# 2.5KVA transformer on square wave 1000hz mains

Started by April 21, 2012
```I just bought a 50/60 hz transformer rated 575VAC pri to 230VAC sec
rated at 2.5KVA on ebay. I remember reading some time ago that
transformer companies rated their transformers to 1000 hz.

I'm needing like a 3KVDC at about 1.5KVA semi regulated supply but
don't like to use lots of mfd's of filtering at higher voltages to get
there. For a 4X1 PA ima building.

So I figure getting there by bridge switching 4 IGBT's at 1000hz
square wave on a 130VDC supply to the 230VAC winding on the
transformer. AT square wave I figure 325 VAC output, from the 575VAC
side.

Running a 10 step ladder multiplier I get 3250VDC. On a half wave
multiplier I figure the last step being hit by 100 hz pulses.

So Am I just spinning my Wheels and stick with 60hz and lots of mfd's
with a autoformer configuration getting more VDC per step and going
with 5 steps? And no easy way of regulation?

73
de N8ZU somewhere in 8 land
```
```On Apr 21, 4:27=A0pm, raypsi <rborow...@gmail.com> wrote:
> I just bought a 50/60 hz transformer rated 575VAC pri to 230VAC sec
> rated at 2.5KVA on ebay. I remember reading some time ago that
> transformer companies rated their transformers to 1000 hz.
>
> I'm needing like a 3KVDC at about 1.5KVA semi regulated supply but
> don't like to use lots of mfd's of filtering at higher voltages to get
> there. For a 4X1 PA ima building.
>
> So I figure getting there by bridge switching 4 IGBT's at 1000hz
> square wave on a 130VDC supply to the 230VAC winding on the
> transformer. AT square wave I figure 325 VAC output, from the 575VAC
> side.
>
> Running a 10 step ladder multiplier I get 3250VDC. On a half wave
> multiplier I figure the last step being hit by 100 hz pulses.

That voltage is very optimistic. Cockcroft walton ladders even at 3
stages give you nowhere near the Vout you'd expect, even on 1mA out. A
10 stager outputting 1.5kVA would be like building a 100ft tower out
of marshmallows.

> So Am I just spinning my Wheels and stick with 60hz and lots of mfd's
> with a autoformer configuration getting more VDC per step and going
> with 5 steps? And no easy way of regulation?
>
> 73
> de N8ZU somewhere in 8 land

I don't see much point using a 2:1 transformer, it doesn't gain much.
Not everyone agrees, but imho once the voltage you're working with
gets evil, isolation from mains becomes of little value, if the V is
ground referenced. 3kV ref ground or neutral is little different.

The most practical way to get 1.5kVA is direct from a transformer, no
multiplier. And the most practical 1.5kVA transformer is going to be a
switcher, not 50/60Hz.

Regulation is most easily obtained by switching the transformer feed
on and off as the required Vout is reached.

NT
```
```On 2012-04-21, raypsi <rborowiak@gmail.com> wrote:

> I just bought a 50/60 hz transformer rated 575VAC pri to 230VAC sec
> rated at 2.5KVA on ebay. I remember reading some time ago that
> transformer companies rated their transformers to 1000 hz.

IIRC tape wound toroidial transformers are often good to 1000Hz,

> I'm needing like a 3KVDC at about 1.5KVA semi regulated supply but
> don't like to use lots of mfd's of filtering at higher voltages to get
> there. For a 4X1 PA ima building.

microwave oven transformers?

--
&#9858;&#9859; 100% natural

--- Posted via news://freenews.netfront.net/ - Complaints to news@netfront.net ---
```
```NT wrote:
> On Apr 21, 4:27 pm, raypsi <rborow...@gmail.com> wrote:
>> I just bought a 50/60 hz transformer rated 575VAC pri to 230VAC sec
>> rated at 2.5KVA on ebay. I remember reading some time ago that
>> transformer companies rated their transformers to 1000 hz.
>>
>> I'm needing like a 3KVDC at about 1.5KVA semi regulated supply but
>> don't like to use lots of mfd's of filtering at higher voltages to get
>> there. For a 4X1 PA ima building.
>>
>> So I figure getting there by bridge switching 4 IGBT's at 1000hz
>> square wave on a 130VDC supply to the 230VAC winding on the
>> transformer. AT square wave I figure 325 VAC output, from the 575VAC
>> side.
>>
>> Running a 10 step ladder multiplier I get 3250VDC. On a half wave
>> multiplier I figure the last step being hit by 100 hz pulses.
