High frequency, high voltage transformer

[Harshana]

Hi all,

I want to build a transformer which runs on switched DC (rectified
mains). Here are the pecs.

switching frequency 1kHz ~ 50kHz

output winding 2kV (100mA).

input winding 300Vx2 (cetre tapped winding)

Will it be easier for me to wind my own transformer,

in which case, what are the windings turns, wire guages, air core
(preffered) / ferrite core?

or

Can I use some thing like a TV flyback transformer, motor bike ignition
coil, or something else...

Please help...

Thanks,
Harshana

 

[Tim Shoppa]

Hi all,

I want to build a transformer which runs on switched DC (rectified
mains). Here are the pecs.

switching frequency 1kHz ~ 50kHz

output winding 2kV (100mA).

input winding 300Vx2 (cetre tapped winding)

Will it be easier for me to wind my own transformer,

in which case, what are the windings turns, wire guages, air core
(preffered) / ferrite core?

or

Can I use some thing like a TV flyback transformer, motor bike ignition
coil, or something else...

Your specs (200W, 7:1 voltage ratio, AF to supersonic frequency range)
are most similar to a VERY LARGE audio output transformer.

It may also be described as a "modulation transformer".

One very strange thing for a switching supply is your large desired
range of operating frequencies. Your range doesn't exactly match up
with audio output transformers but it's only off by a factor of 2 or 3.
If you try to run a typical audio transformer up that high you'll
probably be hitting a resonance of some kind (which may in fact be
desirable for your operation, I dunno!!!)

If you can aim for a substantially lower fixed frequency, you may find
a circa 400Hz unit commercially.

Hint about sizing transformers: TV flybacks and ignition transformers
are certainly not in the 200W class you're aiming for.

If you want circa 2-3kV DC at in the 200W range, there are MUCH easier
ways of doing this that are straight out of a textbook or ARRL
handbook.

Tim.

 

[colin]

Hi all,

I want to build a transformer which runs on switched DC (rectified
mains). Here are the pecs.

switching frequency 1kHz ~ 50kHz

output winding 2kV (100mA).

input winding 300Vx2 (cetre tapped winding)

Will it be easier for me to wind my own transformer,

in which case, what are the windings turns, wire guages, air core
(preffered) / ferrite core?

or

Can I use some thing like a TV flyback transformer, motor bike ignition
coil, or something else...

Is this a power supply or some other application, the 1khz is very low for a
smps.
otherwise wouldnt be too hard to do from stocked E cores or from say a scrap
pc power supply.
some cores wich arnt glued come apart ok.
air core will be a bit large.

You just need to follow the formulea wich gives number of turns based on
cross sectional area of ferrite and frequency,
number of Turns N, Core cross sectional area A (m^2) RMS Voltage V, and
Saturation flux density B (use 0.15 to 0.2 Tesla for most ferrites) :

V = 4.44.F.N.A.B

wire diameter to give a reasonable amps/sq cm.

You can seriously kill yourself with this voltage/current though.

Colin =^.^=

 

[Winfield Hill]

Harshana wrote...

I want to build a transformer which runs on switched
DC (rectified mains). Here are the pecs.
switching frequency 1kHz ~ 50kHz
output winding 2kV (100mA).
input winding 300Vx2 (cetre tapped winding)

Will it be easier for me to wind my own transformer, in
which case, what are the windings turns, wire guages,
air core (preffered) / ferrite core?

or

Can I use some thing like a TV flyback transformer,
motor bike ignition coil, or something else...

I haven't gotten into scavenging household electronics,
as some have, but except for the primary center tap,
those 'pecs look much like a microwave-oven transformer.
Look up half-bridge and H-bridge driving circuits to see
why a center-tapped transformer isn't necessary or wise.

 

[Homer J Simpson]

I want to build a transformer which runs on switched DC (rectified
mains). Here are the specs.

switching frequency 1kHz ~ 50kHz

output winding 2kV (100mA).

input winding 300Vx2 (cetre tapped winding)

Why? What is it supposed to do?

 

[Harshana]

Hi all,

Its a small experiment on electrolyzing water.

I want to apply 2kV AC across two plates and measure the HHO output
with frequency in the range 1kHz to 50kHz. (I'm expecting a peak around
20~30kHz)

Thanks,
Harshana

PS:
Now don't mark me as one of those people who's experimenting on over
unity gadgets. I just want to get the frequency response on HHO output
for specific voltages.

 

[Harshana]

Hi all,

Its a small experiment on electrolyzing water.

I want to apply 2kV AC across two plates and measure the HHO output
with frequency in the range 1kHz to 50kHz. (I'm expecting a peak around
20~30kHz)

Thanks,
Harshana

PS:
Now don't mark me as one of those people who's experimenting on over
unity gadgets. I just want to get the frequency response on HHO output
for specific voltages.

 

[jasen]

Hi all,

I want to build a transformer which runs on switched DC (rectified
mains). Here are the pecs.

switching frequency 1kHz ~ 50kHz

output winding 2kV (100mA).

200W (well 200VA)

input winding 300Vx2 (cetre tapped winding)

Will it be easier for me to wind my own transformer,

in which case, what are the windings turns, wire guages, air core
(preffered) / ferrite core?

200W air core at only 50kHz ???

