Trying to understand why my MPPT burned up.

[Guda]

I had a MidNite Classic 250 burn up after 15 ish months. I was at about 180v & 20a ish. The 250 can handle 250v 63a in. I was well below that. Now I am thinking it might have been my pv arrangement. I had 6p-6p-3p. Arrays ideal for summer, winter & east. The reason for that is a few. 1st array was summer on the roof. Making a array at 45 degrees on the roof was not what I wanted to do. 20 degrees was easy & the wind does not grab it at that angle. Next I put up east to get power as early as possible. Then the winter array is mounted so I can get the snow off easy.

The 250 months ago was reporting "max amps" & a yellow light was on. This was strange as I was only putting about 20amps at max. I'm not so sure there wasn't a issue with the 250 as it was a strange version.

I just found a burned connector on a pv from the east array that gets shaded. Although I have ideal diodes & breakers, I think something is going on with the different arrangements. Like the "T" in MPPT is for tracking. So if there is different V & A coming from the pv then the MPPT must not be tracing correctly.

At this point I am on the summer array only. Something is up with the winter array that I need to figure out. I will have another AIO in a couple weeks. When that comes in I'll connect the winter array to it. The east array is just going to do nothing until I get the other 2 AIO.

I am not sure whats going on technically but I am not going to have differently aimed pv on 1 mppt any more.

 

[time2roll]

If you replace with same I would limit the max amps and clip power vs running at max every day.
Did it ever get warm/hot?

 

[Hedges]

"180V & 20A", that would be Vmp.
What matters is Voc. But for 180 Vmp, Voc is probably around 215V, shouldn't be a problem.

"6p-6p-3p" - what do you mean?
Between funny array and burned connector, I'm wondering if something going open caused voltage to pop higher.
Need to understand your array series/parallel configuration, and see panel specs.

 

[Guda]

This video shows my pv setup. As the 250 was dying it was reporting very high voltage. 260v-360v. I never noticed any heat. But one connector and some wire was over heated on a pv.

My pv
Used
250w 8a 30v
265w 8a 30v

I might be off by a little. I have 2 different sizes.

 

[Hedges]

A one hour video isn't a useful way to document PV panel wiring. How about a sketch, showing the panels as 15 rectangles with lines between them for wires?

Reported 260V to 360V explains why it was killed. Midnight Classic has "Hyper Voc", which tolerates a slight over-voltage, like if Voc of 240V goes a ways over 250V on a cold morning.

Voc of your panels is critical to evaluating the array.
"30V" might be Vmp for a 250W panel, not likely Voc.

Given your sketch of array wiring and panel parameters, model, label, or spec sheet I can calculate Voc.
You need to get it right for your new controller or that would die too.

 

[Guda]

It was so you can see orientation. 6p-6p-3p. Those each went into a 4 to 1 pv combiner. Each had a breaker/fuse & ideal diode. The 6p are 250w, the 3p are 265w

The high voltage was only in the 250. The 250 was not getting high voltage. It was faulting up to high voltage.

Max power 250w 265w
Open Circuit V 37.5v 44.2v
V Pmax 30.3v 35.33v
Short circuit current 8.85a 8.2a
Current pmax 8.27a 7.5a


The "new" MPPT is only running 3p2s of the 250w. I not to worried about the "new" system as I am running 2 MPPT and no east array. I am mainly trying to wrap my head around why the 250 burned up.

 

[Guda]

The forum moved my #'s. They were all spaced out nicely when I posted it.

 

[RobertGreen]

The high voltage was only in the 250. The 250 was not getting high voltage. It was faulting up to high voltage.

It's entirely unclear, to me at least, what it is that you are trying to say.

You say that "As the 250 was dying it was reporting very high voltage. 260v-360v."---- Was the voltage actually going up that high? Or are you saying that this reported voltage was a measurement error on the part of the MPPT unit?

 

[Hedges]

The "new" MPPT is only running 3p2s of the 250w.

"3p2s" is not a typical PV panel configuration.

Please draw a schematic of this new array configuration, and of the old array configuration.

 

[Steve_S]

proportional fonts mess up TXT use the "CODE" under more options for strict Courier ascii txt.

This is a Midnite Classic, they will ONLY tolerate Batt-Volt above Rating courtesy of HyperVOC function. Cross that by even 2 Volts and Poofdah ! Considering that the Majority of SCC's have no HyperVOC capability, you would have likely fried other SCC's.
The MAX a CL-250 with 48V Battery System can take is 298VDC Period !

With Midnite, always use their Sizing Tool to configure your solar panel array

MidNite Solar - Classic Sizing Tool.

MidNite Solar is the industry leader and manufacture of quality Renewable Energy System electrical components and E-Panels.

www.midnitesolar.com

The Midnite Solar Forum has all the Key Players watching it, including Robyn and if you want good proper answers with regards to Midnite Gear, that is the place to discuss it. http://midniteftp.com/forum/index.php

I use Midnite SCC's and have deployed several, they are Great Products but do NOT go outside of spec or push them beyond design limits & specs. Midnite can & will repair their products as well...

 

[Guda]

You say that "As the 250 was dying it was reporting very high voltage. 260v-360v."---- Was the voltage actually going up that high? Or are you saying that this reported voltage was a measurement error on the part of the MPPT unit?

The 250 was saying the voltage was that high. Voltage was not that high going in

I drew it up but iPhone has become a worthless tool if you need content off of it. No idea where the photo went. I see the video that some how came off my gopro & wont delete...

6p means 6 pv in parallel. So when I say 6p-6p-3p. It means 6 pv in parallel with another 6 pv in parallel with another 3 pv in parallel. All 3 of those parallel groups are in series.

