Heavy-weight power supply: 1KVA at 115V and 400Hz

[Al (astral highway)]

I need some ideas and help, please.

I know that most people might immediately think of a switched-mode power supply using power MOSFETS for this PSU that needs to comfortably dish out 1KVA (115 v at 400Hz).

But can anyone suggest a robust tube design, including the types of tube I could use? If it's a non-starter, I wont be offended if you point this out.

I'm slightly concerned, from my very limited knowlege of valves used as modulators in AM transmitters, that something with this power handling capacity would be quite physically large.

But I don't know to what extent the ultra-low operating frequency of these tubes might enable a smaller device to be used in a well-designed circuit.

Thanks!

[GMB]

You could build a switched mode power converter using valves but you would need quite a few in parallel to control this much power.

Had you thought about using a motor-generator set?

[Ed_Dinning]

Hi Al, Motor generator set it probably is, you may find some ex-gov equipment that was used for aircraft instrument testing. OR use a nice big audio amp (ex reddiffusion perhaps ) with 100V line output. (or an 8R one and wind a transformer). Well within an audio amp's capabilities. If you need 3 phase it will have to be the M-G set or solid state (I led the project that designed on on the Airbus A400M!)

Ed

[Al (astral highway)]

Originally Posted by GMB

.

Had you thought about using a motor-generator set?

That's one option, but the availability of space, the resultant noise and other factors make that a less attractive option.

I think it would also be a costly alternative.

From what I can see, military grade valves are available in super-abundance via russia and the baltic states, and xformers with KVA+ capacity still come up ex USA. It's just a question of knowing the best circuit to go with.

[Al (astral highway)]

Originally Posted by Ed_Dinning

If you need 3 phase it will have to be the M-G set or solid state (I led the project that designed on on the Airbus A400M!)

Ed

Crikey. Well done indeed. But I hope it's not 3 Ph.

[M0AFJ, Tim]

Originally Posted by astral highway

Crikey. Well done indeed. But I hope it's not 3 Ph.

Al, I'm a bit worried about this, I think you may well find that this beast will be 3 phase, I've not seen a radar set running this sort of power on single phase, hopefully I'm wrong

[GMB]

Originally Posted by astral highway

That's one option, but the availability of space, the resultant noise and other factors make that a less attractive option.

I think it would also be a costly alternative.

You might want to check those assumptions. An MG set might actually be smaller than a pile of valves and transformers and might be buildable from scrap parts at low cost. They don't have to be noisey.

[Al (astral highway)]

Originally Posted by GMB

You might want to check those assumptions. An MG set might actually be smaller than a pile of valves and transformers and might be buildable from scrap parts at low cost. They don't have to be noisey.

Okay, I'm happy to revise my assumptions. Actually I know nothing whatever about MG sets, so please tell me more

[Al (astral highway)]

Originally Posted by M0AFJ

Al, I'm a bit worried about this, I think you may well find that this beast will be 3 phase, I've not seen a radar set running this sort of power on single phase, hopefully I'm wrong

Hi again Ed. I've checked again and there's nothing in the technical spec that mentions 3 Ph. I now own this beast and it should turn up here within a couple of weeks.

It's an RCA AVQ -10 - not the whole system, just the transmitter and receiver.

[Sean Williams]

As much as I hate to be a killjoy - AVQ10 is a 16kW weather radar - with a 75kW peak power.

I would be amazed if it isnt 3 phase - a lot of airborne radar is.

Forget trying to build a converter with valves - it will be too big, inefficient, and hellishly expensive.

Mosfet design makes sense, but it will be big.

Rotaries are the way to go - they are normally noisy, inefficient, and hard to get hold of.

If you are serious (I mean serious) then i do know where you can get BIG 400 cycle converters - you will require a 3ph supply, and a suitable forklift and truck to collect - they were used by De havilland for aircraft testing.

Otherwise, you have a doorstop to play with.

We wont go into the legality of running this up without the correct RF load.......

[Al (astral highway)]

Originally Posted by Sean Williams

If you are serious (I mean serious) then i do know where you can get BIG 400 cycle converters - you will require a 3ph supply, and a suitable forklift and truck to collect - they were used by De havilland for aircraft testing.

I am serious, but I'm not in a position to go to such extremes. The point of acquiring the set in the first place was :

1) to have a close look at the construction (layout) and as-built design in particular

2) to restore it electronically

3) to integrate it with a suitable indicator from the same family or to adapt something else.

It's way beyond the 'project scope' to increase much beyond the minimum to achieve 2). If it needs as 3-phase power supply, as you and others suspect, then I've already learnt something unexpected. And that's the main reason I do what I do.

I'd be perfectly happy if I once I'd finished, I demonstrated the set to be operational.

By this I mean even running it vastly derated and for the duration of a short (milliseconds) pulse train.

Originally Posted by Sean Williams

We wont go into the legality of running this up without the correct RF load.......

Good call, but now we've come full circle. In another post, I asked about suitable dummy loads. But please, tell me more. What should I be aware of here?

[wave solder]

What may work is a 3 phase motor inverter as quite a lot of these can be programmed to run up to 400 Hz and you can also program the output voltage. The problem with this of course the input section of the equipment must be an inductive load for the inverter to run correctly.

[Al (astral highway)]

Originally Posted by wave solder

What may work is a 3 phase motor inverter as quite a lot of these can be programmed to run up to 400 Hz and you can also program the output voltage. The problem with this of course the input section of the equipment must be an inductive load for the inverter to run correctly.

Hi there -- another good idea, for which many thanks.

Cheers

[wave solder]

Just found this link with the full specs of the system on it:-

http://www.smecc.org/rca_avq-10.htm

[Ed_Dinning]

Hi Al, some sideways thoughts - dispense with the 400Hz and 3 phase, go round the power input stages and feed the appropriate levels of DC into the supply rails of the beast. These supplies might be a bit easier to design (but you still may need Penguins MAR!).

Ed

[Al (astral highway)]

Originally Posted by Ed_Dinning

Hi Al, some sideways thoughts - dispense with the 400Hz and 3 phase, go round the power input stages and feed the appropriate levels of DC into the supply rails of the beast. These supplies might be a bit easier to design (but you still may need Penguins MAR!).

Ed

Yes, that's a good thought, Ed, for which Skywave (scroll down here...
https://www.vintage-radio.net/forum/...ad.php?t=22849

...also gets credit. Thanks both. It would save an enormous headache!

Handily enough, I do have an ridiculous thyratron of the cold war period and Soviet manufacture, lurking in the cellar, that I was thinking of designing the 400Hz supply around.

It would be wasted as a simple half-wave rectifier, of course, and what's the point when silicon bridges are so cheap?

[Alf Fisher]

Hi all,
Has anyone seen the design in Electronics World Feb 99 issue by Paul Bennett. Its 400Hz and 3 phase but not sure if it fits the bill in other respects.
Alf

[Al (astral highway)]

Thanks Alf, that's very helpful.

Actually we're going with another solution now - basically to bypass the power xformer and smoothing circuit and provide DC from an external source. The timebase will be supplied by a separate low-power 400Hz oscillator.

Thanks to everyone who contributed to this solution!

Mods, please close this thread.

[Sean Williams]

After discussion with the other moderators it has been decided that this topic does not really fit in with the forums aims.

Thread closed