Problem with new small array

Estragon · October 2017

I've set up two new small arrays to be used mainly to offset self-discharge and charge controller consumption over the winter. The big arrays have had trouble with snow cover over the last couple of winters.

The 12v array works fine. I was a bit concerned it might have too much voltage drop as I reused some old aluminum wire I had around for the ~100ft run from the 2 150w panels wired in parallel to the Morningstar ts45 pwm controller. It worked out okay though.

Surprisingly, the 48v array didn't. It's 2x300w panels wired in series, which should have Vmp of ~75v. Voc of ~90v measures okay, but under load the voltage drops to around battery voltage. I expected some drop, as I only used 10ga copper for the 100ft run, but it shouldn't be anything like 25v. I've checked connections, and they all seem good. Measuring voltage at the connection from the mc4 extender cables to the home run about 20ft from panels still measures ~52v, so I'm thinking maybe a problem with a panel or mc4 connector.

My plan is to wire each panel by itself, and connect an old dead 12v battery as a load with jumper cables directly to the home run connection, on the assumption I should see something like isc (~37v) on one of the panels and not on the other. Is this likely to blow anything up? I also have a couple of old GC batteries I could use to make a dead 24v bank if just the 12v could be dangerous.

BB. · October 2017

You can put Vmp~35 volt panels on a 12 volt battery--It will not hurt the panel (assuming that the battery=>panel polarity is correct).

For the battery, as long as it can absorb the available current and stay below ~16 volts (and not get too hot/boil the electrolyte too hard), it should be fine to be use as a load bank. Of course, since the batteries are no-good anyway, as long as they do not overheat (and begin to self destruct), you can run current through them (don't let them boil dry--Could get an arc internally and set off the hydrogen+oxygen gasses from electrolysis).

The closer you can get the load bank voltage to Vmp-panel, the better you can estimate the panel and wiring viability.

-Bill

Estragon · October 2017

Ok, thanks. Will get organized today and hope for sun tomorrow. Closing up for freeze-up on Monday, and would like to at least know what the problem is, even if I may not be able to fix it.

If I find one of the panels bad, maybe I'll try cutting the mc4(s) off and wire directly.

Raj174 · October 2017

@Estragon
Are the two 300 panels connected to a Classic 150 CC? If they are, try setting the MPPT mode to Legacy P&O. I did this recently and it made a big difference in array output due to low voltage PV.

Rick

Estragon · October 2017

@Raj174 - No, these are connected to a Morningstar 45a PWM controller. I could lower Vabs, as it's really just for floating though.

Do you happen to know if the Classics use less power as tare loss if solar mode is "off"?

Raj174 · October 2017

No, I haven't noticed a difference.

Ocean · October 2017

Estragon said:

@Raj174 - No, these are connected to a Morningstar 45a PWM controller. I could lower Vabs, as it's really just for floating though.

Do you happen to know if the Classics use less power as tare loss if solar mode is "off"?

Maybe I'm a little confused, but it sounds like what you have is two 300 watt panels wired in series for an O.C. voltage of about 90 volts (assuming about 9 amps). OK. But then you say it's a 48v system. OK. But then you say they are wired into a 45a PWM controller. Wait. the PWM controller is not a MPPT controller. When the PWM controller is working, You should see battery voltage on the panel side (input side) of the controller. It's just allowing the panels to push the batteries up to a certain point. Above you mentioned "I should not see anything like a 25v drop" - but you will. Because it's not a MPPT controller. The input voltage will drop all the way down to battery voltage... then start pushing them up as far as they can push - the PWM will stop the panels from charging the batteries when they reach the programmed BULK time limit... then let the float, only allowing enough current through to let them float. BUT it will not do the fancy trick of allowing the panels to run at their Vmp when the battery bank is less (i.e. the batteries are at 50v and the Vmp of the panels would be 72)... that's what MPPT charge controllers are all about.

If it were an MPPT controller, you would see the 65 - 72 volts on the input side of the controller, as you would expect for the Vmp of the panels.

Estragon · October 2017

I've not used a PWM controller before. As you say, I see the array voltage on the input of the mppt controllers, but I'm not sure about the new PWM ones.

