Solar controller meltdown!

Back in late 2018, I installed a solar panel and solar charge controller to run a pump to heat the swimming pool.

I never made any post about the system as I never got round to it! The new revised system following this meltdown will be posted soon.

Anyway, the charge controller developed a fault and stopped charging the battery. When I opened it up I found that it was full of bugs, and one of the MOSFETs had melted and burnt away part of the PCB and the legs of the MOSFET.

I did some probing of the board, and it appeared that it was just the one MOSFET that had failed, so had a rummage about my various bits and bobs, and found an old power supply that I know was faulty (it had come from my old QNAP NAS box which has long been retired and scrapped).

With the current lockdown in place, I couldn't get hold of a replacement controller as the shop in Paphos where I got it was under the closure restrictions. Getting a new one by post was an option, but again, this could take a few weeks as there was also postal delays, and it didn't justify courier costs.

I pulled a larger MOSFET from the NAS PSU and soldered it in place and much to my surprise it was working well!
It is not pretty, but it fits!

As you can see in the above image, the replacement MOSFET was squeezed in and was charging the battery at 0.5A on the test I was doing, the faulty unit always just read 0.1A no matter what was happening. On the lower picture you can also see some slight PCB scorching above the left terminals. These are where the solar panel connects.

I put it all back together and shoved it back outside. the MOSFETs on the board are all squeezed against a metal heat sink by the case screws. When I put the unit into service with the solar panel I was getting the full 5 to 10A charge current.

But, and this is a big BUT, the unit was absolutely roasting when the charge current was over 5 Amps. I also noticed that the back light for the display had stopped working, so suspected there was more at fault than just the failed MOSFET.

To mitigate against the increased current leading to increased temperature, I secured a towel over the solar panel, to limit the current to below 5A when the sun was out and the sky was cloud free.

That failure was mid April 2020. roll forward 2 weeks and I happened to notice the pump was running in the morning which is good, but around half an hour later I was passing in the kitchen and looked out the window and notice it had stopped pumping.

I thought, it had packed in completely now, so went out to investigate.

Well, I hadn't put the towel in place, I had removed it the previous day when it was heavily overcast to at least get the battery charged when the pump was not running. Unfortunately, it was now clear skies, so that means the charge current was probably maximum and the unit has overheated. When I say overheated, I know when I had been sat watching it, the temperature indicator would reach over 80'C until I disconnected the PV panel.

When I went out the back door, I could instantly smell cooked electronics. Went round the corner to see smoking billowing from the area.

On closer inspection the charge controller was melting, the wiring was all melting. I didn't have any tools, and there was no switches, normally I disconnect the solar PV plugs, but no way I was going to grab a hold of them and try to get them off. Fortunately, when I touched the solar connecters, they just fell off the charge controller. The cables feeding the battery were still glowing and melting the insulation. I nipped into the shed and got a pair of plant trimmers and simply cut the battery cable to kill the system dead.

The fault was obviously allowing the battery just to dump all the power in a short of some form.
A bit toasty

Insulation stripping lesson 1

Okay, so it was hot you say?
I guess I need a new charge controller, I think some better fuse protection is also needed. Well noted for the next revision!

Next revision coming soon......


Popular Posts