What are the main factors limiting frequency of operation in High Voltage transformers?

[idijoeteque]

I have designed a switched mode power supply that operates at approximately 250kHz, providing a maximum output voltage of 6kV. Using soft switching techniques, I get a good efficiency of around 95%. I was wondering, theoretically, what limits the maximum achievable switching frequency in HV transformers? I have seen high output current transformers that can switch at 1.2MHz and above with new semiconductor technology/planar transformers, but never see anything like this in HV transformers.

Why not?Is it to ensure enough insulation can be added to the transformer? Why not just encapsulate it in insulating resin?

 

[ronsimpson]

6kV. Using soft switching techniques, I get a good efficiency of around 95%.

That is good efficiency for 6kv. You are doing things right.
I would have started out with 200 to 300khz. It is hard to run much faster with high voltage diodes that are slow. AND Most high voltage transistors/MOSFETs are a little slow also. I think you can not do 1.2mhz at 6kv. If you are making 24V then I would think about 1mhz or at 5V maybe 2mhz.
High voltage transformers need many turns. I can not get that many turns using planar transformers.
Many of my power supplies are in the 15 to 20kv range. We always enclosed the transformers in resin or "transformer goop". Probably a simple resin is OK at 6kv. We enclosed the output diodes inside the transformer because at 20kv any air is a place for a spark to develop.

Please post a schematic.
RonS.
----edited----
It is possible to run into the self resonant frequency of the transformer. If you need help pushing the resonant frequency up we can talk about that.

 

[idijoeteque]

That is good efficiency for 6kv. You are doing things right.
I would have started out with 200 to 300khz. It is hard to run much faster with high voltage diodes that are slow. AND Most high voltage transistors/MOSFETs are a little slow also. I think you can not do 1.2mhz at 6kv. If you are making 24V then I would think about 1mhz or at 5V maybe 2mhz.
High voltage transformers need many turns. I can not get that many turns using planar transformers.
Many of my power supplies are in the 15 to 20kv range. We always enclosed the transformers in resin or "transformer goop". Probably a simple resin is OK at 6kv. We enclosed the output diodes inside the transformer because at 20kv any air is a place for a spark to develop.

Please post a schematic.
RonS.
----edited----
It is possible to run into the self resonant frequency of the transformer. If you need help pushing the resonant frequency up we can talk about that.

Hello Ron,

Thanks for your reply. Regarding the HV transformer turns ratio. Currently, I have multiple secondary windings which are connected in series at the output of the rectifier. Probably quite tricky to do with planar. However, my initial idea was possibly use a Cockroft walton style multiplier and have only a single secondary winding which achieves only a portion of the step-up required, which is about by a factor of >20 in total from input to output. ButI am unsure on this.

In regards to insulation, that is how the company I work for currently achieve the isolation requirements, using some insulating resin and custom winding bobbins for the transformer. With regards to the speed of the diodes, I am also investigating the use of siliconcarbide, gallium nitride switches so theoretically that shouldn't be an issue however in reality I am again not so sure. I will attach a schematic of what I currently have. It is a buck converter cascaded into a half-bridge LCC resonant stage, operating at fixed D, phase, and Freq. The buck converter achieves control of the output. A catch diode is included in the schematic for clamping fault voltages to the input voltage. I do not use this currently in my LTSpice simulations however, which I could also send if that would provide more insight.

Thank you!