What am I doing wrong? Bad solar charge controller or bad panels?

[iwasforcedtopickausername]

I am absolutely perplexed.

2x 250W Santan solar panels in series feeding a PWM Tristar TS-45 (with screen module), which in turn feeds a SOK 48V 100AH server rack battery.

This setup was working really well, until some weeks/months ago, when I saw a high voltage disconnect on the Tristar. With a VOC of 36V per panel, I don't see how I could be hitting that, especially with the battery not nearly fully charged.

After that first high voltage disconnect, I started getting FET faults. I called Morningstar, got a very nice voicemail response, tried playing phone tag once, and just haven't found the time to call them back yet.

Here's where I'm really confused. I bought the TS-45 used and on rare occasion the FETs go out. I've found maybe two people on the internet claiming to have this happen. I assume it does happen, just rarely. So maybe I had bad luck?

Anyway, thinking about how PWM works, at least about how I think it works, it just shorts the panels to the battery and only goes into PWM mode when high voltage is reached. (Am I wrong about this?)

So, I wired the panels straight to the battery last night.

This morning, lots of direct sun. I check with the clamp meter and I'm getting no current out of the solar panels. If I turn off the breaker, I see a higher voltage than when I turn it on. My 10 amp fuse is happy.

Could something in the panel be bad, or am I being dumb?

I'd appreciate any help. Thank you.

 

[Supervstech]

Sounds like a bad controller to me.

 

[iwasforcedtopickausername]

Sounds like a bad controller to me.

I was thinkning that at first as well. Is my method of bypassing the controller sound? I'm not seeing any current with my clamp meter with the panels wired directly to the battery.

 

[sunshine_eggo]

2X 60 cell panels in series isn't optimal for 48V systems. They can be ineffective when the cells get hot.

I would check Voc and Isc of each panel. Not getting any current to the battery when direct connected sounds like a failed panel or bad connection.

 

[HarryN]

Perhaps try to measure the panel Voc and see if it is roughly matching what you expect?

Since you saw a voltage spike, I guess it is useful to see if the BMS has tripped.

 

[iwasforcedtopickausername]

Cells aren't getting very hot this time of year.

Panel specs are here: https://www.santansolar.com/product/santan-solar-t-series-250w/

37.6V open
30.3V max power

I'm seeing 66.6V right now with it disconnected.

I did test with the amp meter yesterday evening and was getting something like 0.8A shorted with no direct sun on the panel.

Haven't seen any such issues on the BMS.

 

[wpns]

Voltage looks OK. You should be getting 8-9 amps short circuit in full sun, if not then there’s something wrong with the panels.

Is it possible your battery is in overvoltage shutdown?

 

[sunshine_eggo]

Cells aren't getting very hot this time of year.

Leave them in full sun for 30 minutes and go put your hand on them.

Panel specs are here: https://www.santansolar.com/product/santan-solar-t-series-250w/

Used panels are a crapshoot.

37.6V open
30.3V max power

I'm seeing 66.6V right now with it disconnected.

This alone demonstrates that temperature is a factor, and the cells are warm. Panels are rated for 25°C. That's not ambient. When ambient is 25°C, it's easy to see the cells at 40-50°C, which results in about an 8% drop in voltage.


37.2 * 2 = 75.2V
(75.2-66.6)/75.2 = 11%

I did test with the amp meter yesterday evening and was getting something like 0.8A shorted with no direct sun on the panel.

Needs to be in full sun at high noon.

Haven't seen any such issues on the BMS.

Have you been looking for them? Does your battery give an obvious indication when you're in OVP?

 

[Supervstech]

With the two panels in series, connected to battery that is low charge, you should be pulling around 8A not .8
Full sun, aimed directly at the sun.
No shadows on any spot on the panel.

 

[iwasforcedtopickausername]

The amp test was yesterday evening with very little sun -- I was just testing for curiousity's sake. I haven't done it again today.

I put my hand on the panel and indeed it's not hot. Maybe 30C or so.

Right now with sun on the panels, I get 6 amps if I short them out.

 

[sunshine_eggo]

The amp test was yesterday evening with very little sun -- I was just testing for curiousity's sake. I haven't done it again today.

I put my hand on the panel and indeed it's not hot. Maybe 30C or so.

If that's the case, the I would be very concerned that your panels are deteriorated.

Right now with sun on the panels, I get 6 amps if I short them out.

Seems a little on the low side.

Between low Voc under mild temps and low Isc, it sounds like your panels are performing well under spec.

 

[wpns]

Right now with sun on the panels, I get 6 amps if I short them out.

So they are functional, dunno where you are or the sun angle or intensity, but that says they are at least doing something. What's the voltage at the batteries if you connect them directly to the batteries?

 

[GXMnow]

Right now with sun on the panels, I get 6 amps if I short them out.

This does not sound too bad. 6 amps instead of 8 is 75% which is about what I get on average days with reasonable sun.

So this points to the batteries being in OVP. The LFP cells in the SOK battery might be a little out of balance, which can cause it to go into a single cell over voltage even with the total voltage looking good. Do you have a way to connect to the BMS and see the cell voltages and the protection status?

Connect a load to the battery and bring it down a bit. The BMS should act like a diode and still allow discharge with a cell in high voltage protect mode. If it becomes too discharged, it works in the opposite direction. In under volt protect, it will shut off loads but still allow charging. Once you pull the battery voltage down, try connecting the solar panels again and see if you get some current flow. But DON'T leave the panels connected without a charge controller. If that works, you can try going through the charge controller again, but it may very well have died. In fact, the FETs may have failed shorted and that is why the battery went into over voltage protect.

Do yourself a big favor and get an MPPT charge controller. You can get a 100 volt 15 amp Victron pretty cheap now. Amazon is showing it for just $78.20 with Prime.

 

[iwasforcedtopickausername]

The battery is at 50.9 volts right now and if I connect the pannels, it stays at 50.9 volts.

Seems like the voltage must plummet when there's a draw. I'm wondering about setting up a load or 24V battery to power with one panel at a time.

There's a load active (about half an amp) that the panels could be powering as well.

 

[sunshine_eggo]

The battery is at 50.9 volts right now and if I connect the pannels, it stays at 50.9 volts.

Seems like the voltage must plummet when there's a draw. I'm wondering about setting up a load or 24V battery to power with one panel at a time.

There's a load active (about half an amp) that the panels could be powering as well.

Between low Voc under mild temps and low Isc, it sounds like your panels are performing well under spec.

 

[wpns]

The battery is at 50.9 volts right now and if I connect the pannels, it stays at 50.9 volts.

Is there current flowing into the battery?

 

[Supervstech]

The battery is at 50.9 volts right now and if I connect the pannels, it stays at 50.9 volts.

Seems like the voltage must plummet when there's a draw. I'm wondering about setting up a load or 24V battery to power with one panel at a time.

There's a load active (about half an amp) that the panels could be powering as well.

Battery voltage isnt likely to rise quickly with 8A max input, but is the pv input measuring amps?