Typical working voltage of toroid inductors

[AlienCircuits]

I am considering designing a custom buck SMPS that will convert a 1000V power source to a regulated 800V DC voltage. It will source very very low current.

I know that inductors cannot really be rated for a working voltage, since they can only have voltage transients that originate from Ldi/dt, so I cannot find datasheets that tell me the maximum voltage I can have across inductors.

I'd like to use something likes these:
**broken link removed**


Does anyone have any idea how to choose an inductor based on the voltage you plan to have across it?

 

[D.A.(Tony)Stewart]

The critical factor is the maximum Gauss and maximum DC current.
Calculate ET value and follow their instructions.
For flux density (B), calculate ET (V-µsec) for the application, and multiply by ET10 factor from the table. Max Gauss depends on Core, which is not specified for this series.

Core Loss (mW) = k2 * f^1.26 * B^2
Copper Loss (mW) = IDC^2 x DCR
Temp Rise (C) = (Total Losses (mW))^.833 * KO (from table)
Total loss in the inductor is 80 mW (ROS 1) and 280 mW (ROS 2) for 55°C temperature rise above ambient.

Perhaps Bmax losses are the over-riding factor not HV dielectric breakdown. Consult factory

Limit the DC bias (H) to 46 orsteds. Calculate H by multiplying H1 from the table by IDC of the application.

 

[FvM]

Dielectric strength is a problem on it's own. 1000 V working voltage would usually require dual coated enameled wire, or layer insulation.

Some manufacturers are offering inductors designed for higher working voltage, e.g Wuerth has some inductors specified for buck converters with 400 VDC input. By implication are regular inductors not necessarily suited for this voltage level.

I agree with SunnySkyguy, that a manufacturer statement should be obtained. Most likely you'll end up in defining a user specific part.

 

[mtwieg]

Probably the only way to find out aside from asking the manufacturer is to do a hipot test on the winding insulation yourself. Also avoid low resistivity ferrite cores which can develop significant eddy currents due to conservative E fields caused by high voltage.

 

[FvM]

I find a 500 Vrms test voltage specification for inductors with dual winding from Talema. It suggests a working voltage of 100 V according to usual safety standards. For 1000 V working voltage, we would apply at least 2200 Vrms/1min test voltage.

 

[D.A.(Tony)Stewart]

It may be prudent to coat the inductor with additional insulation. Kapton tape may be best for 5kV insulation and low µ. but silicone spray may offer several kV/mm of insulation and work too. THe breakdown between windings is most likely distributed so the maximum potential is not seen against adjacent turns. This may add unwanted capacitive coupling and lower the SRF and there may be better products for improving its safety, perhaps heat shrink tubing if it is small axial type.. The best way to find out is test it with a 1MΩ series resistor so any arcing will be <1 mA and not melt any insulation further or cause damage. Get a 25kV ignition coil from a car. and test it. Put 1MΩ in series and test for current. Resistor case must be >>1kV/mm. 1/4 res are rated for 500V . You can connect in series.

I had designed power supplies a long time ago as a 2nd source for Lexmark laser printers up to 25kV using sine resonant oscillators from 12V stepped up in resonant mode.. No spikes, No pulses, no EMI, no shielding, low cost,, about 25 cents for a multi-output transformer using 10k/day. Regulation was simple R divider tree using enough 1/4W R's to provide feedback at low voltage with 5 outputs some with PWM control from +15kV to -10kV , all < 1% regulation over all specs and cost < $25 in OEM qty. but $350 as a field spare part.

 

[sky_123]

Funnily enough, (for 1-off) even normal varnish is pretty good, to well beyond 10kV as long as it is not an extremely thin coating.
Or, if you want to go old-school, shellac. Try at your own risk! : )