I am building a high voltage RF transformer and I'm wondering what considerations I need to take in order to build it correctly.The transformer primary will be powered by a 12V square wave signal around 5-20kHz. The secondary needs to put out at least 1kV at very low amperage (in the mA range, the lower the better).What considerations I need to take into account in building the transformer?Thanks for all the help!
very last transformer I made had a 1000 volt secondary, a 28 volt time-honored, ran at 22kHz, and a 20mA potential. It change into wound on a ferrite style and change right into a dice about a million.5 inches on each and every area. I had lots of problem protecting it from shorting out. I used lots of particular HV insulating tape between winding layers, and managed to get some operating, notwithstanding it change into complicated. It change into used in a voltage doubler to generate 2500 volts DC, so the entire winding had to face up to that voltage. undergo in concepts that there are purely some volts between consecutive turns, yet after one hundred turns, that builds as a lot as thousands of volts. .
Obviously, you need a turns ratio of 1000:12, or 83:1. Your power requirement is low (1000V x 1 mA 1 Watt). I've designed a lot of RF transformers. To me, the main consideration is operating frequency. Your magnetic core must perform well at your operating frequency. (The higher the reactance and the higher the Q, the better.) Core manufacturers provide that information on data sheets, and sometimes provide it by number of turns, which I really like. You can use that information to predict your no-load current. At high frequency, high-mu materials don't work well. At low frequency, you need high mu to get enough flux in the common magnetic circuit. At high frequency, you need fewer turns as well, and need to have a shorter magnetic circuit. You also need to be careful (especially at low frequency) to keep well away from saturation. If you saturate the core, your current spikes and the transformer no longer works. Be careful not to saturate the core with DC. How much magnetization (flux) it takes is measured in Webers or volt-seconds. At low frequency, you have a lot more volt-seconds. Wind the core carefully to make sure that as much flux as possible passes through all the linkages. Leakage inductance means you take more reactive power to get your output. Do your input and output need to be isolated? If not, you can use a tapped transformer. At 1kV output, dielectric breakdown is fairly easy to handle. Using standard magnet wire for the secondary should be fine.
I should think your primary concern should be the safety angle. How will the transformer be isolated from the public?
Last transformer I made had a 1000 volt secondary, a 28 volt primary, ran at 22kHz, and a 20mA capability. It was wound on a ferrite form and was a cube about 1.5 inches on each side. I had a lot of difficulty keeping it from shorting out. I used lots of special HV insulating tape between winding layers, and managed to get a few working, but it was difficult. It was used in a voltage doubler to generate 2500 volts DC, so the entire winding had to withstand that voltage. Remember that there are only a few volts between consecutive turns, but after a hundred turns, that builds up to hundreds of volts. .
