From Batteries to High Power HV Methods

[Farmhand]

Hi all, I am finding myself in the position where I want to use some reasonably high power for various projects and experiments, however I don't want to have to use the grid to for example run a Tesla Coil at 1000 Watts input for fun.

I have a few options as I see it, I've got a 1000 watt (2000 surge) inverter that failed without any of the protections shutting it down, it has a pair of HF transformers with push pull primaries, each one should be able to handle 500 Watts at the appropriate frequency and output 240 v or more, so the first stage of the plan is to make a small DC to DC converter that has a three wire output ( maybe two separate two wire outputs can have the outputs stacked), a simple oscillator - driver a pair of mosfets for each transformer input and a 3 Amp FR bridge rectifier at the output and I could have a 500 Watt 240 volt DC output from each one which I could stack for a three wire supply for half bridge switching or I could parallel them for Full H Bridge switching. Of course the FWBR's would charge appropriate capacitors but not necessarily in the actual power supply, on the board with the switching circuit maybe. These little transformers are only about 30 mm x 50 mm x 50 mm so quite small.

Now a question, if I power two separate transformers from the same battery could it be more problematic than using a separate battery for each transformer since the outputs might be stacked ? I figured that the higher inverter output would exert more stress on it's primary windings and circuit due to being double the voltage if powered from the same battery. Or would it be OK to run two inverters from the same battery and stack the two 240 volts FWBR outputs ? I have four appropriately large identical batteries.

I'll post an image of the destroyed inverter circuitry, it seems one of the IRF460 mosfets at the output stage failed (the gate resistor is smoked and the driver chip exploded) then it would seem that the transformers went ballistic trying to keep supplying power and melted the IRF1404 mosfets switching them along with one of the MUR TO220 diodes. Some other resistors are smoked and another chip blew out ect., but the transformers look ok although I have not tested them yet or removed them from the board.

It used IRFP460's and IRF460's in the output H bridge. Some other good parts are still oK and I have taken some bits already. 

The idea is to build a power supply so that if it does fail I can fix it/upgrade it, I hope to ensure no operational failures by using simple protections.

With that done I would then look at using the power supplies to run HV transformers and such fun things with lots of input power of my own harnessing.

Two identical stackable stand alone units would be more useful but require more parts.

Or I could use a large center tapped steel core transformer at 400 Hz or something, but I doubt the power handling capability of the ones I have and they're too darn big.

Cheers

   

[Farmhand]

OK so below is a picture of the old inverter board with the two transformers on the left at the input side. They are 45 mm x 30 mm x 50 mm high.


Stage 1)

Also below is a simplified drawing of how I intend to build two identical units, I'll probably use an SG3525 on each to provide the alternate signals to a TC4427 dual driver chip and use IRF1010's for the switches, I can make provision for controlling them via the shutdown pin ect. of the SG3525's. A high signal from a micro controller to the two SG3525's would shut the outputs off quickly, similarly in the later stages the same signal can be supplied to the IRS2110 driver chips to turn off the drive to the H Bridges in the event of a fault condition. Stage 1 should be two 500 Watt units combined to make a 1000 Watt - 240 or 480 volt supply

[Farmhand]

Here are some scope shots from the primary of a transformer switched at about 400 Hz, first by a half H bridge (first two shots) then the same transformer switched by a full H bridge ( I used 24 volts and two primaries in series for the full H bridge). I decided to prototype a full H bridge since I just tried a HHB and see any differences before I put the proto-board away. It's a small 20 VA toroid transformer with a 220 uF load which charges in a flash of the current needle, idle input power is very low of course.

To me the full H bridge looks better, but maybe much the same. Dead time is about 22 or so microseconds as can be easily seen in the first shots. I think I've worked out the parts list, I need a few things.

..

[Farmhand]

Ok so I got the transformers off, I did it by cutting them off the PCB with a skinny disk on a grinder and kept cutting so that each pin had just a tiny square of PCB on it and they came off real easy one by one then with the soldering iron.

Looking at the transformers one of them I'll call "T1" has two secondaries, one is the main output winding (about 4 mH) and the other winding (60 uH) is for I think an isolated supply for the output control and switching supply.

I think I've drawn the drawing below as the power circuit is on the old inverter from the transformer to the H bridge supply, although I only drew one transformer "T1" the inverter has two transformers but the other transformer "T2" has only the main output winding and it measures (about 3.2 mH). There is only four diodes for the two transformers so I imagine they must be synced and in parallel, but I can't tell now since that part of the PCB is in many pieces, haha.

Across the output of the diode bride is a 10 nF 600v cap in series with some big resistors measuring 1.3 KOhm and along with the 800 uH inductor would that section in the dashed box be a filter or something ?

I can't see any visible heat damage to the transformers so I'm wondering why they vary in inductance so much.

Any idea's ? Can anyone tell by the components the approximate frequency the transformer should be operated at ? 60 kHz ? 30 kHz ? 120 kHz ? I'll guess about 60 kHz. 

I also drew how it looks as though the output section power parts go in the old inverter. No that important for now.

..