Static inverter for Gee

[Sean Williams]

Hi all,

I am looking into designing a static inverter to run a Gee system.


The power supply needs to generate a high frequency 80v sine wave (1000cycles) from memory.

My current thought would be to build a large audio type amplifier, fed by a sine wave oscillator.

Efficiency isnt a major concern at present, proof of concept would be a good start.

Im paddling out of my depth here, so constructive comments would be good

I do not want to get involved with building any high frequency transformers....

Cheers
Sean

[Steve_P]

How much power does this thing have to have?

Cheers,

Steve P.

[arjoll]

In terms of audio amps I was thinking something like the good old ETI480 would do the job but it might be a bit low on the voltage. The original recommended transformer was 56V CT which is only going to give you +/-39V DC and I'm not sure how much you could crank the supply voltage up.

Its a 70's Aussie design so if its not common 'over there' I could scan it and send it over.

[Sean Williams]

Steve,

I will have a look at the info for the indicator unit - as yet I do not have the receiver.

Andrew,

Thanks for that - will have a google and see what is about

[GMB]

Unless you really want to be totally authentic you will save yourself a lot of difficulty by at least supplying some of the basics more directly (e.g. heater supply and HT).

I'm not sure about GEE but I know of a power unit which looks comparable and it would need about 500W input, much of it to heat the room I think.

[arjoll]

I've had more of a read on the ETI480 and its probably not going to cope with a continuous 80V sine wave. There's some analysis of the circuit at http://home.alphalink.com.au/~cambie/ETI480/.

[Al (astral highway)]

Originally Posted by Sean Williams

The power supply needs to generate a high frequency 80v sine wave (1000cycles) from memory.

Hi Sean, how about this basic idea? It's an indicative, not an actual design of course, but component count would be low and there's not much that could go wrong here.

I have used a similar design with another tuned stage to drive the MOSFET and at a higher frequency.

[HMV 1120]

http://www.spitfirespares.com/Spitfi...adio%202l.html

Not too sure if this is useful, but this guy has original WW2 GEE powersupply and choke available. He is expensive though and I guess they would need major work to be on the safe side.

[Sean Williams]

GMB - yes, I could do that, but the indicator unit I have is all rigged for high cycles supply - it would mean seroius cannibalism to go with direct supplies (the unit is so immaculate that I will not do that to it!)

Spitfiere spares.... Hmm, not sure what that box is, but I do know it is not the rotary converter for Gee! - I reckon my indicator must be worth a fortune going by his prices!

Al,

Yep, that might work - I would be a bit worried about the ferrite saturating though, as I think we are talking ab out 3-400 watts of power....

[Al (astral highway)]

Originally Posted by Sean Williams

I would be a bit worried about the ferrite saturating though, as I think we are talking ab out 3-400 watts of power....

ha ha, I thought there might be a little surprise there. But a MOSFET half-bridge into air-cored xformer with very close coupling would work, wouldn't it, with this sort of current?

[Ed_Dinning]

Hi Sean, yes you want to avoid using SMPS invertor if possible as they are complicated. (I have designed them for a living to Mil specs). Have a look at what is available in the form of "brick" amplifiers up to this power and consider using 2 of then fed anti-phase in a bridge circuit. This may give you sufficient voltage.
A transformer for 500-1000Va may not be too bad to build as the higher frequency will mean a reduced turns/volt figure. Let me know and I can help you with the design.

Ed

[Danny]

Hi ive been asked about this vary problem from a good friend hew has several of these working and is fed up with the humming his rotary converters make. there is an article comming up in one of his news letters showing the construction of such an inverter. though if well silenced rotary trance former would be the easy option. as cheap generators are floting about . remove the engine and fit a motor then speed it up till you get the 80 cycles. but orig converters do somtimes turn up. Danny

[Danny]

Hi thinking again i wonder if an inverter of the type used on campers could be reved up to produce the required frequency. thay are quite cheap now and a 1kw one will cost about £70 but we would need the diagram for it.
another idea would be to use a simple multivibrator to drive transistors.
though the step up tranceformer would probrably be the biggest problem.
Danny

[wave solder]

Some of the competition (not the cheap boy racer ones) car audio amplifiers are capable of a continuous output of the required voltage, current and frequency. (with a signal injected on the input of course !)

