Inverter back in business

[frackers]

Those who have been following the saga of getting everything working only to have the inverter fail may be pleased to know that its back up and running again. Two problems were sorted out - a pair of MOSFETs in the output H bridge were short circuit and the crimps on the internal connectors were done all wrong during manufacture.

The original eight 500volt 14amp irfp450 MOSFETs were replaced by STY34NB50 which are rated at 500volt 34amp. Higher drive requirements (twice the gate/source capacitance) but much lower ON resistance so should run cooler. So far so good!!

The internal connector lugs were all recrimped and soldered with a blowtorch and heatshrink sleeving applied where some of the cable insulation shrank back.

Very interesting having a good look inside - there are two main PCBs, one with 1kw of DC-DC converters (4 sections) and the H-bridge MOSFET output stage and the other main PCB with 8 DC-DC converters for the other 2kw of capability. A display PCB for one end with V & A bar LEDs and a fan at the other end and thats it!! The PCBs fused on each DC-DC converter with labeling that indicates that the same PCB is used for both 12 and 24 volt variants of the inverter.

[Ntrain2k]

What kind of invertor was it?

Have any pictures of the work?

Thanks..

[frackers]

Its a generic Chinese 24v 3kw modified sine wave inverter, 6kw peak capable (allegedly) with remote control. This is where is came from which gives a few clues but I didn't take any pictures whilst it was in bits apart from the wires that had fallen out of the crimp lug.

[frackers]

Still got problems - tractored the battery bank from the paddock to the garage so I could do some testing last night and the inverter is putting a LOT of noise into the battery leads - looks like one of the DC-DC converters is failing...

I'll be pulling fuses tonight until I identify the section that has gone bad (and hope that its only one!!).

I'll try and remember to take some pictures!!

[frackers]

Looks like a false alarm...

After pulling fuses without making much of an impression on the noise across the shunt and playing with filtering on the input to the battery monitor chip all the problems went away when I put it back as it was as I gave up in disgust!!

The current measurement problem seems to be with one of the bullet connectors in the sense leads - the noise issue is a total red herring.

So tonight its back into the paddock with the battery bank and get some power flowing again :-)

I hope to be publishing the Linksys router software for the 1-wire controller over the weekend, depends on whether the boss has any more jobs for me...