Solar charge controllers overcharging during full sun hours.

Hey guys,
I own a wisp and have about 15 solar/battery sites on the top of mountains outside of San Diego. I have various batteries/charge controllers/solar panels at each site.

Currently I have an issue where during peak sun hours my 24v system will shoot up to 30v. I have tried multiple charge controllers both cheap pwm and MPPT and they all seem to follow the same issue. I'm starting to think maybe the solar panel could be an issue. I'm using 2x old Trina 36v/8 amp panels in parallel and 24v Vestwoods 300ah LIFPO4 battery. I've talked to the makers of the charge controllers but they have never heard of anything like that and couldn't help me. Was wondering if anyone else ran into something like this. The panels are old is there anything I should do the service them. Besides overcharging the batteries, the radios used to service customers shut off once the voltage hits 30v+, which generates support calls.

Attached is a graph from my voltage monitoring program.

krumbles said:

Hey guys,
I own a wisp and have about 15 solar/battery sites on the top of mountains outside of San Diego. I have various batteries/charge controllers/solar panels at each site.

Currently I have an issue where during peak sun hours my 24v system will shoot up to 30v. I have tried multiple charge controllers both cheap pwm and MPPT and they all seem to follow the same issue. I'm starting to think maybe the solar panel could be an issue. I'm using 2x old Trina 36v/8 amp panels in parallel and 24v Vestwoods 300ah LIFPO4 battery. I've talked to the makers of the charge controllers but they have never heard of anything like that and couldn't help me. Was wondering if anyone else ran into something like this. The panels are old is there anything I should do the service them. Besides overcharging the batteries, the radios used to service customers shut off once the voltage hits 30v+, which generates support calls.

Attached is a graph from my voltage monitoring program.

Click to expand...

Not sure what to make of your chart. It doesn't make sense in any context unless they are showing 10X actual battery voltage.

Your LFP batteries are almost certainly entering charge protection due to internal cell imbalance. One or more cells have hit the protection limit, and the BMS terminates charge. The batteries cut themselves out of the charge current, and the charger can't react quickly enough, so the charger voltage spikes.

Try lowering bulk/boost/absorption to 27.2V and boost/absorption time to 6 hours, float to 27V. This should eliminate/diminish spikes if this is an imbalance issue, and the longer absorption period will give the BMS more time to balance.

Before deploying a new battery, it's a good idea to confirm they will charge up to 3.55V/cell without abnormal termination, i.e., go to 3.55V/cell and have the current taper to 0A - not instantaneously plummet to 0A.

Sounds like the BMS is shutting the batteries off to prevent overcharging them. Can you query the BMS? Can you set the charge controllers for something like 27.5 volts?

SOLVED

Since my load only consumes 4ah. having 2 solar panels that create 16ah was overloading the charge controller. I decided I don't need both solar panels so I unplugged one of them. The charge controller has been able to keep the voltage @ a steady 27v during sunlight hours. I guess I'll run 1 panel during the summer and 2 during winter.

I did spend some time chatting with chatgpt about the issue and AI recommended a Zener diode to break away any voltage over 30v. but I would need to dissipate 96w which I was not able to find a diode big enough for that.

Unplugging the solar panel was a simple solution to a complicated issue.

None of this makes any sense. Where is the 300v in the graph coming from?
Not the 24v battery, or the two 36v solar panels.
And if the panels were actually wired in parallel, removing one doesn't change the maximum voltage. (Unless they are two different types of panels)

krumbles said:

SOLVED

Since my load only consumes 4ah. having 2 solar panels that create 16ah was overloading the charge controller. I decided I don't need both solar panels so I unplugged one of them. The charge controller has been able to keep the voltage @ a steady 27v during sunlight hours. I guess I'll run 1 panel during the summer and 2 during winter.

I did spend some time chatting with chatgpt about the issue and AI recommended a Zener diode to break away any voltage over 30v. but I would need to dissipate 96w which I was not able to find a diode big enough for that.

Click to expand...

I have yet to see a sensible answer from chatgpt.

Your system is improperly sized.
Your solar panels are connected to a charge controller. The charge controller is connected to a battery.
The battery is connected to an inverter.
The loads are connected to the inverter and receive power from the inverter.

