Hi everyone: I have been designing a circuit that will output a very low level pulse. The pulse train is two pulses, with each pulse at 10 mV for 50 uSec, and -10V for 50 uSec. This is repeated every milliSec. The outpus needs to be isolated, which (at least in my current attempts) has employed a transformer. I have all the drive and controls worked out and prototyped. I can't seem to get a good pulse across the transformer. The circuit runs off a single AAA alkaline cell with a boost controller to produce 3.3V. If everything works, I may reduce that to about 2V.
Can some one suggest a way to get to a solution?

Thanks in advance.

 

Can you supply a part number and specs on the pulse transformer?

Have you considered using an opto-isolator?

 

Can you supply a part number and specs on the pulse transformer?

Have you considered using an opto-isolator?

Thanks MrChips:

The problem with an optoisolator is the need for some kind of power on the isolated side. I have none, and can't put one there in this project.

I have tried a number of transformers:
The 100o series from Murata,
The 786 series from Murata,
Miniature SMT transformer T6437-DL from coilcraft.

One of the controlling factors for a pulse transformer is the VuS product, how long it is designed to holdup a particular voltage in uSec (on the primary). Based on the frequency of the square wave I am passing, that would be 50 uSec multiplied by the voltage on the primary.
Also, I am using a center tapped primary on the transformer, with each leg driven to ground by a pulse on a 2N7000/2N7002 transistor.

On the output side, I can resistor divide down to get the voltage levels I need.

I did neglect to mention that the output load is a nominal 600 ohms. I had thought to use an audio transformer, but cannot find enough info in the specs to be able to make a decision.

 

Can some one suggest a way to get to a solution?

Sure. First step is to post your circuit here so we can get a look at it; then maybe we can figure out what's going wrong.

 

What sort of load is on the output pulses? In other words, how much current does it need to sink at -10V?

I did neglect to mention that the output load is a nominal 600 ohms.

Ah, I see. Cross posted.

 

Sure. First step is to post your circuit here so we can get a look at it; then maybe we can figure out what's going wrong.

You are right:
I have posted the output circuit I have been working with. The NPN transistor is currently a 2N3904. BV1 is a AAA alkaline battery. VCC is 3.3V from a step up controller. The transformer CT can be attached to any of these at this time, and will stay attached to the source that actually works.

 

Hmmm... all I can think of, looking at the datasheet for the T6437-DL (as an example) is that perhaps you are exceeding the V*T rating of the transformer; if that's the case, the only remedy I can see (other than finding a transformer with higher V*T) would be to drive the transformer from as low a voltage as possible, and then not attenuate its output so much. You're looking for output pulses of only +/- 10 mV, correct?

 

What sort of load is on the output pulses? In other words, how much current does it need to sink at -10V?



Ah, I see. Cross posted.

I see a typo in my original post. the output goes from +10mV to -10mV

 

Hmmm... all I can think of, looking at the datasheet for the T6437-DL (as an example) is that perhaps you are exceeding the V*T rating of the transformer; if that's the case, the only remedy I can see (other than finding a transformer with higher V*T) would be to drive the transformer from as low a voltage as possible, and then not attenuate its output so much. You're looking for output pulses of only +/- 10 mV, correct?

I agree with your initial finding. That's why I tried using the op amp current source (it is compliant when there is no pulse, with about 100 mV on the output.) That is also why I tried the T6427. It has a much higher Vt rating. However, it is designed for 10 KHz operation, and it does not do so well at 1 KHz.

 

Do you have any scope traces you could post? It might help if we could get a better idea of the waveforms you're actually getting from this circuit...

 

Do you have any scope traces you could post? It might help if we could get a better idea of the waveforms you're actually getting from this circuit...

I will put them together and post. It may take me a couple of days though, I am on vacation and my lab is not available (also don't want to upset my wife too much by working).

I will do that next week.

So far you have all been extremely helpful and I appreciate that a great deal. Thanks.

 

You're posting here about pulse transformers WHILE YOU'RE ON VACATION?????

That's just... just... just WRONG!!!

 

You're posting here about pulse transformers WHILE YOU'RE ON VACATION?????

That's just... just... just WRONG!!!

Ah well, More to come.