oscillator for driving a transformer

[panfilero]

Hello,

I'd like to drive a small high frequency transformer at around 100kHz-300kHz with a 24Vpk signal, I tried building up an astable multivibrator with a couple of caps and transistors but i couldn't get it to drive the transformer, it wouldn't even make anything close to a nice square wave.... anyone know of a better simple oscillator circuit I could use to do this?

thanks!

[Tim Williams]

Yeh, what's *on* the transformer? Not much point in driving a transformer
unless ya want it for something...

If you have/can put a CT on the transformer (or use a CT'd inductor), you
can use the classic chopper circuit. Even better, if you can add a
feedback winding (if you have to do it with a CT'd inductor, it still
works), you don't need as many resistors and as much bias current.

No caps, if you try making it resonant (or quasi-) you need a choke in
series with the supply voltage. Lack of one causes crappy waveforms.

Tim

--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms

"panfilero" <panf...@gmail.com> wrote in message
news:d4d20e3c-816c-4172...@googlegroups.com...

[John Larkin]

How much current do you want?

This worked pretty well:

https://dl.dropbox.com/u/53724080/Circuits/ESM/ESM_power_B.pdf

[Bruce Varley]

"John Larkin" <jjla...@highNOTlandTHIStechnologyPART.com> wrote in message
news:jt563851ku95ttpul...@4ax.com...

Half bridge drivers are ideal for this, they can handle beanpole voltages up
to the hundreds, at frequencies in the 100s of KHz, they can be driven with
a 5V logic signal, and (at least with the device that I built using them)
they avoid problems from simultaneous turnon of the two power devices.

For a simple oscillator, you can't go past a schmitt CMOS inverter, R from
output to input nad C from input to ground. If duty cycle has to be exactly
50%, just put it through a flipflop divider.

[Bill Sloman]

On Aug 21, 6:01 am, panfilero <panfil...@gmail.com> wrote:
> Hello,
>

> I'd like to drive a small high frequency transformer at around 100kHz-300kHz with a 24Vpk signal, I tried building up an astable multivibrator with a couple of caps and transistors but I couldn't get it to drive the transformer, it wouldn't even make anything close to a nice square wave.... anyone know of a better simple oscillator circuit I could use to do this?

Peter Baxandall's class-D oscillator produces a tolerable nice sine
wave, typically with about 1% harmonic content if you can keep the Q
of the tank circuit above about 5, and it's happy as long as the
frequency of oscillation is below the self-resonant frequency of the
transformer (which can easily be less than 100kHz if you aren't
careful).

http://home.planet.nl/~sloma000/0344_001_Baxandal.pdf

http://home.planet.nl/~sloma000/Baxandall%20parallel-resonant%20Class-D%20oscillator1.htm

As you can see, it can be a pretty simple circuit - one centre-tapped
transformer, one capacitor, one inductor and two MOS-FET's.

If you really want a square wave, a Royer inverter can be even simpler
- no inductor - but at 300kHz the ringing on each switching-edge cans
be nasty.

--
Bill Sloman, Nijmegen

[Phil Hobbs]

Nice. You get a lot of mileage out of those little ISDN transformers.
Is ISDN a good long term bet, though? I'd be starting to worry about
them going away. (Also zeners D6-D8 are mislabelled.)

Cheers

Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510
845-480-2058

hobbs at electrooptical dot net
http://electrooptical.net

[John Larkin]

On Tue, 21 Aug 2012 09:26:59 -0400, Phil Hobbs
<pcdhSpamM...@electrooptical.net> wrote:

>On 08/21/2012 01:10 AM, John Larkin wrote:
>> On Mon, 20 Aug 2012 21:01:28 -0700 (PDT), panfilero
>> <panf...@gmail.com> wrote:
>>
>>> Hello,
>>>
>>> I'd like to drive a small high frequency transformer at around 100kHz-300kHz with a 24Vpk signal, I tried building up an astable multivibrator with a couple of caps and transistors but i couldn't get it to drive the transformer, it wouldn't even make anything close to a nice square wave.... anyone know of a better simple oscillator circuit I could use to do this?
>>>
>>> thanks!
>>
>> How much current do you want?
>>
>> This worked pretty well:
>>
>> https://dl.dropbox.com/u/53724080/Circuits/ESM/ESM_power_B.pdf
>>
>>
>>
>Nice. You get a lot of mileage out of those little ISDN transformers.
>Is ISDN a good long term bet, though? I'd be starting to worry about
>them going away. (Also zeners D6-D8 are mislabelled.)
>

The original design had three 25 volt zeners, so the energy sensor
photodiode could be selected to run at 25, 50, or 75 volts, which made
sense from the data sheet. The customer claims that they will get
breakdown in the sensor wiring (1 torr hydrogen atmosphere) so made me
drop the selections to 5 and 10 volts. Of course, the photodiode
capacitance is nanofarads at that voltage, slowing down production at
what may be gigabuck consequences. As a concession, they allowed me to
include the 15 volt range, under the condition that it require a
32-bit password to enable it, and that they not know the password.

