Pure Sinewave Inverter burned, fake ir2110s?

[FabianLPHD]

Hey guys,

i bought a used pure sinewave inverter branded with Edecoa.
I first took it apart.

I thought it was not bad build.
It consists of two parts. DC-DC Step-UP with 400V and the SPWM inverter board, it's build like egs002.

Now to the problems. I have an 1400watts amplifier connected to it. It has a big transformer.
I turned the inverter on and the amplifier started working, but after a second it went down. No magic smoke, but no output power anymore.
Took the inverter apart and messured the outputs of ir2110s, because FETs are not shorted or anything seems blown.
The Output of high side of ir2110s was very bad, but the input was correct.
I had original ir2110s and solderd them in and it worked. 230V AC, brilliant.
But they look different to my original ir2110s, maybe you can verify if they are fakes?


But now i tried to solder in an egs002 board. Thought RS232 and LCD were an upgrade.
It had the same pinout and worked well.....BUT now, i turned on the amplifier while the inverter was on, the ir2110s(gnvs branded?) blow again  , gone short circuit completly from this board and a pair of 20n60 blow with it.

Now the other question is... Does the transformer of the amplifier create a peak voltage and destroy the ir2110s? or are they fakes and bad?
How i know that the ir2110s blow first? First the output voltage was low, some peaks and then the IGBT burned down.
I now orderd new 20n60 and solder in the original spwm board with original ir2110s. How long will this last? Hope, more than 2 Days 

Maybe you can follow me, sorry for my englisch 

Thanks for answer me, have a nice day!

[Messtechniker]

Yep, the inrush current kills it. German: Einschaltstromstoß.
You need some form of inrush current limiting between inverter output and your amplifier. Search for German:  Einschaltstromstoßbegrenzer.

[FabianLPHD]

Okay, yeah thought that too, but this is a 3,5kw/7kw rated inverter. It should handle that?
I want to connect an air conditioning, but with an inrush current limiter, the compressor can't turn on?

[FabianLPHD]

And added to say, the first ir2110s kill, was while the inverter powered up with soft start, so inrush current shouldn't be a problem?

[Messtechniker]

The soft start you mentiion applies to the primary
side of the inverter and not to its output.

By the way: powering up 3 ordinary tower PCs
(rated at around 300 W each) via a central switch
caused a 10 A circuit breaker to respond
every now and then. 

[FabianLPHD]

Okay, okay. But is an air conditioner working with inrush current limit?
Also, my modified sine wave inverter worked over a year with this amplifier, also directly plugged in without inrush current limiter.
This Inverter has a H-Bridge with 8 IGBT, they can handle much more current than the single Mosfet of the Square Wave Inverter. Why should be that be now a problem?
I don't think that inrush killed it first time, because the amp was on already, inrush was over. Second time possible, or can it be if dead time control is set wrong? That it short out itself while high current peaks?

[Messtechniker]

Only measurements can tell. Not easy considering
the currents, voltages and timing involved.

[FabianLPHD]

Okay, i solderd the old control pcb in, with the new IR2110S and replaced the blown IGBT with a Mosfet for debugging now, while trying with 12V.
A big difference is, that the original board modulate drives all 4 gates with spwm and create a unipolar output waveform.
The EGS002 did only spwm on 2 gates and "fundamental wave output"(datasheet) on the 2 others. This created a bipolar waveform.
I think this is the key why my IGBT exploded. Hope with the original circuit inverter will run propper. Parts are ordered.

I will report.

[FabianLPHD]

Maybe you can, or anybody else can, explain to me the difference between unipolar with only 2 IGBT SPWM and one conductor, or 4SPWM and two conductor.

How mine is build.


How the EGS002 thought it was build.

Schematics are not 1:1 the same, only for following

Is current flowing different in unipolar or bipolar operation?
Want to know the advantage or disadvantage.
Hope you can follow me.

[TheMG]

Unipolar if there's only a +VDC high voltage supply. If there's both +VDC and -VDC high voltage rails then it would be bipolar.

Disadvantage of unipolar I can think of is that the output neutral (N) usually can not be bonded to ground, as it has half the AC waveform on it. If neutral isn't left floating, it can cause problems.

[NiHaoMike]

For a large motor load, you'll want V/Hz ramp up to eliminate or greatly reduce inrush.

[FabianLPHD]

Okay, thank you guys!

Now i measured the DC Output from the DC/DC StepUp and it was 430V, is this not too high? Should it not be 380VDC?
With 430V there is to much power loss, or i'm wrong?

Thanks

[TheMG]

Now i measured the DC Output from the DC/DC StepUp and it was 430V, is this not too high? Should it not be 380VDC?

Normal. This gives the inverter some headroom to compensate for various losses at differing loads and different DC supply voltages.

In these types of inverter the DC-DC converter part is usually not regulated (or very loosely regulated), so this voltage tends to go up and down as the battery supply voltage fluctuates. It has to at all times be greater than the peak AC voltage + losses, even at the lower limits of battery voltage.

[FabianLPHD]

Thank you! That's a good explanation.
Now to the point that the IR2110S burned out the High Side. How can happen this without blowing the IGBT?
So VS of the IR2110S is directly connected to one half bridge output. So my logical conclusion is, that from my AC Output a voltage spike killed it. Right or wrong?
The max Voltage that can be there is my 425V, but IR2110S is up to 500V, maybe i should build in an IR2113S, it's up to 600V?

[FabianLPHD]

Hey guys,
yesterday my parts arrived, yeah.
I replaced three completly shorted IGBTs, put the stock inverter controller back in with original IR2110S.
Also i set the DC Voltage to 380V, near VACp of 230Vrms, more headroom for all components.
Now i tested first with a microwave oven with 1000Watts and it worked.
Second my never starting air conditioner started now right away without any trouble.

In conclusion, it's working. Hopefully more than 2 days 

Thanks guys!