[JHGibson]

It is not exactly a static set up but perhaps you could use a automobile alternator driven by a mains electric motor. You would have to tap off one of the phases before the rectifiers. Even though the alternators are from a 12 volt system, the ac phases will output 110 volts if the rotor is fed with enough DC amps.
John.

[PJL]

I have no idea what the PSU on the unit loooks like but does it really need a good sinewave? A 1KHz sq. wave push-pull driver and some LC filtering might do the trick. Anything else is going to need a big heatsink!...Peter

[Kat Manton]

Hi Sean,

Originally Posted by Sean Williams

My current thought would be to build a large audio type amplifier, fed by a sine wave oscillator.

Not a bad idea IMO.

But as a confirmed lazy wossname who avoids building something if it can be bought for less than the cost of building one, I'd be looking for an old rack-mount PA amp, probably a faulty one as they're usually pretty easy to fix.

A quick calculation suggests that if it's rated at 1600W into 4 ohms, that'll provide the 80V you're after. If it's allegedly proper PA kit then it ought to handle continuous operation, flat-out, without melting. The class-D ones may be the best option if you're thinking of heading down this route as they're fairly efficient, smaller and lighter. Many are two-channel with provision for switching into a bridged configuration too.

Cheers, Kat

[RobinBirch]

This is interesting as I have been repeatedly looking at static inverters for 115V at 400 Hz to power aircraft instruments in gliders (Have to take 12V and convert). Haven't got round to building one yet but the simplest one to give a goodish sine wave that I could design consisted of a PIC running a mark space signal out of an eprom. The mark spaces are chosen to average out to a sine wave and clocked out with the PIC controlling the effective frequency, this allows soft start and so on for the motors in the instruments. This is then used to switch a high current FET driving a high frequency transformer. Then put some smoothing on the output.

It's open circuit in that it's final voltage depends on the load unlike a proper switched mode system that monitors the output voltage but we needed something very simple.

The only issue really is the construction of the transformer which would probably be a ferrite one.

Robin

[Ed_Dinning]

Hi Robin, I have built quasi sine inverters for testing aircraft instrumnets in the past. This uses a stepped wave and gives approx the same RMS current into the instrument as a sine wave supply, so no overheating problems. I used transformers (3 off) of a simiklar size to a DAC90 O/P transformer, so it was reasonably compact. Ferrite transformers, which operate on a lower flux density will be larger.

There were some designs in WW in the 90's for some power inverters up to a few hundred watts at 400Hz.

PM me if you need more details.

Ed

[John H.]

Hi Sean,

I have been thinking along these lines as well for a while and would also prefer to run mine as originally designed.

The starting point, as extracted from the AP, is that it was powered from an engine driven generator (not a dynamotor) which supplied 80 volts at 500 watts (6.25 amps) at a frequency of between 1300 Hz and 2600 Hz depending on engine speed.

This output fed a voltage control panel Type 3 which among other things supplied 24 vdc via a carbon pile regulator to the exciter winding of the generator to keep things constant.
The Indicator unit only requires the high frequency AC for it's valve and CRT heaters, since the associated R1355 receiver provides the HT and EHT from it's own internal PSU also driven by the 80v HF ac.
As previously mentioned the easy way would be to build a conventional mains PSU and carefully disable the HF transformer wiring on both units. Outputs need to be HT at 300 vdc, EHT at 1650 vdc and given the amount of valves in the RF unit/R1355 receiver and Indicator unit, 6.3 volts at 11.25 amps, plus 4 volts at 1 amp for the CRT heater.

Of course, at the other end of things, there needs to be a "Black Box" to simulate the Gee pulse signals. From my (very poor) memory there were A,B,C,D pulses. An "A" master at 500 P.P.S and "B,C,D" slaves at 250 P.P.S in the frequency range 20 Mhz to 85 Mhz depending on the RF unit in use.
Regards, John