0
\$\begingroup\$

I am assembling an electronic product using off-the-shelf components. I want to connect the device directly to the power socket, so I plan on using a switching power supply (this) to get the 12V output I need.

I'm considering whether I should add some EMI filtering before the power supply. However, the datasheet for the power supply indicates an inrush current of up to 65A, while the passive filters I've found so far (e.g. this one) usually handle far less current.

Should I take the inrush current of the power supply into account when selecting an EMI filter, or should I only consider the rated current during normal operation? Could the filter fail upon device startup due to high inrush current?

At this point I'm also wondering whether there is there a fundamental flaw in my plan to add an EMI filter before the switching power supply. I expect the power supply itself might include some filters, but I do not have access to its schematic, so I don't know exactly what to expect.

Finally, if you could point out any good resources I could study regarding this topic in general it would greatly appreciated.

\$\endgroup\$
3
  • \$\begingroup\$ How long duration is the inrush current, or even better, how big is the I^2*t area for the inrush? \$\endgroup\$
    – winny
    Commented Apr 24 at 11:33
  • \$\begingroup\$ Sadly, it doesn't say it in the datasheet. Maybe I could measure it, but given the large amount of current involved it's not obvious to me how to do so. \$\endgroup\$ Commented Apr 24 at 11:53
  • \$\begingroup\$ Oscilloscope and current shunt or current clamp. \$\endgroup\$
    – winny
    Commented Apr 24 at 12:06

1 Answer 1

1
\$\begingroup\$

The inrush current of the power supply would be flowing through the inductors of the EMI filter. The current rating of inductors is based upon two factors -- the ability of the inductor to withstand the heating effects of the current, and the core saturation current, for solid-core inductors.

Core saturation will prevent inductors acting as inductors. EMI inductors are not meant to limit current, so the main effect of core saturation in EMI inductors is that they may not filter properly. Since inrush current is a transient effect, you may or may not care how well your EMI filter works during the period of inrush current.

Also, because inrush current is transient, the extra heat produced in an inductor will be transient. If the wires are capable of handling the transient, you should be OK. The core may heat up as well, but because of the core's thermal mass, and the transient nature of the inrush current, core heating should not be a problem.

So, that leaves the transient current handling capacity of inductor wires as the main issue. Generally inductor datasheets will have little or no information about transient current handling capabilities. If you purchase name brand inductors, it is likely that a test of one (or a few) inductors will be representative of any inductor with the same brand and part number. So, my suggestion is to simply test the inductor with the inrush current, and see if the temperature rise is not so great as to damage the enamel or insulation. I suspect that, in many cases, there should be no problem.

\$\endgroup\$
4
  • \$\begingroup\$ Note that saturation depends on the direction of magnetization; in a CMC (also called a current-compensated choke), load current normally cancels out, and only the differential-mode aspect of the core can exhibit saturation. This mode includes a large air gap in the magnetic path, so the saturation current is very high indeed. (This isn't a well-documented threshold, I think; and I say this as someone having designed and tested many such parts myself! But indeed, it's hardly consequential, so we're not missing out on a lot.) \$\endgroup\$ Commented Apr 24 at 10:58
  • \$\begingroup\$ I see, thanks for the explanation. Altough, I don't want to design the EMI filter from scratch, so I actually don't know the inductor included in the commercial solutions. Maybe I can test the whole EMI filter and see if after a few cycles of turn-on/turn-off the transfer function is different. \$\endgroup\$ Commented Apr 24 at 12:04
  • \$\begingroup\$ @AjejeBrazorf if you do not have access to the inductor and cannot measure or at least estimate its temperature, then you will not know, from a few cycles whether the enamel might be damaged. It might be better to do a sacrificial test in this case. If the filter can survive, say 30 seconds of continuous current at the inrush amperage, then I would imagine it would survive many thousands of 1 second cycles. Not exactly analogous, but many motors can withstand locked rotor conditions (i.e. high current) for 30 seconds or more without insulation failure. Different design of course. \$\endgroup\$ Commented Apr 24 at 13:13
  • 1
    \$\begingroup\$ Good point. Thanks, I'll post the result if I'll end up doing this measurement. \$\endgroup\$ Commented Apr 26 at 9:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.