Help with creating customized pulse transformer

[Bryan Stone]

Goal: We wish to use an existing high frequency, saturable core, toroidal resonant windings, pulse transformer with the following specs:

saturable soft core resonant, off the shelf pulse transformer capable of producing an asymmetrical negative polarity pulse of approximately Constant period of 21.7 milliseconds at a frequency to be specified.

The pulse should have a rise time of at least 2 microseconds or less and 6 milliseconds dwell time pulse duration.

The duty cycle will be approximate a period between pulses of 20,7 milliseconds.

The two power levels will be 500 volts, and a second at 1000 volts with a current flow for both of 3 amperes or 1,5kv and 3kv at 46khz frequency.

We wish the narrow pulses, which are asymmetrical and of negative polarity, to have as short a rise time as possible.

We intend to control the characteristics of the pulse outputs
with MOSFET gating and control circuits customized to our specifications above.

Does anyone know who can reliably complete this project?

 

[Sceadwian]

This must be for a sonar application, the power levels you're using are no joke! What 'off the shelf' transformer have you investigated so far? I'm not even sure you'd want one, it'd be expensive, you'd be better off investigating someone that had the knowledge to design a transformer that would meet your specs.

 

[Bryan Stone]

I will try to get specific examples. Is this within your capabilities to create a prototype?

 

[Sceadwian]

Nope definitely not, magnetics definitely aren't my forte.

 

[Mosaic]

This item looks like a decent starting point:
**broken link removed**

 

[Ubergeek63]

Goal: We wish to use an existing high frequency, saturable core, toroidal resonant windings, pulse transformer with the following specs:

saturable soft core resonant, off the shelf pulse transformer capable of producing an asymmetrical negative polarity pulse of approximately Constant period of 21.7 milliseconds at a frequency to be specified.

The pulse should have a rise time of at least 2 microseconds or less and 6 milliseconds dwell time pulse duration.

The duty cycle will be approximate a period between pulses of 20,7 milliseconds.

The two power levels will be 500 volts, and a second at 1000 volts with a current flow for both of 3 amperes or 1,5kv and 3kv at 46khz frequency.

We wish the narrow pulses, which are asymmetrical and of negative polarity, to have as short a rise time as possible.

We intend to control the characteristics of the pulse outputs
with MOSFET gating and control circuits customized to our specifications above.

Does anyone know who can reliably complete this project?

the simple answer is it's not going to happen. there is no not so simple answer.

your power levels have contradictions.... 500 and 1000V will generate currents based on your UNSPECIFIED LOAD,

your post seems to beg for a 6mS 1500W pulse (at the low end) with a 21.7Hz repetition rate.

you do not specify what your power SOURCE is....

you would be better off at a higher switch frequency...

and so on

In short, you are in WAY OVER YOUR HEAD!!!

 

[johansen]

why don't you just build a pulse forming network and use a 1200v igbt to turn the power on and off?

there are no off the shelf transformers that will work for you. there are off the shelf magnetic materials that can store enough energy for this application but it seems to me the problem being the very long on time (21mS) compared to the turn off time. if you want the transformer to saturate in less than 2 microseconds, but you want the volt seconds to appear at the secondary for 21000 uS you're going to need an exceedingly high permiability core. now note that you still have to reset the core and you only have 6 mS to do that in.. so you tell me what the voltage is going to be at the output!.. if you can't handle positive voltages then you will have to disconnect the load while you reset the core.

 

[Ubergeek63]

better off just charging a cap to the required voltage and dumping it to the transducer