Attenuator build problems

[Chuck H]

I have built this twice now with different parts in different enclosures. Both did the same thing to me. Lowish impedance at full CW and near 0 impedance at full CCW.

By my figures I should have about 6.5 ohms at either end and about 9.5 ohms at halfway up.

I've redrawn my shematic from my wiring diagram and redrawn my wiring diagram from my work just to be extra sure there are no mistakes. But it's not working and nothing is wrong!?!

I'm hoping there is some mistake in the circuit that I for some reason can't see. So here's the schem... What's wrong with it?

Thanks

Chuck

[dai h.]

hey Chuck, I think the plan(that schematic) looks wrong so if you built exactly according to that I don't think it will work. Try drawing it out on a piece of paper and I think you'll see where you went astray.

[Steve Conner]

You have the tip and ring contacts of your jacks mixed up. The cold line should be a solid connection all the way through, and all the attenuating, etc. should happen in the hot line.

[dai h.]

the 0.5mH and 12mH I think these should be reversed also

[Chuck H]

Dai, the inductors are correct in the build. But your right. They are reversed in this schem. I put it together pretty fast to post it here.

Steve, you nailed it. Only if I were using isolated jacks (and I'm not) could it work as I have it now. Doh! Notice the lack of a chassis ground MISSING from the schem!?!

I knew it had to be right there, but I couldn't see it. That's shameful. I'll swap the +/- leads on the jacks.

Thanks

Chuck

[dai h.]

hey I've done the same thing, lol...

I was making something from my own drawing and it didn't work until I finally noticed the instructions to myself (which I made) were wrong.

[Chuck H]

Well I rewired the jacks and it's working great except... There's a very high resonant frequency that is slightly audible and wants to feedback with the pickups at high attenuation levels. The problem is proximity sensitive. If I can squash this problem the design is very transparent otherwise. I would guess it's the active componants. Any ideas? Maybe a .5uf shunt?... Where?...

Thanks

Chuck

P.S. I'll be out of town for a few days. So if I don't reply right away it's not because I'm rude (though maybe some other things are).

[dai h.]

"proximity" meaning if you get close with the gtr.? The reactive load is the same as a speaker and you will get feedback getting close to it. From my experience I think it's good to put the physical placement in order, that is, gtr. on one side to amp input, then amp output(speaker out), then reactive load, then speaker. Having the speaker signal (via the cable radiating and coupling) and the load near the gtr. and gtr. cable is not good for stability. I used to have the load facing the amp head with the spk. cable running near the gtr. cable paralleled to it and when I re-arranged the load to the side(as above) it was more stable. And a reactive load is supposed to be more liable to instability than a purely resistive. Even a long spk. cable can make a difference (as I understand it).

[Chuck H]

Absolutely Dai. The rub is that I want to install these IN the head cabinet with the amp chassis. The amp chassis is aluminum so there is no real EMF shielding. Placing the attenuator in the head cabinet makes the problem worse. I may have to go with a resistive attenuator for the built in application. I could try shielding the attenuator case before I give up.

Chuck

[Enzo]

Ever notice when you stand right up by your amp, your pickups pick up the power transformer field and you hear it as hum?

I suspect the inductors in your attenuator are radiating their field as the signal current flows through. Your pickups are picking up that field. Put a steel cage around the inductors.

I can't imagine such a thing inside the amp chassis working very well. it would be like the worst lead dress problem in a Fender multiplied many times. or so it would seem to me.

[daz]

Enzo, i don't think he's talking about putting it in the chassis, but into the cabinet like where you'd put a reverb pan.

[Enzo]

Oops, you're right, that is indeed what he said. I got a mental picture of this thing inside the chassis and never stopped to think it through.


Well then, never mind.

[Chuck H]

Well, Being as my chassis for both the amp and the attenuator are aluminum, and this is a head cabinet so it's close no matter what, all the same issues exsist as if I were putting it in the chassis.

As I said, I might try shielding the attenuator. But I like the idea of shielding the inductors better. I can run a ground lead outside the attenuator case and ground it to the chassis right at the AC cord ground. That should do it since it almost works now.

My other option would be to use only resistive elements for the on board attenuators. But ear testing by bypassing the reactive elements showed that the reactive design sounds much better than a purely resistive one.

Thanks

Chuck

[Enzo]

I don't know what you are using. Would they fit under a Fender capacitor vault cover? Would they fit under an old 8EB2C1B reverb pan body? Put a bottom on either of those and you have a pretty good shield. Ready made.

[dai h.]

I have no idea if it'll work, but I'd think the same principles would be involved in placement (try not to let it feedback or radiate backwards towards the input especially into sensitive portions). (As an obvious example) combo amps do exist (where of course the speaker/s can be close to the amp), so maybe it can be done. And even if it's not magnetic shielding, something like aluminum still might help.

As far as orientation of inductors in a reactive load (from examining my Power Brake) I can tell you that in the PB, the inductors do appear to be placed so as to not affect ea.(they are also pretty close in the PB since there isn't a whole lot of space). In my Marshall SE100 though (which has pretty much the same reactive load, plus two networks for attenuation-- -6 and -12 which are connected in series for -18dB) all of the inductors don't appear to be (some are side by side in the same orientation--although that may be a design defect possibly).

you can check out some innards pics here:

powerbrakepics

(geocities is being eliminated BTW, so anyone who wants something from a GC site may want to save it in case it disappears entirely although with these in particular I can upload them, etc. since they are mine.)