oscillator for driving a transformer

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:

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!

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?

How much current do you want?

This worked pretty well:

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.

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?

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).

scillator1.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

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?

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

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?

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?

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 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...

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?

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...

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.

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).

On a sunny day (Tue, 21 Aug 2012 10:26:37 -0700 (PDT)) it happened panfilero wrote in :

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.

For doing something like that, a circuit like this works well:

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"  wrote in message 
news:d88db390-c12c-4dc4-91f8-7c8000d8973e@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-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!

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...

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?

We use this type of oscillator via a tube to create the

100Khz for the rectifier stack on a 2Mev irradiation unit.

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

Jamie