How to prohibit negative pulses out of the transformer secondary?

[neazoi]

Assuming a transformer secondary outputs only positive pulses and negative pulses, a single diode can be used across the secondary, so that only positive pulses are out.

Is there any other non-semiconductor way to do it by using a second transformer or a special winding etc?

 

[Venkadesh_M]

Why don't you use a vacuum tube rectifier ?? !!

 

[neazoi]

Why don't you use a vacuum tube rectifier ?? !!

Ok, I had to be more specific in my first post. I am trying to do it using transformers only, and resistors and capacitors of course.
It seems difficult to me to do it using only transformers.
I am thinking of a special winding that would allow positive pulses to come out but cancels out negative pulses. Can you think of any way that this would be possible?

 

[Venkadesh_M]

I am thinking but when you saying "cancels negative pulses", It is don seem to be possible without non linear devices.....

- - - Updated - - -

Hi i think I got Idea, The core of the Transformer is a non linear device so it is possible to rectify only with transformers but it is some more tedious process...

 

[neazoi]

I am thinking but when you saying "cancels negative pulses", It is don seem to be possible without non linear devices.....

- - - Updated - - -

Hi i think I got Idea, The core of the Transformer is a non linear device so it is possible to rectify only with transformers but it is some more tedious process...

Hm... so a core saturation up to the point that the BH curve becomes non-linear you say...
No phasing of any kind, not at all?

 

[Venkadesh_M]

Try this, for this you have to design a new transformer.... which with stands the current for saturation.........

- - - Updated - - -

I am not sure about the polarities but its needed to be verified once again give some time......

 

[neazoi]

Ok you are trying to use it as a saturate reactor, setting the DC bias at the point were the BH curve becomes non-linear.

 

[Venkadesh_M]

Yes I do, still not found the polarities.....

- - - Updated - - -

Hi this is called as ,magnetic half wave amplifiers It is found in the lot of websites and you can also refer the exact design.........

 

[neazoi]

Yes I do, still not found the polarities.....

- - - Updated - - -

Hi this is called as ,magnetic half wave amplifiers It is found in the lot of websites and you can also refer the exact design.........

Perfect, I also have two books for these in my website qrp.gr
I will have a closer look

 

[neazoi]

I have red the book part that refers to the circuit you propose. RX is a diode, so it is not good to me. If a diode is to be used, i would use it directly across the windings of the secondary of the transformer and avoid using another transformer to do the job.

 

[FvM]

A saturable inductor with DC bias can be used to cut signal halfwaves similar to a rectifier. But why would you do it?

 

[neazoi]

A saturable inductor with DC bias can be used to cut signal halfwaves similar to a rectifier. But why would you do it?

You mean saturable reactor? http://en.wikipedia.org/wiki/Saturable_reactor
How? In the picture shown in the wikipedia, to input the negative/positive signal at one end of the inductor (G) and then take the output from the other end?
According to this page http://www.vias.org/feee/trans_08.html
saturable reactors tend to regulate AC power only in one direction: in one half of the AC cycle, the mmf's from both windings add; in the other half, they subtract. Thus, the core will have more flux in it during one half of the AC cycle than the other, and will saturate first in that cycle half, passing load current more easily in one direction than the other.

 

[FvM]

I don't mean special "saturable reactor" designs, just a small inductor/transformer similar to the first circuit suggested by Venkadesh_M in post #6, but biased by a current rather than a voltage source.

The core's magnetization curve should have sharp saturation and low hysteresis like silicon steel or even better amorph metal tape.

Are you talking about a principle concept or a real design?

 

[neazoi]

Are you talking about a principle concept or a real design?

Yes, after brainstorming with an English fellow, we have theoretically successfully come up with a design of a NAND gate. The difference from the old designs of the 50s is that this gate does not use special cores with stiff B-H curve, but modern ferrite cores like the FT23-43. It uses them as simple transformers, it does not use their saturation region, where they are permanently magnetized in one or the other direction. That to say, any core could be used. I will post the results as long as I have physically constructed and tested the gate.
The only "problem" of this design is a diode that is used to remove a negative pulse, whereas let positive pulse come out.
This diode is what I am trying to avoid and use a transformer in such a way to do the job. I have not yet thought of a possible operation with a split winding or something like this, the only possible solution I have seen (as suggested) is a core that is saturated so that it can pass mostly the positive pulses and much less the negative ones.
But this image http://www.vias.org/feee/img/02161.png shows this core in series, like a rectifying diode in series.
I winder if the circuit in post 6 let also positive pulses go down to the ground.

 

[FvM]

In simple words, you need a non-linear element, either a semiconductor or a saturated core.

 

[neazoi]

In simple words, you need a non-linear element, either a semiconductor or a saturated core.

It seems so, according to post #4. A transfomer operating in the non-linear region could theoretically do the job. Unless another transformer way exists, something that uses special windings in phase and out of phase, to cut-off negative pulses.

 

[FvM]

A transfomer operating in the non-linear region could theoretically do the job.

That's more than theoretical. The problem is about design effort and performance.

Unless another transformer way exists, something that uses special windings in phase and out of phase, to cut-off negative pulses.

This can be safely excluded. A non-linear operation can't be performed by any combination of linear circuit elements.

 

[Venkadesh_M]

Hi what FvM said is purely true and you can't cancel only negative pulses with any kind of winding connections... the only possibility is using a saturable core, for your more exact operation you have to choose a core which is more linear before it saturates, so that will give you un affected +ve pulse and no negative pulse.....

the circuit in #4 is
The core will be +ve ly saturated and If you apply a positive pulse there will be no more increase in flux so the primary is no more inductor and just a wire so it short circircuits and gives you 0 Volt..
But when giving negative voltage The flux reduces and the transformer acts as a inductor and drops the voltage to be on the output....

another thing is the source shd be a current source because if its a voltage source it will give some extra current when the flux is reducing and it will make it neutral effect...
and sorry for giving the circuit with diode and i didnt noticed..

 

[neazoi]

Hi what FvM said is purely true and you can't cancel only negative pulses with any kind of winding connections... the only possibility is using a saturable core, for your more exact operation you have to choose a core which is more linear before it saturates, so that will give you un affected +ve pulse and no negative pulse.....

the circuit in #4 is
The core will be +ve ly saturated and If you apply a positive pulse there will be no more increase in flux so the primary is no more inductor and just a wire so it short circircuits and gives you 0 Volt..
But when giving negative voltage The flux reduces and the transformer acts as a inductor and drops the voltage to be on the output....

another thing is the source shd be a current source because if its a voltage source it will give some extra current when the flux is reducing and it will make it neutral effect...
and sorry for giving the circuit with diode and i didnt noticed..

Could you make a guess about the turns ratio? A starting value to experiment with.

 

[Venkadesh_M]

Hi first forget about the primary, design a core of known area and material, calculate the maximum flux density.
Now calculate the Ampere turns needed to make it saturate, and test it after winding primary coil..
The bias shd be made by a constant current source, A small core would be better for your application
because it will saturate for a lower Ampere turns..