So I made a booboo and reversed the polarity on the MPPT PV input, the DC breaker melted instantly and released a bunch of smoke and it took me a while to unplug a MC4 in the chaos. The 20A fuses are all fine by the looks of it and didn't blow so the 63A Breaker doesn't inspire much confidence. I even checked everything with the multimeter before plugging in to make sure voltages are what they should be etc but alas.

The manual talks about the battery not having reverse polarity protection but what about the MPPT? I dont want to do more damage so interested if someone knows? I measured the resistance on the MPPT's one being +-0.7M Ohm and two +- 1.4M Ohm but I am not sure what they are normally but hoping that means they are still OK.

 

Edited October 29, 20223 yr by Beachless

10 hours ago, Beachless said:

the DC breaker melted instantly and released a bunch of smoke

Well, that's a huge fail. Is it really DC rated? 

10 hours ago, Beachless said:

The 20A fuses are all fine by the looks of it

Hopefully that means that the current was limited to about 30A, depending on how quickly you broke the current. 

10 hours ago, Beachless said:

it took me a while to unplug a MC4 in the chaos.

You should really replace the MC4 connectors, they are not designed or rated for disconnecting under load. If course, you were right to disconnect in this emergency, rather than waiting for sunset 🔥😵

 

10 hours ago, Beachless said:

The manual talks about the battery not having reverse polarity protection but what about the MPPT?

Almost certainly there will be nothing there. But the flywheel diodes in the MOSFET or IGBT will have conducted, and panels are inherently current limited, and the flywheel diode would be rated for full panel current. However, that doesn't explain what happened to the DC breaker. Perhaps it's a rare polarised kind, so the reverse polarity would have made its arc protection much worse rather than better. I believe that polarised DC breakers are ni longer legal in many countries, including Australia. I have no idea about South Africa. 

So cautiously, my guess is that after replacing the MC4s that disconnected under load (no need to replace any others) and of course the DC breaker, it should be safe to try out. I suggest waiting for low sun conditions though. Your meg ohm resistance readings are very encouraging: usually, semiconductors fail short circuit, which I'd expect to see at the MPPT inputs. 

3 hours ago, Coulomb said:

Well, that's a huge fail. Is it really DC rated? 

Hopefully that means that the current was limited to about 30A, depending on how quickly you broke the current. 

You should really replace the MC4 connectors, they are not designed or rated for disconnecting under load. If course, you were right to disconnect in this emergency, rather than waiting for sunset 🔥😵

 

Almost certainly there will be nothing there. But the flywheel diodes in the MOSFET or IGBT will have conducted, and panels are inherently current limited, and the flywheel diode would be rated for full panel current. However, that doesn't explain what happened to the DC breaker. Perhaps it's a rare polarised kind, so the reverse polarity would have made its arc protection much worse rather than better. I believe that polarised DC breakers are ni longer legal in many countries, including Australia. I have no idea about South Africa. 

So cautiously, my guess is that after replacing the MC4s that disconnected under load (no need to replace any others) and of course the DC breaker, it should be safe to try out. I suggest waiting for low sun conditions though. Your meg ohm resistance readings are very encouraging: usually, semiconductors fail short circuit, which I'd expect to see at the MPPT inputs. 

Thanks Coulomb.

I recently bought the complete wired combiner box so its all new and the breaker is rated 63A, 600v DC no polarity indicated anywhere on the breaker. The MC4 connectors were just used as they came already wired into the combiner box but I'm glad they were there as that was the only way to disconnect things.

The good news after inspecting things early this morning and switching in a spare DC breaker everything is working as it should. The MC4 looks good I hit the power button on the inverter somewhere in the chaos and I think either the melted breaker or switching off the inverted dropped the current so the connector wasn't under much load.

Why the breaker self combusted with less than 11A load I don't know but I'm just glad its the only thing that seems to have been damaged. 

My guess would be it is not related to AC or DC breaker. Pure bad workmanship on the internals causing a hot connection. I guess that DC breaker just had an internal hot connection.

This is an example of a 15A toggle switch under a 7A load for about an hour. I switched it on and had a chat with the neighbour and on my return I was shocked to find this remains. It was a new toggle.

5 hours ago, Scorp007 said:

My guess would be it is not related to AC or DC breaker. Pure bad workmanship on the internals causing a hot connection. I guess that DC breaker just had an internal hot connection.

This is an example of a 15A toggle switch under a 7A load for about an hour. I switched it on and had a chat with the neighbour and on my return I was shocked to find this remains. It was a new toggle.

 

 

That made in China sticker frequently lives up to its name. Its really hard to trust any of it.

Edited October 30, 20223 yr by Beachless

7 hours ago, Beachless said:

its the only thing that seems to have been damaged.

Thinking  about it a little more, it sounds like the reversal was at the inverter. So the breaker would not have seen reverse polarity anyway, just panel short circuit current, which is maybe 5% more than working (Imp) current. Which has to be waaay less than 63 A.

4 hours ago, Beachless said:

That made in China sticker frequently lives up to its name. Its really hard to trust any of it.

The problem is becoming, that it's nearly ALL made in China. Even the best brand name stuff.

Maybe branding is actually the answer, except when the knockoff manufacturers make junk that looks just like a brand name product.

Maybe block-chain is somehow the answer, so the buyer can be sure that he's getting a product that can be traced back to a reputable manufacturer. So then you just have to worry about the "unofficial night shift" where some workers use the boss's equipment and credentials, but they bring in junk parts for the night shift.

22 minutes ago, Coulomb said:

Thinking  about it a little more, it sounds like the reversal was at the inverter. So the breaker would not have seen reverse polarity anyway, just panel short circuit current, which is maybe 5% more than working (Imp) current. Which has to be waaay less than 63 A.

The problem is becoming, that it's nearly ALL made in China. Even the best brand name stuff.

Maybe branding is actually the answer, except when the knockoff manufacturers make junk that looks just like a brand name product.

Maybe block-chain is somehow the answer, so the buyer can be sure that he's getting a product that can be traced back to a reputable manufacturer. So then you just have to worry about the "unofficial night shift" where some workers use the boss's equipment and credentials, but they bring in junk parts for the night shift.

I would say not more than a few amp above 20 did flow as 20A fuses did not blow. Thus irrespective of reverse polarity the short in the breaker was between + &  -. I have no idea of what current could flow from MPPT to the panels in such a connection.

11 hours ago, Scorp007 said:

I have no idea of what current could flow from MPPT to the panels in such a connection.

Unless the solar charge controller is damaged, and it seems it's ok, power can't flow out of the PV input terminals, apart from a nip ir splat from the charged capacitors. Nothing that should smoke a contactor. 

Power can however flow from either PV input to earth. Is that possible? 

I'm really not sure but I'm going with a bad breaker. It burned holes on either side of the breaker melting to the one next door. The fuses are in line on positive and negative from the panels so there should be no way more than 20a flowed through the breaker from the panels. 

I don't believe power can flow from the MPPT through a fused breaker back to the MPPT without there being damage somewhere either.

My suspicion is that the breaker was not actually rated at 600V DC. That is to say - it's not rated to break a load at 600V DC. To do that most breakers have a spring loaded mechanism in order to make the gap big enough, fast enough to quench the arc. (most likely also in combination with magnets to direct the arc away from the other contact  - one of the reasons why DC breakers are usually polarized.)