PV OUTPUT WILL BECOME ZERO IN DAYLIGHT ONCE THE BATTERIES ARE FULLY CHARGED AND THE LOADS ARE ALSO VERY LOW. ONCE THE LOADS BEGIN TO DEPLETE THE BATTERIES, THE CHARGE CONTROLLER SHOULD WAKE UP AND PV HARVEST WILL RESUME WITH A CHARGE CONTROLLER running the charging program built in..

YOU MUST USE CAUTION IN THIS REGARD. ALL CHARGE CONTROLLERS HAVE A VOLTAGE IMPUT LIMITATION CALLED Voc. This is the maximum output voltage from the panels to the charge controller.
All charge controllers have a maximum amperage input too. Connect a group of panels that exceeds the maximum allowed charge controller amperage or voltage and you will fdamage the electronics of your charge controller.

krumbles said:

Unplugging the solar panel was a simple solution to a complicated issue.

Click to expand...

As others have noted, this 'solution' is a workaround that doesn't solve the real problem, so you are probably in for more confusion in the future. That's a lot of sites and a lot of customers to have balancing on the edge of "I think that fixed the problem I don't understand", when everyone here is telling you it's probably charge controller settings causing the batteries to trip out on over voltage.

I am in San Diego. You can send me a PM if so inclined and maybe we can talk on phone and solve issue.

krumbles said:

I did spend some time chatting with chatgpt about the issue and AI recommended a Zener diode to break away any voltage over 30v. but I would need to dissipate 96w which I was not able to find a diode big enough for that.

Click to expand...

AI is really making breakthroughs nowadays, it's reached the intelligence of a relatively new forum enthusiast who's gaining steam but hasn't yet learned we don't add diodes anymore.

Sorry been busy with family stuff.

My graphs need to be divided by 10 to make it 30.0. I'm just used to looking at 300v and realizing its 30.0v.

The way I am making sense(I maybe wrong here) of this is my load is 27v@4amps 24hrs/day which becomes 2592 watts. A 36v panel at 8amps with 7 hours of daylight gives me 2016watts. 2 panels would double that to 4032watts. I think after the Lifepo4 batteries hit their limit of absorption the bms on the battery cuts the current flow which increases the voltage. I believe the charge controller has some ability to absorb some of the current but when its overpowered it allows the voltage to pass though to the battery/load.

My radios are direct dc(no inverter) with a ~30v max limit before shutting down due to over voltage.

My chat with chatgpt was asking it to limit the amount of voltage coming from the solar panel. I asked it to limit the solar panel from 36v to 30v. It recommended a zener diode(not a bypass/blocking) that would activate once the voltage goes over 30v and dissipate the excess power. Its what was recommended, I have not tried it and at $15(amazon - 1N3324B ZENER DIODE, 50W, 30V) for the diode I think i'll pass for now.

I guess to sum it all up there is such a thing as having too much power.

Thanks for the replies. I attached more graphs. If your interested.

What’s a wisp?

Wireless Internet Service Provider, I presume.

Bluedog225 said:

What’s a wisp?

Click to expand...

Wireless Internet Service Provider.

I have been doing exactly what this guy is with several wireless sites for 16 years. He is probably using Ubiquiti or Mikrotik gear, as am I. It runs off anything between 10 VDC to 24.

BUT:

I am running 12V FLA/AGM systems. And with those, have never had this problem.

However, I have had it with two non-WISP LFP systems. Problem is exactly what sunshine outlines in post #2. If the BMS kicks off due to overvoltage, the CC can't react fast enough and things spike. In my case, it was 36V on an 80VOC panel array on a 24V system. That would kick the inverter. In one case (Samlex EVO-2224) it would recover when the voltage went back below 32V. In another case (cheap MSW inverter), manual intervention in the form of a power cycle was needed.

And his solution is the correct one. Back off the charge voltage a couple tenths. That completely fixed the problem here.

One of these days I will take a spare test system and figure out just what the voltage can be before things go sideways. Unfortunately, there are so many variables I'm not sure you can predict what it will be. Just have to FAFO.