 

[Guda]

I use Midnite SCC's and have deployed several, they are Great Products but do NOT go outside of spec or push them beyond design limits & specs. Midnite can & will repair their products as well...

I had a 150 running my wind setup. Loved it. Lost all that in the fire. This 250 was a weird deal. A company ordered a custom lot that was not normal 250. Company went under & I got one for $450. Here is the whole story

 

[Hedges]

The 250 was saying the voltage was that high. Voltage was not that high going in

I drew it up but iPhone has become a worthless tool if you need content off of it. No idea where the photo went. I see the video that some how came off my gopro & wont delete...

6p means 6 pv in parallel. So when I say 6p-6p-3p. It means 6 pv in parallel with another 6 pv in parallel with another 3 pv in parallel. All 3 of those parallel groups are in series.

I thought the oddball 250's were made with some features missing, but not different specs.


Three such groups in series, even if all the higher 44.2Voc, would only be 132.6Voc. Cold would raise that < 1.2x, so 159V or less. No where near limit of Midnight Classic 250.

Six in parallel 8.85 Isc is 53A max.
Output current of 250 is 55A to 62A depending on battery voltage, derate to 52A at 40 degrees C.

"Derating" ought to be accomplished by the SCC, it isn't what you have to do in array design.

Interesting that your PV array short circuit current could equal Classic's output current limit (if voltage pulled down so 3p is bypassed with diodes, 6p supplies current.) But I don't find an Isc limit in their manual.

Nothing here should cause excessive voltage input.

Rather than n panels in parallel, m groups of those in series, I'd rather do m panels in series, n strings of those in parallel. Each string can have a different orientation.

The burned connections are worth understanding. If 53A was put through one MC connector, sure that would overheat.
Or, if two brands were mixed (cable connector to panel connector), one might be inferior.

https://www.midnitesolar.com/pdfs/Classic-SL.pdf

 

[Hedges]

Your array configuration of 6p && 6p && 3p presents two peaks for the MPPT to find.
One is the current of 3p and the voltage of 3s, for the power of 9 panels.
The other is the current of 6p and voltage of 2s (activating bypass diodes of the 3p group), for the power of 12 panels.
Midnight would probably find the higher current, lower voltage peak (unless voltage too low given battery voltage).

That means current 6x Imp of the panels gets forced through bypass diodes of 3 panels, 2x the current they were designed for.
Current through a diode rises exponentially with voltage across it, and rises with junction temperature.
Parallel diodes can't be expected to share current evenly. One will go into thermal runaway, taking more current because it is hotter and getting hotter because it it taking more current.
Eventually, one diode is carrying all the current of 6 panels in parallel, and it fails.

I suggest you check performance of each panel, and determine if bypass diodes are good, open, or short.

I've had degraded panels that had reduced power output. Some showed reduced Isc and/or Voc. Others looked same as the rest as far as those parameters go, but when loaded (I used a few ohms from some electric oil-filled radiator heaters), their V(load) and I(load) were reduced.
Electric heaters have switches and thermostats that work for AC but not DC. I turned those on high so they wouldn't try to interrupt DC current, and used a DC rated switch to connect he panels for testing.

 

[Guda]

Am I confusing parallel & series? Parallel ups volts, series ups amps correct?

I've already read all this 3x. I understand more each time. I think this is the answer I was looking for. Now I just need to study it to understand it all. Much thanks!

Testing pv is tomorrow hopefully

 

[RobertGreen]

Am I confusing parallel & series? Parallel ups volts, series ups amps correct?

That is incorrect. The opposite is true.

 

[Hedges]

Schematics of the panels would be useful, showing what the array was like before, and now.
Thinking about it, I've come up with several ways things can go wrong. I'm good at looking for trouble.
(Prime contractor's engineers and program managers didn't seem to appreciate it, but better to fix up front before catastrophic failures.)

The other thing that helps is testing voltages before connecting wires together, and before connecting electronics. Make sure it meets prediction.
But, there are some bad configurations with failures that could cause voltage to pop up later.

 

[RobertGreen]

Thinking about it, I've come up with several ways things can go wrong. I'm good at looking for trouble.

What are your thoughts as to how an array that was operating normally for 15 months can develop a problem and increase in voltage so much?
He's saying that it was reading as way above the open circuit voltage of the array? If things didn't get rewired so as to put more things in series, how could the voltage rise like that? (other than a measurement error being the cause of the crazy high voltage reading of course)

 

[Lt.Dan]

I had a MidNite Classic 250 burn up after 15 ish months. I was at about 180v & 20a ish. The 250 can handle 250v 63a in.

This is what concerns me the most. It seems you mix up a lot of terminology. The 250 Classic can only handle 63a OUT, as in, OUT to the battery. Assuming you are running a 48v battery, that could mean 180v X 20a = 3600w. 3600w / 48v (or even 51.2v nominal for LiFePO4) = WELL OVER 70a. If you were 20 ish, then that could be 25a or more which is almost 88amps OUT! Sure overpanelling is fine for most, but for how long? I guess yours was 15 months.

You over taxed the MPPT and it burnt out. Once it burnt out, somehow it raised voltage, or it was a fluke and reading 350v incorrectly because it was toasted inside.

 

[Hedges]

What are your thoughts as to how an array that was operating normally for 15 months can develop a problem and increase in voltage so much?
He's saying that it was reading as way above the open circuit voltage of the array? If things didn't get rewired so as to put more things in series, how could the voltage rise like that? (other than a measurement error being the cause of the crazy high voltage reading of course)

If you made a series string 9s, and paralleled it with one 6s, it would be held down around 250V
If a wire connecting the 6s went open circuit, it would jump to around 375V

That's why I would like a schematic showing how all 15 panels were wired.