Solar panels under load drop to 37V from 66V

[Any name you wish]

I have a small 4 panel string that Voc is 66V. When a load happens (charging batteries, discharging to car) the panels voltage on the SCC drops to 37V. The other SCC with similar panels only drops down to 55V which is closer to what it is supposed to be. I thought it was connections and changed them out and it still doesn't seem to change things. Once the batteries are full, it idles at 66V with no issues. Wiring is great. No voltage drops back to the breakers and inputs to the SCC. Suggestions?

 

[SpongeboB Sinewave]

Does this 4 panel string have ANY shade on the panels at all ? That would certainly make the maximum power point lower.

 

[Hedges]

Could be one bad panel. A large panel with diode-bypassed sections can behave similarly.
Try removing each panel from the string and testing with the remaining three. You may find one is bad. If so try replacing it with one panel from other (good) string.

It is possible for Voc and Isc to both look good, but Vmp/Imp is low.

What are specs of individual panel?

 

[Any name you wish]

Does this 4 panel string have ANY shade on the panels at all ? That would certainly make the maximum power point lower.

No. It is a 2S2P setup as well FYI.

Could be one bad panel. A large panel with diode-bypassed sections can behave similarly.
Try removing each panel from the string and testing with the remaining three. You may find one is bad. If so try replacing it with one panel from other (good) string.

It is possible for Voc and Isc to both look good, but Vmp/Imp is low.

What are specs of individual panel?

These are BenQ panels. I need to find my picture of the data sheet.

 

[sunshine_eggo]

I have a small 4 panel string that Voc is 66V.

A string consists of panels in series. You should have simply referred to it as an array or stated the config, e.g., 2S2P.

No. It is a 2S2P setup as well FYI.

Since it's 2S2P, you can separate the two strings, try each 2S string individually.

These are BenQ panels. I need to find my picture of the data sheet.

The label on the back of the panel should have all pertinent information as well as the model number, which should allow you to look up the datasheet if you need more info.

 

[Any name you wish]

Sun is setting now so I can try the separate string thing tomorrow.

 

[Hedges]

I think you are seeing voltage 25% lower than expected when under load.
With 2s2p, possibly the panels have 2 diodes, and cells on one half are damaged, not able to deliver much current.

Well, data sheet says 3 diodes.

https://www.solaris-shop.com/content/PM250M00-260%20Specs.pdf

Test with 2 panels 2s, as Sunshine suggests. Then the other 2s. If different, try swapping panels to find the weak one.

 

[Any name you wish]

Alright, here is what I am seeing. Started charging my car again. fullish sunlight. Batteries started at full. Here is the solar voltage trend.

Going out to seperate the 2S2P into 2S now.

 

[Any name you wish]

Here is the first set. The drop to zero was breaker off to make the change and reset the MPPT to track again.

 

[Any name you wish]

And here is the second set. Me thinks we have a winner.

So can I use a diode tester to get some readings off of these to see which is the issue? The second voltage drop was me flipping the solar breaker a second time just to make sure the MPPT got the hint.

 

[Hedges]

Swap which two panels are used in a string. You should be able to narrow it down to 1 bad panel.

I'm thinking cells bad, might not be something a diode tester can show.
I used a resistor (couple of electric radiators) to load my 165W panels to near Imp/Vmp, measured voltage & current.

 

[Any name you wish]

I did voltage measurements on the 2S that was low. One measured 35Voc, the other 29-31Voc. I checked Isc on both and they were both around 3A. I have 6 spares available. I think I will swap the lower voltage one out for another. Will verify the voltage on it before beginning to see how this goes..

 

[Hedges]

In that case, just Voc of all panels should identify odd man out.
I had 5 bad panels out of 18. 3 of those showed up by Voc or Isc. The other 2 by V(load) or I(load).

 

[chrisski]

I'm thinking cells bad, might not be something a diode tester can show.
I used a resistor (couple of electric radiators) to load my 165W panels to near Imp/Vmp, measured voltage & current.

I've heard you say this before and I think its a great idea and would like to someday see or get a link to a radiator like you mentioned.

I have some ideas on how to do this, but I'd probably go through a few things trying to figure it out on my own.

 

[Hedges]

When used for DC, turn thermostat up all the way and leave switched on. The controls are made for AC, might arc and burn with DC.
The ones I've got have elements for 600W and 900W at 120Vrms (same for VDC).
I plugged a couple into a power strip, checked resistance while switching the elements, selected 8 ohms because at 36Vmp, that gave about Imp and panel wattage.

I used suitably rated knife switches to open/close circuit as I plugged in different panels, giving Voc and Imp/Vmp (or load, not exactly maximum power point). And to short out the panel for Isc measurements.



I actually have one under my bench at work right now, to draw current from a power supply or Variac to generate magnetic fields for test measurements.

They are low temperature, especially at reduced voltage, so nicer than having 100C+ resistors burning things. I've also had resistors operating at 400VDC; those I put in a box to avoid killing my intern (or me).

 

[efficientPV]

Don't forget the third option for even lower wattage, the two heater elements in series. I have one of these I used as a dump heater with solar. At 400W it wasn't touch safe. I think they can only take continuous full power if the room is very cold.

 

[Hedges]

(for anyone who might try without doing electrical analysis first ...)

If operating above rated (120V) voltage, for series configuration make both same wattage setting, or calculate voltage division.
And certainly be sure switch and thermostat can't open. They might work at reduced DC voltage, but likely to pull continuous arc at high DC voltage.

You can blow a fan over the radiator to carry away much more power while keeping temperature low.
Thermostat I think measures air. It may have an over-temperature switch in the oil, not sure, but if it opened it would arc too.

 

[Any name you wish]

Finally got around to swapping panels today and success!



Not too bad on my 30A controller. Glad to be back up to full production.