Charge controller throttling

[plataoplomo]

Hello,

I have noticed a decrease in the production from one of my solar strings.

Background.

48 volt system. 3 Mitsubishi MF175 in series feeding an Outback MX60 charge controller. 30voc @ 7.92a short circuit for each panel.

Short circuit testing the string is 76 vdc @ 5.6adc. We are in the tropics. With the heat I think that is about normal.

I walked the voc/short circuit tests through the wiring all the way up to the charge controller inputs with no change in what I measured.

Here is the question......

Only 2 or 3 adc are seen by the charge controller.


Anyone ever seen this before?

[Mike90250]

is it only one string of panels that is deficient, or the entire array ?

If the batteries are full, the controller won't be trying very hard to charge them. Are the batteries in good condition, right amount of water ?

What happens about 9am, when batteries are low and sun is beginning to be high enough to produce good power

[plataoplomo]

is it only one string of panels that is deficient, or the entire array ?

Hi Mike. Thanks for the response.

The array I am having problems with is active tracking mounted. Originally rated at 4kw.

I have 1000ah of Changhong NiFe cells, and 920ah of refurbed Edison NiCd cells. I typically run 38 cells in a string.

My system has been in production for about 10 years. I noticed the drop off in watts over the last month or so.

On the array in question I have 8 strings of 3 panels each. These are my oldest panels. About 9 years old. 7 of those strings have diminished production.

I am getting 2 -> 3 amps out when I think I should be getting 5 -> 6. On the 8th string I am getting 5 -> 6 amps out. The 8th string was connected a year later than the first 7 strings. So it is a year younger.

As I said I tested each panel voc and short circuit. They measured about 30% less than the lab values from the manual. I chalk that loss up to heat and age.

I then walked those measurements back for each string all the way up to the charge controller. Nothing is being lost in the lines or breakers.

I then pulled a Midnight Solar Classic 150 that I have as a running spare and Installed it. Same diminished outputs on the same strings.

So then I thought that my controller my be struggling with bank resistance.

So for a test I lowered the number of cells in series seen by the controller down to 34 cells. In my mind that should have been low enough to encourage the cc to run the panels at full blast bulk. Same lowered productivity.



My mornings are a little different down here on the equator. Sun is up at about 0600. I angle and track on it about 0730 for close to full production until about 1630.

Right now on each of the 7 strings in question voltage and current rise up to the max production, (which I think is about half of what I should get, half of what I measure short circuit),


I have inspected each of the panels very carefully. I have one or two that have micro cracks in the wafers. The conductors across those cracks appear intact. No de lamination, scoring, or burning. Most of my panels have some faint brown discoloration on the solar cell interconnect busses.


Half splitting has led me to think that something has degraded 21 panels all at once in a couple of months time.


Insights?

[Mike90250]

If you have 2 controllers that give roughly the same output, it seems like the panels crashed. Wow. to loose them all in a couple months..
If you have the local app on the classic, it's easy to adjust your charging voltage. Crank it up. I run 42 cells in a string and charge at 67V, just under the inverter warning .

[plataoplomo]

Wow. to loose them all in a couple months..

Yup, that's what I thought too. Which is why I'm asking around. Anybody seen anything like this before?

What is stumping me is I can get the power I am used to with a short circuit test. That would lead one to believe that the panel is capable of it. And 21 identical failures all within a few months?

I always seem to get the good problems.


I have up-cell-ed a few times with my NiCd bank. Ran 42 cells for a week or so. Ran 44 for a day. You have the same settings ability with Outback gear under the advanced menu.

[Mike90250]

Short circuit current is not the same as power.

A realistic test, rig up a 150W incandescent flood lamp, and see if your string can get it to glow. Measure the volts at the bulb, use a meter (clamp-on or inline) to measure the amps.

Don't be breaking or making a connection with the MC4 connectors while under power. You will destroy the contacts and ruin them

[plataoplomo]

rig up a 150W incandescent flood lamp

Good idea. Thank you. Working on it.

[plataoplomo]

Sorry for the delayed response. I wanted to wait a bit before I called it.

I believe that I had degraded MC4's. That caused heat to build up in the panels themselves. In some cells this manifested by collector runs de-laminating. Other cells browned on the edges.

Troubleshooting led me to this document. You may find it of value.

"https://www.researchgate.net//publication/274717701_Review_of_Failures_of_Photovoltaic_Modul es"


I replaced all of the connectors for that array. Crimped/soldered connections. Then a healthy application of NO-OX-ID.

That array is back in service at about 70% of oem specs.

Moral of the story. Check your connectors.