Good day every one.

A very loud bang came from the room where my 5 kwh Meritsun Powerwall LIFEP04 Battery was being recharged by the 5kva FOSHAN OUYAD ELECTRONIC (Axpert type) hybrid inverter. We had a power failure of about one hour earlier and the inverter has now been recharging for the last 30 minutes or so.
Upon investigation I found the battery indicating an overcharge alarm and the inverter showed battery disconnected.

Checking the battery indications and consulting with the battery supplier in China it was concluded the battery to be ok. However the battery terminals showed zero volts instead of 50v.

I disconnected the two battery cables between battery and inverter.

I found the inverter to have a short circuit across its battery terminals. A diode test also confirmed a dead short.

With the battery still disconnected I switched  the battery off and on and 50v appeared across the its own terminals. The battery SOC was showing 100%.

I removed the inverter cover and could not see any sign of any blown component. With the inverter switched on again  the only abnormality is the "Battery disconnected" warning.

What could have made the loud bang and is it safe to retry connecting the battery?

I have not yet connected an PV panels.

I am an "out dated" electrical engineer and a newby on the site.

Advice will be greatly appreciated

 

Neville

My guess would be that the BMS disconnected the battery to protect it from being over-charged. The problem with that is that if you were charging from PV, then depending on your MPPT controller, the DC bus voltage could have gone to the open-circuit voltage of the PV array. Which would likely destroy your inverter.

Thanks Stanley but no pv panels connected

15 hours ago, nevinvert said:

A very loud bang...

I found the inverter to have a short circuit across its battery terminals. A diode test also confirmed a dead short.

...

What could have made the loud bang and is it safe to retry connecting the battery?

The loud bang is almost certainly several thousand amps blowing the legs off some shorted MOSFETs, and or splitting them. It is certainly not a good idea to reconnect the battery; the inverter's battery terminals are a dead short after all. Not rnough legs must have blown off to still show a dead short.

My guess is that this occurred because the battery charge voltage settings were too high. Unfortunately, the most common manufacturer (Pylontech) recommends very high charge voltages, and others seem to follow this bad recommendation. It may have been exacerbated by the clone not measuring the battery voltage accurately, or the inverter's voltage control algorithm may have overshot.

15 hours ago, nevinvert said:

I removed the inverter cover and could not see any sign of any blown component.

How about the sniff test? Usually blown power semiconductors exude a distinctive smell from the burned epoxy.

The MOSFETs that are most likely to be blown are all along one heatsink, along the right side of the unit as viewed conventionally, almost hard against the metal wall. That's on a real Axpert; the Ouyads in my preliminary searches appear to be clones, but the clone manufacturers usually copy the layout pretty well.

You would have to take out the main PCB to see the carnage in detail. However, you might be able to get the general idea using a mirror and torch, or perhaps looking through some ventilation holes on that side.

Repair of an inverter with this sort of [ edit: continued ] failure is difficult if you don't have the right equipment and knowledge. For example, a desoldering station is very useful. See the repair posts on PIP-4048MS/5048MS inverter-chargers here (use the index to find the relevant posts).

Edited June 16, 20206 yr by Coulomb
Completed last paragraph; I can't recall why it wasn't completed.

33 minutes ago, Stanley said:

The problem with that is that if you were charging from PV, then depending on your MPPT controller, the DC bus voltage could have gone to the open-circuit voltage of the PV array. Which would likely destroy your inverter.

Actually, the models that use high voltage PV strings actually boost the PV voltage to the bus voltage; the bus is happy at 500 VDC (the rating of the main bus capacitors). So the open circuit panel voltage has to be under 450 V (500 V for one model; it must have higher voltage capacitors and perhaps IGBTs as well).

As noted above, no panels were connected, but unless installed incorrectly, they could not have caused this problem anyway.

As I said earlier, I believe that the BMS disconnected the battery from the inverter. Some models are designed to work without a battery, but even with those having the battery disconnected suddenly might cause some sort of instability or voltage spikes leading to shoot through of the battery-to-bus DC-DC converter, and hence the loud bang.

7 hours ago, Coulomb said:

Repair of an inverter with this sort of 

I'm eager, though dreading,  to know the completion of your sentence.

Thank you for a very informative reply.

I bought the battery and inverter directly from suppliers though Alibaba. Trusting equipment to be reliable. Guidance on setting up may have not been that good. Meritsun the battery supplier was familiar with the inverter and said to select AGM battery and the rest would be ok. The BMS being self-sufficient. Regular bleeps without indication came from the battery. I was advised to modify the settings and I suspect that is where things went wrong.

I will post the new values here later

Thanks and good night

6 hours ago, nevinvert said:

I'm eager, though dreading,  to know the completion of your sentence.

Oops! I don't know what happened there. Here it is:

Quote

Repair of an inverter with this sort of [ edit: continued ] failure is difficult if you don't have the right equipment and knowledge. For example, a desoldering station is very useful. See the repair posts on PIP-4048MS/5048MS inverter-chargers here (use the index to find the relevant posts).

Edit: I meant to add that it seems that there is a lot of cloned equipment about these days. It's making it hard for buyers to know what they are getting. There is an expectation that anything sold commercially should "just work", but sadly it isn't always the case.

Edited June 16, 20206 yr by Coulomb

On 2020/06/15 at 1:21 PM, Coulomb said:

My guess is that this occurred because the battery charge voltage settings were too high. Unfortunately, the most common manufacturer

The battery supplier originally said to choose "AGM" mode on the inverter without any further changes. The Bulk Charge (CV) was 56.4 and Float 54.0 However after regular beebs from the battery, I was advised to reduce bulk charge voltage to 54 and float to 52.5.

Here is possibly the crux of the matter:

I could not change the values without changing fro AGM mode to "User Defined". I did not change back to AGM after the new voltage settings were applied. 

I only discovered too late that the new voltage settings would remain even changing back to AGM mode.

My question is: would the User mode in stead of AGM have had a detrimental effect? 

15 hours ago, nevinvert said:

I only discovered too late that the new voltage settings would remain even changing back to AGM mode.

My question is: would the User mode in stead of AGM have had a detrimental effect? 

The changed bulk/absorb and float settings only take effect when you have setting 05 at USE (user battery type). It keeps your changed voltage settings in EEPROM, but uses fixed AGM appropriate voltages any time the battery type is set to AGM.

So you won't have done anything detrimental, and you need the battery type setting at USE for LFP-suited settings to take effect.