'security bits', a bit like this one.
Amazon product ASIN B01GZWTHFK

Thanks have ordered some similar.

 

[donkey]

I thought the way to test a solar panel was to check open circuit volts and close circuit amps, that is with the panel disconnected from vehicle and in decent sunshine.

Geoff

 

[Wh05apk]

I thought the way to test a solar panel was to check open circuit volts and close circuit amps, that is with the panel disconnected from vehicle and in decent sunshine.

Geoff

Now you've lost me!

All I know is that if the panel is disconnected or covered, and nothing changes then something is wrong, if the volts are there but it is providing no additional current then is the panel dead? or is there something faulty with the connections etc?

 

[autorouter]

With the panel disconnected from anything, measure the voltage (Voc) when the sun is shining on it. Should be quite high, maybe 28V, depending on how many cells it has.

The other test is to measure the 'short circuit current' (Isc) in amps. Set the meter to amps. You may need to move one of the meter probe wires to a different socket on the meter, you are expecting several amps. Connect the meter across the panel wires to see what amps flows.

Note that it is a REALLY bad idea to do this short circuit test on anything other than a solar panel. If you do this test on a battery for example, you will instantly blow the meter fuse if you are lucky, and cause all kinds of damage if you are not.

^{_{Subscribers  do not see these advertisements}}

 

[Wh05apk]

With the panel disconnected from anything, measure the voltage (Voc) when the sun is shining on it. Should be quite high, maybe 28V, depending on how many cells it has.

The other test is to measure the 'short circuit current' (Isc) in amps. Set the meter to amps. You may need to move one of the meter probe wires to a different socket on the meter, you are expecting several amps. Connect the meter across the panel wires to see what amps flows.

Note that it is a REALLY bad idea to do this short circuit test on anything other than a solar panel. If you do this test on a battery for example, you will instantly blow the meter fuse if you are lucky, and cause all kinds of damage if you are not.

Will give that a go, the voltage was about 27v, I think we did try but no amps, will try again.

 

[Wh05apk]

UPDATE:

At the weekend, we tackled the panel, I obtained some tri-wing bits but they didnt work, in the end we just drilled them out as they appeared to be rivets?
Once we removed the panel to access the rear, we took the cover off the connections and on the live terminal the solder looked dodgy so re-soldered it, and as if by magic we have now achieved 401w max so far today from 440w of panels which previously was 253w from 280w, so the panel is re-fitted and working, and currently charging nicely.

However.....we still seem to lose about 40% of the output from charger to the display in the van:

Difficult to tell from here, but this is the layout:

Looks a bit of a mess, that will be my next job, this is how the dealers left it...

Any explanation why we are losing 40% between the app and the van? the 40% seems constant even at lower states of charge, and lower outputs, at a loss where it is going, if it was a constant 2/3 a I could understand there is a drain somewhere, but this just appears to be about 40% regardless.

 

[Raul]

Meter is wired wrong, or somehow is missing some load/charging that bypasses it before it makes it to the battery.
Also your solar charger to battery, the cable is undersized, hence to voltage difference. It should be 10mm2 at minimum, for a very short distance, or 16mm if longer than 1,5m.

 

[autorouter]

You can't just 'lose' about 10A of current. It's not like voltage drop that just disappears along the wire. There must be a pathway that is not being picked up by the display meter. How does the display measure the amps?

It's unlikely that the whole current passes through the display box, so it must measure the current either inside the distribution/fusebox or at a separate point, most likely near the leisure battery negative terminal.

Wire thickness will make no difference. If the wire is too thin, then the 10A will still be flowing, but the wire will heat up.

 

[Wh05apk]

Meter is wired wrong, or somehow is missing some load/charging that bypasses it before it makes it to the battery.
Also your solar charger to battery, the cable is undersized, hence to voltage difference. It should be 10mm2 at minimum, for a very short distance, or 16mm if longer than 1,5m.

Thanks for your quick reply, the cable was as supplied by Zamp for the panels (admittedly 2x160w) but only just fits in the Victron terminal, so will look to upgrade when I can. It just seems weird how the difference in output/charging is always 40% (presumably a short run of undersized cable wouldnt account for this?) at least now I have some decent power going in, I will ignore the meter and see how we get on for now.

^{_{Subscribers  do not see these advertisements}}

 

[Wh05apk]

It does seem to be a big drop, about 120w, i've no idea where/how the meter is wired it is the factory fit one, here is a pic of the negative terminal of the battery which I assume is some form of shunt, where presumably it is picked up from:

 

[Raul]

The drop I refer to, is the voltage from 13,87 to 13,6v. The power 120w You refer to, unaccounted for, it’s probably bypassing the meter. Something is wired wrong( after meter and not counted), or meter decalibrated, or a hidden load, or, or, etc.

 

[Wh05apk]

The drop I refer to, is the voltage from 13,87 to 13,6v. The power 120w You refer to, unaccounted for, it’s probably bypassing the meter. Something is wired wrong( after meter and not counted), or meter decalibrated, or a hidden load, or, or, etc.

That all makes sense, but the battery % increases roughly in line with the charge on the gauge not the app (I appreciate the % is only an estimate) 5a on the gauge roughly putts on about 2% an hour, whereas if the app was correct at about 8a I would expect about 3% an hour?

 

[68c]

Most battery load meter shunts are fitted to the battery negative terminal. All other negative connections are to the other end of the shunt. In that way all current flowing into or out of the battery can be measured. If a circuit return line has been connected directly to the battery negative terminal it will bi-pass the shunt and not be read. Perhaps a device to top up the vehicle battery, burglar alarm etc has been fitted incorrectly. Another problem can be the shunt itself, many are a metal strip with a thin saw cut to set it's resistance, this slit can get filled with dirt and affect the resistance and give a low current reading, similarly poor connections can give a false high reading. Better designs have a built in circuit that reads the voltage drop then output as a digital signal.

 

[autorouter]

The MPPT is telling you its amps output, and the meter shunt is measuring the amps going in/out of the battery. The difference between them will be the habitation loads, which will use some of the amps instead of it going into the battery. Is there some load taking power that isn't switched off?

^{_{Subscribers  do not see these advertisements}}

 

[Raul]

We did ask, and even if is, that’s quite a considerable 12v load at 120w. Even two compressor fridges on ours is less than that. It may be the meter. The mppt data is very accurate

 

[Wh05apk]

At the moment, I am letting things settle and see how things work, the readout in the van is a constant 60% of the Victron app, with the solar switched off the draw is about 0.3a, so I can only assume somehow the van display is reading wrong, the important things is that I now have 440w of solar no functioning properly, currently producing up to 2 KW a day! enough to run the fridge and aircon for a couple of hours and still have the batteries charged by evening.