power supply chokes for solid state?

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.

i know chokes (inductors) have long been used in tube power supplies, where the benefits (good ripple rejection, better regulation and smaller cap size than capacitor-only supplies) outweigh the drawbacks (higher power supply impedance, size and weight of chokes), but recently i've noticed some solid stage gear (the Musical Fidelity stuff in particular) using chokes in the supply to seemingly good effect. so i'm considering trying a choke-regulated supply in a solid-state preamp project i'm working on.

does anybody know any good references on designing choke power supplies? all of the information i've found so far pertains to high-voltage (tube) supplies and i'm trying to figure out how to optimize such a supply for the low-voltage rails i would use (e.g. +/- 24V). basically i'm looking for design tips, rules of thumb, basic formulas - anything to help me find optimum component values (i'm an EE so i can deal with the math). also, are there any good sources for the kinds of chokes i would need? they seem rather hard to find.

And they're quiet, too...
Inductors are out of fashion, due to weight and cost. I've had trouble finding larger current capability units, myself.
One option is to strip an old transformer core and wind your own. Look up the current capability of various wire gauges, derate somewhat due to the fact that the wire will be in a confined space, and start winding. (If you use heavy enough gauge wire, the R of the inductor can be quite low.) Use ordinary shellac to cement layers in place. If you want to isolate layers from one another (lower capacitance), use kraft paper (i.e. grocery bags) to wrap each layer.
Yes, it's a pain in the ***, but since inductor values for power supplies are not as critical in value as they would be in the signal path, you have plenty of latitude as to final H values. The values calculated from the formulas for ripple are minumum values; anything over is gravy.
Anyone who's been fiddling with electronics for a while ends up with a few dead power transformers in a dusty box somewhere. Why not put them to good use? Think of it as recycling...you're helping the environment.

Inductance For Your Supply design
by Fernando Garcia Viesca
Audio Amateur 4/87 pages 22-26

"We place the inductor before the rectifier, and therefore the core is AC powered and reset every half cycle.
Core saturtion is no longer a problem and we are free to include as much or as little inductance as the circuit requires"

Also Onkyo use an inductors prior to the bridge recifier in some of their amps, they call it a "Real Phase Power Supply"

Regards James
thanks for the info guys, i'll look into that ON Semi handbook and so forth. incidentally, i've only found one source for power supply chokes, this esoteric tube audio shop in switzerland:


they look very expensive but pretty hard-core. anyway, i'm sure if i keep looking hard enough i'll be able to find one in the states (sorry grey, i'm too lazy to wind my own... =p).

Search the web for "psu designer" + "Duncan" and you will find PSUD2 which is a great amateur power supply design package that I was involved in betatestin for. I really like it -- and it supports inductors.

You can get some pretty good chokes in various flavours from http://www.hammondmfg.com -- but I had custom toroids made up. It seems http://www.plitron.com are planning to make chokes as well.

hmm yeah, i've actually had Toroid Corp. of maryland build me a couple power xformers for my amp project last year, i saw that they could do chokes too. i've also heard (though i could be completely wrong) that traditional E-core inductors/xformers have better performance characteristic than toroids, other than EMF radiation. anybody care to comment on that?
Can of worms -- toroids versus traditional core

I am opening up a can of worms here. Technically, the Toroid is always superior to a traditional form factor. Space, leakage etc.

However, many claim that toroids don't sound as good as EI's etc. Having separated windings does seem to make sense + the small air-interruptions/gaps in the magnetic path may actually be useful.

I went with toroids myself from a space/cost/availability consideration.


As far chokes go, I believe that air cores are the best because you avoid problems of core saturation but you have to pay the price of larger size and cost. I have only used E-cores myself and they might be better than toroids(being less efficient than toroids, E-cores might suppress charging pulses better? This is only speculation on my part).
Ayre acoustics use toroid chokes in their poewer amplifiers maybe if you called Charles Hansen (at Ayre) he could shed some light on the subject.
I have been using toroids for power transformers and prefer them over E-cores by a big margin. Lately I have been playing with R-core transformers and I believe that they might be better than toroids, unfortunately not may manufacturers produce them. I will keep you posted.


PS. Has anyone else had experience with R-core power transformers?

[Edited by jam on 06-15-2001 at 07:15 AM]
Something doesn't make sense here, or maybe I'm missing something. You want to use a choke to clean up the power rails (+/- 24V) in a pre-amp? What's the total current draw? 1-2 amps at most?

Is there a reason you don't want to use a linear regulator? If you're concerned about crosstalk, use a pair (+/-) for each channel. Even simpler, put LM7824/LM7924 regulators near each section. Of course, you'll need significant input (1000uF) and output (220uF-1000uF) capacitors on the regulators, along with a metallized polypropylene capacitor (~1uF) bypassing the output capacitor.

If you want to use a choke to clean up the pre-regulator power being distributed, use a 16 or 18 guage air-core inductor (an iron core would magnetize; would this be good?), built for crossovers, and shield it (otherwise, you're spraying 100/120Hz through the enclosure). I think the capacitance multiplier supply from Elliott Sound Products ('http://sound.au.com/project15.htm') would work better though.

Good luck.

[Edited by thoth on 06-15-2001 at 07:27 PM]
actually a choke-input power supply is a type of power supply design in itself... it uses an LC filter after the rectifers instead of just shunt capacitors, and has some nice characteristics that make it superior to a capacitor-only circuit when designed properly. you are probably thinking that the choke is just being used as a RF filter (like a ferrite bead or something) but it is actually an integral part of the rectifier circuit. compared to a conventional capacitor-shunt supply, it has better regulation, far better ripple rejection, and more stable current characteristics.
It seemed to me that you were looking at using the choke in a PI filter on the power rails. The transistor in sited capacitor multiplier performs similar functions.

As I see it, with a PI filter in the power supply, you have cleaned up most of the line noise and ripple, but you still have to deal with component-generated noise and ripple. With distributed LM7824/7924 regulators, this problem goes away too.
dorkus said:

i would use regulation downstream close to the circuit, though not LM78xx... i prefer the LT1033/1085 which have superior regulation characteristics. i may also considering building a opamp-based regulator circuit.

I just found 'http://www.audiocom-uk.com/q_power.html'. This should work even better then the LT1033/1085. The bad news is that these are a bit pricey.

Good luck.
Thoth said:
I just found 'http://www.audiocom-uk.com/q_power.html'. This should work even better then the LT1033/1085. The bad news is that these are a bit pricey.

oh cool! that is exactly what i was looking for... a high-performance regulator based on a good opamp like the AD797. i'm sure it's easy enough to make one yourself, they probably charge a premium for the compact board, assembly, etc. also you could probably get better performance by not having it be a plug-in module like this one, just do it right on the board by your circuitry. i've seen a bunch of precision regulators based on opamps like this, i wonder if they do anything different?
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.