Output Inductor Power Requirement

[gearheadgene]

Need some advice on output inductor power requirements - probably current, actually.

Assume a 100Wrms system, so Pmax = 200W. Switching F = 500KHz. Output load > 4-Ohm.

Imax is going to be SQRT(Pmax / R), or SQRT(50) = around 7A. Irms is something like 5A.

Based on that, do I choose an inductor that can handle 7A? Even though audio has the crest factor thingy going on (8:1) and will likely never have 7A on a continuous basis. When do the cores fail? Is it when the max current is exceeded. I believe one data sheet described the max current at the point when the inductance dropped to 80% value (core is starting to saturate).

Anybody seen the transformers from Panasonic that are meant specifically for class-D? Check out ETQA17B220. Digikey even has them for around a buck or 2.

One more thing - how does the switching frequency affect the inductance, or any other paramete? The inductors I've looked at are spec'd at 10KHz. What happens at 500KHz?

thanks

gene

[BWRX]

Originally posted by gearheadgene
Based on that, do I choose an inductor that can handle 7A?

A good choice would be to look for an inductor that can handle 7Arms. Ideally you don't ever want the inductor to come close to saturating.

Originally posted by gearheadgene
When do the cores fail?

The cores don't fail, they saturate. When they saturate they are no longer in the nicer part of the B-H curve.

Originally posted by gearheadgene
I believe one data sheet described the max current at the point when the inductance dropped to 80% value (core is starting to saturate).

Different manufacturers define this differently. Some are wild enough to use a figure of 50% drop in inductance to determine the maximum current rating...

Originally posted by gearheadgene
how does the switching frequency affect the inductance, or any other paramete? The inductors I've looked at are spec'd at 10KHz. What happens at 500KHz?

Inductance drops off as frequency rises. Most inductor specs at 100kHz will be reasonably safe to use for 500kHz as well. 10kHz ratings aren't good to go by if you'll be using it around 500kHz.

[Eva]

Conventional SMPS inductors are hardly useable at 500Khz. Also, conventional solid magnet wire is not very suitable for such high frequencies due to skin effect, and due to eddy currents when gapped ferrites are used.

[fumac]

Originally posted by gearheadgene
Need some advice on output inductor power requirements - probably current, actually.

Assume a 100Wrms system, so Pmax = 200W. Switching F = 500KHz. Output load > 4-Ohm.

gene

if rms= 100w , Pmax<>200w

Pmax is Power of P2P or ??

[gearheadgene]

Originally posted by Eva
Conventional SMPS inductors are hardly useable at 500Khz. Also, conventional solid magnet wire is not very suitable for such high frequencies due to skin effect, and due to eddy currents when gapped ferrites are used.

Hi Eva. I'm not sure what you're trying to tell me. Certainly many people are running class-D amps at 400Khz or so. Without being a magnetics guru, I can't say with certainty they have problems - yet they are using toroidal inductors at this power level and frequency.

Originally posted by fumac


if rms= 100w , Pmax<>200w

Pmax is Power of P2P or ??

Yeah, that one got me too a while back
Prms = Irms *Vrms
Ppk = Irms*sqrt(2) * Vrms*sqrt(2)
Ppk = Irms * Vrms * 2
Ppk = 2 * Prms

[fumac]

Yeah, that one got me too a while back
Prms = Irms *Vrms
Ppk = Irms*sqrt(2) * Vrms*sqrt(2)
Ppk = Irms * Vrms * 2
Ppk = 2 * Prms [/B][/QUOTE]

i think u r still wrong

take a look

it seem Vrms=V_p2p/2.8285
do u think so?

[gearheadgene]

it seem Vrms=V_p2p/2.8285
do u think so? [/B]

Your math works out that way because Vrms for a sine wave, is Vpk / 1.414 (i.e. sqrt(2)). For peak to peak, the voltage is 2X, so you divide by 2.8185 as you wrote. But your thinking is incorrect. You don't refer to a sine wave of amplitude X as 2X, even though the peaks are at +/-X. A 1 volt sine wave is 2V peak-to-peak, average is 0V, RMS is 0.7071 [i.e. rms = 1V / sqrt(2)]

Are you confusing peak-to-peak with peak power (or peak voltage or peak current)?

Check my first post, I wrote Ppk which meant peak power - not peak-to-peak.