Inductor selection for VFD filter appplication

[xj25]

Hi guys,

We had some problems with voltage reflections discussed in thread237-300370

We had two cases, 7 fans in parallel for which we selected a commercial sinus+common mode filter, and just one fan, for which we don´t have the space for that filter and there is no commercial solution for such low power (about 0,7A 220V 60Hz).

For this latter case I did some tests with a compact custom-made LC filter, made with PCB components for power supplies.

After looking for some alternatives I have some doubts about the inductor ratings:
- About the PCB inductor used in prototype, I don´t see any info about voltage ratings in the datasheet, apart from being used for "power supply" applications ¿what do you think?. Current rating is 0,9A DC (AC no idea, but probably more).
- I checked some commercial 3phase inductors (2,5A minimun) and seems to be of two types, for "line" or "motor", but not any clue in the datasheets about the real difference between them.

I am a little worried about voltage and thermal ratings of PCB inductor, but can´t use a "big" 3phase inductor without modifying quite a lot the existing cabinet.

I adjoint info about the inductors and some waveforms of the working filter prototype (including voltage over inductor and current).

It gets a quite nice sinus (or sine ;-)) at the output.

Good weekend

 

[MAGTiger]

Regarding the attachment, you asked about what is the switching frequency. This is the sometimes referred to as the carrier frequency. that is the frequency of the carrier pulses in the drive output. From your waveforms, the carrier frequency appears to be quite high. I'd guess higher than 8khz. Possibly 12 to 16khz. That more than anything is why the waveform looks so good out of your filter. In the end, if the inductors are not overheating and you get this kind of waveform then your doing pretty good. The only other issue would be the caps. you probably should measure how much current is going into the caps so they don't get overloaded. The inductors you've selected give you and advantage over the 3 phase line/load reactors in the presence of common mode inductance. The 3 phase models are typicall used for higher power than you appear to be dealing with and lower carrier frequency.

Neil

 

[xj25]

Hi Neil, thanks for your reply.

Switching is 18kHz, about capacitor it is used 1.5uF prolypopylene type rated for 850Vdc. Current is about 0,8A peak, 18KHz, and checking with rated load depending on cap losses in datasheet should be fine.

What it puzzled me a little was the quite clear parameters in the capacitors datasheet to select and size them, and the not-so-clear parameters for selecting the inductor in relation with voltage/Tª ratings.

About the 3phase motor side inductor, may be the switching freq. rating related specially with thermal capacity of the core?

Regards

 

[MAGTiger]

The switching frequency will have an effect on the temperature rise both in the core and the winding. However, this effect is not linear with frequency and will not be linear between core and coil.

Neil

 

[crthompson]

Unless you have to have your carrier frequency at 18khz for a reason, I would drop the carrier down below 10khz... I have read several application notes on different drives and it seems that most of them want you to run at the lowest carrier frequency possible that doesn't cause any adverse effects. That may help you out...

 

[crthompson]

I would recommend trying it at around 4khz...

 

[xj25]

Thanks cr,

VFD can be configured in 18,12,6 and 3 kHz pole switching freq.

Line voltage really is that freq.x2 so inductors are working with 36,24,12,6 kHz pulses.

If freq. is low, current consumption by the filter itself rises (L impedance reduction is stronger that C increase) and the VFD trips at start if set up with 3kHz. I am not very confident in permanent mode with 6kHz, so probably the safer point would be at least 12kHz (24kHz pulses line to line).

I got confirmation by Murata that the PCB inductor used is not suitable for that voltage levels, due to realibility issues with voltage stress over the first coil turns (however there is no problem with frequency, I think that they use insulated powder iron core)

So I am quite lost now, I will let you know if advance in some way....

 

[MAGTiger]

It is true that VFD will operate more efficiently and cooler at lower carrier frequency. However, the opposite is true with a sinewave filter. Typical sinewave filters will have a resonant frequency below the carrier. So the higher the carrier, the less carrier current the filter has to deal with. But lowering the VFD carrier so that it gets close to the filter resonant frequency causes the carrier current to increase in the filter and temperature rises. Caveat: this is for sinewave filters. dv/dt filters do not follow this rule as their resonant frequency is higher than the carrier.

Neil