The input on the 12v PWM was about array voltage (~Vmp, 18v - 2x150w panels in parallel) IIRC, and the controller LEDs indicate charging mode. The input on the 48v is roughly battery voltage, and the controller LEDs indicate state of charge only. When the sun comes up, I'll check if ~Vmp is present on the input of the 12v PWM controller.,

I'd love it to turn out that you're right as I have to get this working today.

mcgivor · October 2017

Personally I'm a little confused, due to the mixing of information of 2 systems. Dealing with the PWM system, input voltage should be pulled down to just above battery voltage, the indication LED's on the Morningstar are for voltage during charging, not the status when the sun goes down. During the day a solid green represents bulk, a flashing 1/2 second pulse represents PWM, if the culumitve time in PWM falls below 30% for 1 hour, the controller will go into float, represented by a flashing green 1 pulse per second. Float is rarely seen unless the battery is healthy with no loads. Here is the explanation from Morningstar tech support.

The duty cycle is used to determine when to transition into float. The duty cycle is derived from a square waveform and the percentage of time - out of each complete cycle - when current is allowed to flow. Current is regulated by opening and closing the PV-battery circuit -- the essence of a series charge controller. A 50% duty cycle means that current is flowing, and interrupted for, half of the cycle. Our controllers pulse with a frequency of 300 Hz, so each cycle lasts 1/300th of a second. By varying the duty cycle, we can control the average current going into the battery so that the regulation voltage is maintained but not exceeded. As the battery reaches full charge at a particular voltage, it will require less and less current to maintain that voltage and the duty cycle will decrease. The duty cycle will approach zero %, but will never actually get there because losses and battery self-discharge require at least some current to maintain regulation voltage. A Morningstar PWM controller's float transition algorithm waits for the duty cycle to reach 30% or less before starting a one hour timer. When the timer expires, the controller moves from absorption to float voltage. A 30% duty cycle simply means that 30% of AVAILABLE charging current is allowed to pass. Duty cycle is the best measure we have to determine how deep a battery may be in regulation. Factors affecting the accuracy of this method include: 1. Reduced radiation and current. With less available charge current, the duty cycle will inevitably be higher making the transition to float more difficult; 2. Load on the battery. Loads on the battery drain current from the battery which causes a higher duty cycle. If the load is large enough (greater than the amount of solar current ) it can pull the controller into 100% duty cycle and out of absorption - into bulk charging; 3. Aging batteries. Old batteries tend to absorb more current at regulation, keeping duty cycle high. Sulfation leads to higher internal resistance, and more energy loss in the form of heat, which leads to excessive water loss.

At least that's for the PWM side of things.

Estragon · October 2017

The explanation of the lights helps. The 12v was blinking, which I though indicated charging, but could well have been in absorb. That bank was not discharged much, and could have been fully bulk charged already by the mppt array before I checked the new PWM system, so the blinking just indicating out of bulk and into absorb. I haven't checked, but I assume the SOC indication doesn't light at night, so the solid green or whatever SOC indication indicates bulk?

This may also explain why I saw nearly Voc on the 12v PWM input and near Vbatt on the 48v PWM input. If the duty cycle in absorb on the 12v bank was low enough, presumably that would allow the input voltage to rise to near Voc, whereas the lower SOC on the 48v bank would mean the 100% duty cycle pulls the input voltage down to near Vbatt?

Looks like it will be a nice sunny day today, so should be able to confirm this. If the input voltage being pulled down to near battery voltage is normal, that means there likely isn't a problem

Thanks.

Estragon · October 2017

The hoped for sun didn't show up much today, ended up having to charge with the big arrays and generator. Hard to tell if the small array is working properly, but the trimetric did show varying current with just the small new arrays charging as clouds varied in thickness.

About the best I saw was ~3a briefly with not quite full sun into the bank with no loads at ~60%SOC. The pair in series have an Imp of a bit over 8a. I really should get a clamp-on ammeter. Hopefully Sears will still be around when I'm in the US in Feb.

Problem with new small array

Comments

Categories