>
> That voltage is very optimistic. Cockcroft walton ladders even at 3
> stages give you nowhere near the Vout you'd expect, even on 1mA out. A
> 10 stager outputting 1.5kVA would be like building a 100ft tower out
> of marshmallows.
>

Sorry, but there I have to disagree. I personally built a three-stage
Cockroft-Walton that delivered a solid 2kW plus. It could deliver a lot
more but the plates in my tubes would have been glowing white instead of

In the end what matters in a Cockroft-Walton cascade is the size of the
diodes and the size of the caps. I stepped up from 230VAC to 900VDC and
it held that rail like a rock. IIRC I put around a dozen 470uF caps in
there. That whole area had about the volume of half a shoe box.

>
>> So Am I just spinning my Wheels and stick with 60hz and lots of mfd's
>> with a autoformer configuration getting more VDC per step and going
>> with 5 steps? And no easy way of regulation?
>>
>> 73
>> de N8ZU somewhere in 8 land
>
> I don't see much point using a 2:1 transformer, it doesn't gain much.
> Not everyone agrees, but imho once the voltage you're working with
> gets evil, isolation from mains becomes of little value, if the V is
> ground referenced. 3kV ref ground or neutral is little different.
>
> The most practical way to get 1.5kVA is direct from a transformer, no
> multiplier. And the most practical 1.5kVA transformer is going to be a
> switcher, not 50/60Hz.
>

If you have the \$\$\$, sure. The other way would nowadays be a switcher.
That makes the transformer really small but you have to roll your own
which isn't for the faint of heart. Plus EMC will be very critical
because it seems this is for a ham radio station.

> Regulation is most easily obtained by switching the transformer feed
> on and off as the required Vout is reached.
>

The utility isn't going to like that :-)

--
Regards, Joerg

http://www.analogconsultants.com/
```
```raypsi wrote:
> I just bought a 50/60 hz transformer rated 575VAC pri to 230VAC sec
> rated at 2.5KVA on ebay. I remember reading some time ago that
> transformer companies rated their transformers to 1000 hz.
>
> I'm needing like a 3KVDC at about 1.5KVA semi regulated supply but
> don't like to use lots of mfd's of filtering at higher voltages to get
> there. For a 4X1 PA ima building.
>
> So I figure getting there by bridge switching 4 IGBT's at 1000hz
> square wave on a 130VDC supply to the 230VAC winding on the
> transformer. AT square wave I figure 325 VAC output, from the 575VAC
> side.
>
> Running a 10 step ladder multiplier I get 3250VDC. On a half wave
> multiplier I figure the last step being hit by 100 hz pulses.
>
> So Am I just spinning my Wheels and stick with 60hz and lots of mfd's
> with a autoformer configuration getting more VDC per step and going
> with 5 steps? And no easy way of regulation?
>
> 73
> de N8ZU somewhere in 8 land
Driving the 230V winding at 1KHz will allow a much higher voltage;
maybe 600V or so; output will still be about 2.5 times higher.
Autoformer mode would give 3.5 times.

Naturally, lower value capacitors are needed for filtering on the
secondary at the higher frequency.

Continuous (sine?) drive allows standard diode voltage multiplier scheme.
Driving with a pulse gives something like a pulse at the output,
meaning you might try a 50V-200V supply and a FET to switch, like a
flyback, giving maybe 1-2KV peak primary and autoformer mode 3.5 times
that (roughly).
Regulate by varying the primary amplitude.

```
```"raypsi"  wrote in message=20

> I just bought a 50/60 hz transformer rated 575VAC pri to 230VAC sec
> rated at 2.5KVA on ebay. I remember reading some time ago that
> transformer companies rated their transformers to 1000 hz.

Toroid transformers will easily handle 1000 Hz, and possibly as high as =
15=20
kHz, as I have demonstrated. When you go up in frequency, you can also =
go up=20
in voltage (and wattage) by a proportional amount. So you could use a=20
150-300 VA toroid transformer with a 24 VAC secondary and 240 VAC =
primary.=20
Create a 300 VDC bus from the 220 VAC line, and then drive MOSFETs or =
IGBTs=20
at 1 kHz into the secondary. The primary should be a 2200 VAC square =
wave=20
which you can easily rectify and filter to get the 2500 VDC you want. =
You=20
could also use a center tapped version and use a push-pull drive. I'm =
doing=20
that with a PIC16F684.

> I'm needing like a 3KVDC at about 1.5KVA semi regulated supply but
> don't like to use lots of mfd's of filtering at higher voltages to get
> there. For a 4X1 PA ima building.

The 1 kHz reduces the filtering requirements.

You might have problems with an ordinary line voltage rated toroid or =
other=20
type transformer when running at 3000V. The insulation might be rated at =
4kV=20
if it's a very good transformer, but that is just a surge breakdown =
rating=20
and not designed to be continuous. You may need to make the transformer=20
using high voltage wire. At 1 kHz you can probably get 5 volts/turn, so=20
you'd need several hundred turns (as is typical of small 220V power =
toroid=20
primaries).

A better idea might be a microwave oven transformer. Maybe two in =
parallel=20
if they are identical. Or two in series if you can use a +/-1500V =
supply.

> So I figure getting there by bridge switching 4 IGBT's at 1000hz
> square wave on a 130VDC supply to the 230VAC winding on the
> transformer. AT square wave I figure 325 VAC output, from the 575VAC
> side.

> Running a 10 step ladder multiplier I get 3250VDC. On a half wave
> multiplier I figure the last step being hit by 100 hz pulses.

> So Am I just spinning my Wheels and stick with 60hz and lots of
> mfd's with a autoformer configuration getting more VDC per
> step and going with 5 steps? And no easy way of regulation?

You might look at this website for high voltage stuff. But mostly they =
have=20
lower power sources like 10-50 mA. They do sell 5 kVA "pole Pigs" that =
are=20
14.4 kV, but they are about \$700. However, surplus supply shops and eBay =

might have something.
http://www.amazing1.com/transformers.htm

Good luck, and be careful. I'm almost ready to finish my DC-DC converter =

which should provide 300-350 VDC at 5 amps from 12V, 24V, or 36V =
batteries.

Paul=20

```
```"P E Schoen"  has gone INSANE  !!!
-----------------------------------------

Toroid transformers will easily handle 1000 Hz, and possibly as high as 15
kHz, as I have demonstrated. When you go up in frequency, you can also go up
in voltage (and wattage) by a proportional amount.

** That  UTTER  NONSENSE was PROVED wrong here months ago  !!

While SOME increase in power throughput is possible, it is nothing like
proportional at 20 times.

So you could use a
150-300 VA toroid transformer with a 24 VAC secondary and 240 VAC primary.
Create a 300 VDC bus from the 220 VAC line, and then drive MOSFETs or IGBTs
at 1 kHz into the secondary.

**  COMPLETE  INSANITY.

The insulation would  FAIL  in seconds  !!!!!

The primary
should be a 2200 VAC square wave ...

**  Sppplaatttttt  !!!!!

The 1 kHz reduces the filtering requirements.

** There is virtually NO  filtering needed with a rectified square wave.

You might have problems with an ordinary line voltage rated toroid or other
type transformer when running at 3000V.

** No fooling  ................

The insulation might be rated at 4kV if it's a very good transformer,

** FFS  - WANKER  !!!!

Even if that figure is quoted - it is for 1 minute and from primary
to secondary  ONLY  !!!!
-------------------------------------------

but that is just a surge breakdown rating
and not designed to be continuous. You may need to make the transformer
using high voltage wire.

**  REALLY  !!!

Hundreds of turns of test instrument lead  ????

A  better idea might be a microwave oven transformer. Maybe two in parallel
if they are identical.

** Call the guys in white suits with the straitjackets  -   NOW  !!

( rest of this puke's drivel snipped  - cos it was making me too nauseas )

.....    Phil

```
```P E Schoen wrote:
> "raypsi"  wrote in message
>
>> I just bought a 50/60 hz transformer rated 575VAC pri to 230VAC sec
>> rated at 2.5KVA on ebay. I remember reading some time ago that
>> transformer companies rated their transformers to 1000 hz.
>
> Toroid transformers will easily handle 1000 Hz, and possibly as high as
> 15 kHz, as I have demonstrated. ...

Even normal EI core transformers of good quality do. I have one from the
60's here that I use to test aerospace designs at 400Hz. 15kHz is a
stretch though even for a toroid. Must be a really expensive one :-)

>                        ... When you go up in frequency, you can
> also go up in voltage (and wattage) by a proportional amount. So you
> could use a 150-300 VA toroid transformer with a 24 VAC secondary and
> 240 VAC primary. Create a 300 VDC bus from the 220 VAC line, and then
> drive MOSFETs or IGBTs at 1 kHz into the secondary. The primary should
> be a 2200 VAC square wave which you can easily rectify and filter to get
> the 2500 VDC you want. You could also use a center tapped version and
> use a push-pull drive. I'm doing that with a PIC16F684.