So far, no problem getting ISDN transformers. We have some custom made
for us, less winding capacitance, and they don't cost much more than
standard ones. There will probably be zillions of ISDNs available
forever, if just overstock.

They are handy little transformers.

Does anyone still use ISDN? European telephones?

[panfilero]

Thanks for the replies, I'm basically trying to turn 24V into 72V using a voltage tripler, then I was gonna send that into a voltage regulator to clean it up, the load is only going to draw maybe like.... 5mA.... so I was thinking, take the 24Vdc, oscillate it somehow, feed it to a transformer and then triple it...

[John Larkin]

That will work. Or build a simple boost converter. If you don't need
isolation, that takes a single inductor.

Or buy a couple of SIP dc/dc converters, 24-to-24. Stack their outputs
on top if the 24 input and get 72, likely a bit more at light loads.
The SIPs are cheap, like $4.

[Jan Panteltje]

On a sunny day (Tue, 21 Aug 2012 10:26:37 -0700 (PDT)) it happened panfilero
<panf...@gmail.com> wrote in
<d88db390-c12c-4dc4...@googlegroups.com>:

>On Monday, August 20, 2012 11:01:28 PM UTC-5, panfilero wrote:
>> Hello,
>>
>>
>>

>> I'd like to drive a small high frequency transformer at around 100kHz-300=
>kHz with a 24Vpk signal, I tried building up an astable multivibrator with =
>a couple of caps and transistors but i couldn't get it to drive the transfo=
>rmer, it wouldn't even make anything close to a nice square wave.... anyone=

> know of a better simple oscillator circuit I could use to do this?
>>
>>
>>
>> thanks!
>

>Thanks for the replies, I'm basically trying to turn 24V into 72V using a v=
>oltage tripler, then I was gonna send that into a voltage regulator to clea=
>n it up, the load is only going to draw maybe like.... 5mA.... so I was thi=
>nking, take the 24Vdc, oscillate it somehow, feed it to a transformer and t=
>hen triple it...

For that sort of thing, get a small ferroxcube E core,
and use 1 V / turn, 3 windings, one feedback.
Diagrams have been published here.
No tripler, 1 transistor, 1 tuning capacitor,
1 resistor, 2 decoupling caps, 1 rectifier diode, 1 rectifier cap.
You can do the feedback in the base or in the emittor.

[whit3rd]

On Monday, August 20, 2012 9:01:28 PM UTC-7, panfilero wrote:

> I'd like to drive a small high frequency transformer at around 100kHz-300kHz with a 24Vpk signal, I tried building up an astable multivibrator with a couple of caps and transistors but ...

Transformers (or any inductor) really, REALLY misbehave if you give them any DC
component on their drive. So, a symmetric drive signal is key. Easiest
would be to use a SMPS chip (TL494 will almost work, there's LOTS of 'em to choose
from) that provides two drive transistors, and connect a center-tap on the
transformer primary to your +24V power.

The collector voltage (with 24V on the center tap) can rise to 48V on the un-driven
leg of the transformer; you'll want that much voltage rating on
the drive transistors (the internal transistors of the TL494 max out at 40V,
so you'd have to use external transistors).

[Tim Williams]

For doing something like that, a circuit like this works well:
http://t3sl4.dnsdynamic.net/Images/Deadbug_Sch.png
Notes:
- The base drive on the left-hand side drops a lot of voltage, so it
starts getting impractical over 12V. This is most efficient for low
voltages, where Vbe is relatively large.

- Your step-up ratio isn't very much, so you could skip the tapped
inductor and doubler, and just use a boost setup. This is only a matter
of the inductor and output rectifier; the rest of the circuit remains the
same.

- Note that this circuit was built for current limited drive, because the
output has to charge capacitors as shown. In boost mode, this goes away,
and the switch will saturate efficiently.

- For more power, the current limiting resistor (2.2 ohms) will have to be
decreased, of course. Maybe a higher power switching transistor as well.

- A MOSFET can be used for the main switch, just change the bias
components a bit (i.e., short out the 100 ohm resistors so the 3904 and
3906 fight each other for gate voltage) and add a complementary emitter
follower to drive the gate capacitance. At 24V supply, this will probably
work fine, but for safety margin, a voltage limiting zener should be used
to keep gate voltage below 15V or so.

- Compensation components may need to be changed, of course.

Tim

--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms

"panfilero" <panf...@gmail.com> wrote in message

news:d88db390-c12c-4dc4...@googlegroups.com...

[Jamie]

We use this type of oscillator via a tube to create the
100Khz for the rectifier stack on a 2Mev irradiation unit.

http://en.wikipedia.org/wiki/Armstrong_oscillator

You may also want to look at a basic blocking oscillator circuit.

http://en.wikipedia.org/wiki/Blocking_oscillator

Jamie