Resonant power converter

familiar names are still here.

ing application. I used Bourns 33mH inductor part number RLB9012-323KL whi ch is rated for a maximum DC current of 90mA. In my application it is carr ying a sinusoidal AC current of 85mA peak at a frequency of 44 kHz. Althou gh it was only running at ~ 20 degsC above ambient, after a few minutes, th e inductor failed. It still measured as the correct DC resistance and indu ctance using a bench multi-meter; but blew the switching transistors and th e fuse if I tried to use it in circuit again. It was a hard problem to fin d because I changed every other component first, before suspecting the 33mH . Repeating with a fresh 33mH did the same thing after some minutes of see mingly working OK. I am wondering if I am permanently damaging the ferrite core somehow? But my peak current and therefore peak 0.5*LI^2 is not exce eding the datasheet spec and my frequency is below the SRF. Am I being ign orant here in trying to use this inductor in a totally inappropriate way? Please advise.

Weird, Does the bench meter give you the Q at various frequencies? Can you measure any change in the inductor? I had some weird effects of water (from pcb cleaning) getting in a spoling the Q

What's the peak voltage?

f familiar names are still here.

ting application. I used Bourns 33mH inductor part number RLB9012-323KL wh ich is rated for a maximum DC current of 90mA. In my application it is car rying a sinusoidal AC current of 85mA peak at a frequency of 44 kHz. Altho ugh it was only running at ~ 20 degsC above ambient, after a few minutes, t he inductor failed. It still measured as the correct DC resistance and ind uctance using a bench multi-meter; but blew the switching transistors and t he fuse if I tried to use it in circuit again. It was a hard problem to fi nd because I changed every other component first, before suspecting the 33m H. Repeating with a fresh 33mH did the same thing after some minutes of se emingly working OK. I am wondering if I am permanently damaging the ferrit e core somehow? But my peak current and therefore peak 0.5*LI^2 is not exc eeding the datasheet spec and my frequency is below the SRF. Am I being ig norant here in trying to

About 800V.

It was in series with a 630V rated 270pF capacitor. The voltage across the cap was ~ 1kV peak! At first I thought the cap was breaking down. I repl aced the cap with a 3kV part and saw the same failure.

I'm trying a different LC network now with lower value inductors and lower voltages; however, the frequency is similar; and I'm worried that my peak A C currents are still close to the maximum DC ratings of the inductors.

I've been tabulating L*I^2 and even f*L*I^2 of different circuits to gauge the relative stress. Not sure if this is a valid thing to do.

Peak current doesn't matter... RMS current does.

You are likely arcing the insulation of the wire inside the inductor. Little inductors like this are rarely voltage rated.

The ferrite core may be conductive, too. So you can get wire-wire breakdown and wire-core breakdown.

That is most likely the problem. Dielectric failure sometimes takes time. It'll be hissing in there but you may not hear it.

API-Delevan states a dielectric rating of 2500V but you have to inquire about the working voltage across the coil. Dielectric withstand is destructive over time and the allowed working voltage (continued exposure) will be many times lower. Best to get that in writing.

It could mean a custom coil job. If so, talk to Xfmrs or Renco.

Thank you John and Joerg. You are right. I cut away the shrink jacket from two failed inductors and I can see the point where it was arcing through the enamel. It might be better if I used axial rather than radial inductors!

Newava has been good to us lately. Todd.

Single layer solenoids are less likely to arc wire-to-wire.

You might put several smaller inductors in series, but you should still test them to estimate breakdown voltage.

There are transformers specifically intended for capacitor charging, or for CCFL backlight drive, that have high voltage ratings.

If there's quantity involved, you could get a custom inductor.

A 33 mH inductor is going to have a lot of turns of fine wire.

from two failed inductors and I can see the point where it was arcing throu gh the enamel. It might be better if I used axial rather than radial induc tors!

Measuring the self-resonant frequency might be useful. Getting a high self- resonant frequency at that kind of inductance means winding the coil as a s eries of "banked" windings which means less voltage drop between each layer (and more - narrower - layers).