>

Careful with that "proportional amount". There comes a point where the
insulation parameters aren't good enough anymore and ... tzzzzt ... *POOF*

>> I'm needing like a 3KVDC at about 1.5KVA semi regulated supply but
>> don't like to use lots of mfd's of filtering at higher voltages to get
>> there. For a 4X1 PA ima building.
>
> The 1 kHz reduces the filtering requirements.
>
> You might have problems with an ordinary line voltage rated toroid or
> other type transformer when running at 3000V. The insulation might be
> rated at 4kV if it's a very good transformer, but that is just a surge
> breakdown rating and not designed to be continuous. You may need to make
> the transformer using high voltage wire. At 1 kHz you can probably get 5
> volts/turn, so you'd need several hundred turns (as is typical of small
> 220V power toroid primaries).
>

Insulation can also age with prolonged exposure to voltages at the
design limit.

> A better idea might be a microwave oven transformer. Maybe two in
> parallel if they are identical. Or two in series if you can use a
> +/-1500V supply.
>

Except they aren't rated continuous duty like Ray most likely needs.
They are often only rated to get a bowl of top ramen to cooking
temperature :-)

[...]

--
Regards, Joerg

http://www.analogconsultants.com/
```
```On Apr 21, 11:27=A0am, raypsi <rborow...@gmail.com> wrote:
> I just bought a 50/60 hz transformer rated 575VAC pri to 230VAC sec
> rated at 2.5KVA on ebay. I remember reading some time ago that
> transformer companies rated their transformers to 1000 hz.
>

You do not remember correctly.  Why would a transformer company rate
their transformers to 1000 hz?    There are a lot of trade offs in
transformer design.  It makes no sense to design a transformer to work
at both 60 hz and 1000 hz.  To work well it 1000 hz, you need to use
thinner laminations.  Designing for 60 only will let you make a
transformer that costs less and can be sold for less money.

This is not to say that one can not use a 60 hz transformer at higher
frequencies.  It will just not work as well as one designed for the
higher frequencies.

Dan
```
```Phil Allison wrote:

> "P E Schoen"  has gone INSANE  !!!
> -----------------------------------------
>
> Toroid transformers will easily handle 1000 Hz, and possibly as high as 15
> kHz, as I have demonstrated. When you go up in frequency, you can also go up
> in voltage (and wattage) by a proportional amount.
>
>
> ** That  UTTER  NONSENSE was PROVED wrong here months ago  !!
>
> While SOME increase in power throughput is possible, it is nothing like
> proportional at 20 times.
>
>
> So you could use a
> 150-300 VA toroid transformer with a 24 VAC secondary and 240 VAC primary.
> Create a 300 VDC bus from the 220 VAC line, and then drive MOSFETs or IGBTs
> at 1 kHz into the secondary.
>
> **  COMPLETE  INSANITY.
>
>   The insulation would  FAIL  in seconds  !!!!!
>
>
> The primary
>  should be a 2200 VAC square wave ...
>
> **  Sppplaatttttt  !!!!!
>
>
> The 1 kHz reduces the filtering requirements.
>
> ** There is virtually NO  filtering needed with a rectified square wave.
>
>
> You might have problems with an ordinary line voltage rated toroid or other
> type transformer when running at 3000V.
>
>
> ** No fooling  ................
>
>
> The insulation might be rated at 4kV if it's a very good transformer,
>
>
> ** FFS  - WANKER  !!!!
>
> Even if that figure is quoted - it is for 1 minute and from primary
> to secondary  ONLY  !!!!
> -------------------------------------------
>
>
> but that is just a surge breakdown rating
> and not designed to be continuous. You may need to make the transformer
> using high voltage wire.
>
> **  REALLY  !!!
>
> Hundreds of turns of test instrument lead  ????
>
>
> A  better idea might be a microwave oven transformer. Maybe two in parallel
> if they are identical.
>
>
> ** Call the guys in white suits with the straitjackets  -   NOW  !!
>
> ( rest of this puke's drivel snipped  - cos it was making me too nauseas )
>
>
>
> .....    Phil
>
>
>
>
And good afternoon to you, too!

Jamie

```