Replacing power supply resistors with chokes

[debk]

I put a Triad C-7X choke into my Eros.

Fairly easy to do, just had to drill two holes, remove the resistor and solder in the two leads.  I also ran a ground wire from a solder lug I placed on one of the screws to the star ground lug on the power tranfo.

It did make a large difference in the sound.  It sounds "brighter" and maybe more detailed now.  I originally had a rear firing super tweeter on top of the speakers to add "a little twinkle to the high end".  After the choke was installed I disconnected the super tweeters, it was just to bright.  I may try and add a L-pad to the super tweeters to tone them down a bit, or play with the capacitor connected to them.

No problem with any 60Hz hum being picked up.

First impression is that I like the way it sounds with the choke.

Debra

 

[Paul Joppa]

Cool! Thanks for documenting this; we all learn when one person experiments. The beauty of community....

 

[Grainger49]

Debra,

Does the plate still fit within the base with the choke mounted there?  It looks awfully close to the edge.

 

[debk]

Yes the plate still fits on the base.  The perspective in the photo makes it look like it hangs over the edge.

I could have mounted it a little closer to the PCB.  I had to sand down the inside of the wood case a couple of mm.  Forgot to take into account the lip the top plate sits on. 

Debra

 

[Grainger49]

Thanks.  I own a rotary rasp that makes quick work of nibbling out wood.  I'll take it to Paul's house.

 

[debk]

I just used a Dremel tool.  It took only a few seconds to sand the area down

Debra

 

[Doc B]

Hey Deb, do you have a scope? Your description of the sound makes me curious to see if maybe the choke is picking up some high frequency interference that makes the sound seem brighter. Might be interesting to look at the noise floor and see if there are any really high frequency spikes riding on it, like maybe up in the MHz range.

 

[debk]

Sorry, I don't have a scope.  probably should look into getting one

Deb

 

[debk]

Actually I do kind of have a scope.  It is the DPScope, a USB scope that goes out to about 1MHz.  I have not used it in a few years and forgot about it.

I used it connected to one of the outputs with no input into the Eros

Since we were looking for possible high frequency spikes, I set the scope to do a FFT

40Hz/div Shows a 60Hz spike

20KHz/div

Don't know if this data is meaningful or not, or if I even made the measurements properly.
Probably should get a real scope with a higher bandwidth.

Debra

 

[Grainger49]

Debra,

1M Hz bandwidth is enough to tell if you have any oscillations.  And the software is more versatile than my Tek digital storage scope.  Your FFT shows no high frequency noise.  I am surprised that the FFT shows such high 60 Hz component, but that is the advantage of it versus the scope trace itself.

 

[debk]

Had to see what it looked like without the Choke

40Hz/div No Choke

20KHz/div No Choke

One more 20KHz with Choke ( Not sure if I captured the trace in the previous post correctly)

Debra

 

[Doc B]

I my limited experience with vacuum tube audio design I have come across things one wouldn't think audible, like digital filters that can radiate spikes in the 10MHz range. Under Buddha's tutelage I wrestled with filtering that kind of stuff out and found that the difference was audible.

 

[JC]

debk, I'm curious about the program you are using, and what is "acquiring" the signal for it to analyze.

 

[debk]

Check out
DPScope.com

Debra

 

[JC]

Thank you, I most certainly will.

 

[johnsonad]

The PC board was designed around that time, and it retains the optional terminals for a choke.

Paul, the chokes are here and I'm curious if there are different terminals for the choke or simply removing the 270 Ohm resistor and wiring in the choke (Paramount use)?  There will be a shorter wire run if I I use the present terminals when mounted above the filament choke.  Lastly, will twisting the wires reduce potential hum pickup? 

Thanks,

Aaron

 

[Paul Joppa]

Paul, the chokes are here and I'm curious if there are different terminals for the choke or simply removing the 270 Ohm resistor and wiring in the choke (Paramount use)?  There will be a shorter wire run if I I use the present terminals when mounted above the filament choke.  Lastly, will twisting the wires reduce potential hum pickup? 

Thanks,

Aaron

I'll attach a picture of the missing silk-screen layer. The two terminals at the far right, labeled "L," were for the choke. (This design is six years old...) Twisting is generally a good idea, and using the resistor terminals may in fact be better than using these terminals, which run through a large loop area on the board. In simplified theory the magnetic field should be pretty symmetric and orthogonal, hence not a problem either way, but the real world always holds unanticipated surprises for us!

 

[johnsonad]

Thank you Paul.  I'll post some pictures once finished.  Another thing, I measured the chokes and one is outside 10% at 238 Ohms and the other is 251 Ohms.  This is to be expected with the tolerances Triad specifies.  Am I looking at any potential problems?

 

[Paully]

Yep, I just measured both of mine and they both came back almost exactly 236 ohms.  But they do state 20% tolerances, hopefully that isn't an issue for the Paramount either.

 

[Paul Joppa]

I think the C-7X design by Triad goes way back. When we first bought them, they were actually labeled MagneTek (they had bought Triad from Litton Industries in the seventies) and made in Mexico. I dug one out of the stash and read 256 ohms.  The current Triad web site has an interesting summary of ownership changes, and are currently owned by Axis of Taiwan, who were by that time (2004) actually manufacturing 95% of the product line. But the current production sold by Allied appears to be made by Hammond (look, feel, and boxes are distinctive) and has changed from the original layer-wound to modern nylon bobbins. No way to tell if anybody changed the internal design during any of those iterations. The C-7X is listed in the current Triad catalog, but I have no idea where it might be made, or if it is the same as what Allied sells.

Anyhow, since this is past the first cap, you can estimate the voltage loss fairly accurately. In a Paramount drawing 80mA, the 30-ohm difference will become 2.4 volts; less with other amps that draw less current. It's nothing in comparison to the effect of different power line voltages!