Can I use some thing like a TV flyback transformer

too high in voltage, maybe too little power handling capacity (one from a
36" crt might be big enough), but the design frequency 17kHz is about right

motor bike ignition coil,

nowhere near enough power handling capacity.

or something else...

maybe something from a PC powersupply, but you'd need to modify the
secondary...

AIUI they run the primary at around 300V

Bye.
Jasen

 

[Winfield Hill]

Harshana wrote...

Its a small experiment on electrolyzing water.
I want to apply 2kV AC across two plates and measure the

Based on what evidence, do you have any references, etc.
for us to read?

BTW, I still think your best bet is two MOSFETs or IGBTs
in a half-bridge driving an old microwave-oven transformer.
In fact, you can probably use the oven's AC-power wiring
and rectifier, etc. There's a lot of detail on the web.

 

[colin]

Hi all,

Its a small experiment on electrolyzing water.

I want to apply 2kV AC across two plates and measure the HHO output
with frequency in the range 1kHz to 50kHz. (I'm expecting a peak around
20~30kHz)

Thanks,
Harshana

PS:
Now don't mark me as one of those people who's experimenting on over
unity gadgets. I just want to get the frequency response on HHO output
for specific voltages.

You might be able to use an ordinary mains transformer in reverse,
as the min frequency is 20 times higher than line freq you can use 20 times
the voltage,
or to be conservative 10 times therefore a 240v primary can be used as your
2kv secondary,
you can chose whatever secondary to use for you primary eg 24-0-24, for
200vac input,
current ratings will remain unchanged, therefore power rating will be
increased by a factor of 10,
although the higher frequency may cuase more heating,
a 20watt transformer is quite small.

Colin =^.^=

 

[Joerg]

Hello Win,

I haven't gotten into scavenging household electronics,
as some have, but except for the primary center tap,
those 'pecs look much like a microwave-oven transformer. ...

Maybe this shows my age now but when I took apart our first microwave
the transformer was a big old 50/60Hz laminate core. Whenever I tried to
use power transformers at higher frequencies their performance became
rather lousay above a few kHz.

If the new microwaves use switchers those transformers probably won't
work at 1kHz.

 

[Winfield Hill]

Joerg wrote...

Hello Win,


Maybe this shows my age now but when I took apart our first
microwave the transformer was a big old 50/60Hz laminate core.
Whenever I tried to use power transformers at higher frequencies
their performance became rather lousay above a few kHz.

Lousy, yes, but how lousy? I think the o.p. can stand some
losses for his electrolysis experiment. If not, he can let
us know how it works out.

If the new microwaves use switchers those transformers
probably won't work at 1kHz.

Yes. How many old defective microwave ovens do you think
have those? I'm pretty sure my 14-year-old model doesn't.

 

[Joerg]

Hello Win,

Lousy, yes, but how lousy? I think the o.p. can stand some
losses for his electrolysis experiment. If not, he can let
us know how it works out.

Maybe it works. When I was a kid I sometimes used small line
transformers to replace a blown audio transformer in a tube final.
Worked ok but not great. Later when I had some more gear I measured one
and it was completely flat above 10kHz. Then I knew why FM stations
sounded so dull after such "repairs".

Yes. How many old defective microwave ovens do you think
have those? I'm pretty sure my 14-year-old model doesn't.

It's surprising what people toss these days just because the decor or
something doesn't fit. A neighbor chucked a perfectly good TV set that
wasn't even 5 years old because the power cord had gotten whacked during
a move.

 

[jasen]

news:[email protected]...

You might be able to use an ordinary mains transformer in reverse,
as the min frequency is 20 times higher than line freq you can use 20 times
the voltage,
or to be conservative 10 times therefore a 240v primary can be used as your
2kv secondary,
you can chose whatever secondary to use for you primary eg 24-0-24, for
200vac input,
current ratings will remain unchanged, therefore power rating will be
increased by a factor of 10,
although the higher frequency may cuase more heating,
a 20watt transformer is quite small.

the insulation might not last real long with that sort of abuse.

Bye.
Jasen

 

[JoeBloe]

200W (well 200VA)


200W air core at only 50kHz ???


too high in voltage, maybe too little power handling capacity (one from a
36" crt might be big enough), but the design frequency 17kHz is about right


nowhere near enough power handling capacity.


maybe something from a PC powersupply, but you'd need to modify the
secondary...

AIUI they run the primary at around 300V

Bye.
Jasen

The bigger question is what is the application. I'd like to know
just what he thinks he needs 100mA for.

You want a ferrite core. You want around 57kHz.

Also, after generation at the transformer secondary, one needs
rectification and storage to make DC, as well as current limiting and
arc suppression considerations.

If it is simply AC output that is desired, a lower frequency and
other than ferrite could be used. Again, however, it depends on the
application involved.

 

[JoeBloe]

Harshana wrote...

Based on what evidence, do you have any references, etc.
for us to read?

BTW, I still think your best bet is two MOSFETs or IGBTs
in a half-bridge driving an old microwave-oven transformer.
In fact, you can probably use the oven's AC-power wiring
and rectifier, etc. There's a lot of detail on the web.

Excellent suggestion. A pair of FETs and an oscillator setup makes
a fine input stage driver (though not likely at 300 Volts). Likely
nice and cheap cost wise as well. The transformer you suggest is a bit
overkill though for his needs in both size and capacity, but it would
have the isolation and would function properly.