![]() ![]() |
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
In Reply to: how much can solid state amps be improved by using CLC filtering? posted by freddyi on June 09, 2001 at 22:30:09:
Hi Freddy,
My interest for choke filters in the power supply was arosen by the big claims by Mark Levinson and Tom Colangelo for the choke input in the Cello's power supply. In the book Transistor Circuit Techniques by G.J.Ritchie (Chapman & Hall) I found the circuit for a choke input filter i.e. a choke between the rectifier and the filter cap. For a supply of 20V and 0.5 A you end up with a minimum value for the choke of 40 mH (that is 40 milliHenry). Considering the current that is going through the choke it is a rather unpractical value.
So I experimented somewhat and decided to do a simple lowpass filter as used in a loudspeaker crossover. I placed the choke between the first supply cap and a second additional cap of the same value. I used a choke of 22mH and the caps are 4700 mufarad. For a dual supply say +/- 15 V you need four caps and two chokes. The nice thing is the ripple voltage is reduced by a factor 10. The choke I used is from Intertechnik Germany;. part#1340984 has a DCresistance of 4.26 Ohm and is about 6 $. The voltage drop is not too large. It is not a air coil but a 'Corobar' coil probably iron dust.
Later I found out Nelson Pass uses the same Pi-type filter in his power amps, with a 2.2 mH choke. I used the 2.5 mH choke from my Klipschorns crossover in the Ampzilla power amp.(Laminated iron core)
With the Ampzilla a heard a significant improvement in sound.{The Ampzilla does not use any active supply regulation}. With the preamp I did not hear a improvement in sound but I retained the choke because the hum is much lower with MC-phono. I do not use preregulation in my preamp as I found the Jung regulators sounded better that way. There is also a thread of me on powersupply's some time ago; could find it right now.
If you insist on using AIR coils you end up with boat anchors indeed, if you want a reasonable low DC resistance!
I hope my post clarifies the matter.
ELSO
Follow Ups:
Elso, would you please post the formula used to compute the minimum sized inductor for a choke I/P filter. I am going to dig the book you refered to out of the library for additional study.I am at the "brainstorming" stage of "designing" a hybrid power amp. A choke I/P filter P/S for the O/P stage fascinates me. The P/S I have in mind will be 40 V. and 6 A. per rail. Boat anchors for sure when you consider following the ARRL Handbook when building means using a LCLC filter. 50 WPC monoblocks are going to be about 35 Kg. each, OUCH.
Eli D.
Hi Eli,
The formula for the choke input filter is I=V/942L where
I = critical current
V = output voltage
L = inductor in Henry's
Also note the choke input filter does work as good at a current lower than I critical.
To calculate the capacitor: rippleV=3.38x10^-6/LC at 50Hz.The maximum current in my Amzilla is also 6A but at 80V! With the inductor I mentioned and four 16.800 mufarad capacitors the pi-filter has no problem. I tested it at full power in a 4 Ohm load! For the choke input filter you have to search a much higher valued choke with 6A current rating. Hopefully we are not talking at cross-purposes.
ELSO
nt
Dear Elso -no surprise in your experiment - if building solid state I'd employ the LC pi filter after the input cap - could use Duncan's PS designer to juggle things for a smooth turn-on the 2nd cap is probably => 2X the input to start the simulation.
a M6 laminated xover choke seems suited for some solid-state designs keeping its axis at a rt angle to power xformer (assuming EI) - I currently have 2-3mH 10ga air cores tied up in a not so great DIY speakers. In larger amps pulling a lot of current there could be some saturation with steel core - wonder where one would run into trouble?
sideband IM problems were brought up in a discussion with Steve Bench over push-pull tube guitar amps. I never had skills nor knowledge to build solid-state - much more comfortable with tubes.
btw - guess who broke a tweeter flange?
Freddy
ps - I really appreciate John Curl's immense contibutions to solid-state art - cool
This post is made possible by the generous support of people like you and our sponsors: