Attenuator treble bypass cap problem, help please...

[Regis]

Some of you know I like to build my own DIY amp attenuators, like my Ampwell House and Havanatone.

In my Havanatone designs I use a dual ganged L-Pad, and wire it to switch one gang in and out to achieve 4 or 8 ohms. I install a bypass cap on the switch to get some treble back when the attenuator is dialed down low. The cap is usually around 12-15 uf, but I've gone higher and lower.

I discovered an issue where the cap doesn't work when I switch in both gangs to get a 4 ohm impedance.

In the 8 ohm setting with just one L-pad section enganged the treble boost provided by the cap is noticable and sounds good. When I switch the other L-Pad gang into get 4 ohms it disappears. It is slightly apparent when the attenuator is dialed down very low, but not much.

For the life of me I can't figure out why this happens. Something related to the impedance maybe? I've tried higher caps, up to 47mf but no difference. I put the cap across the bypass switch.

Any help would be appreciated.

Regis

[Chuck H]

Yup. It's the impedance of the circuit nullifying the effect of the bypass cap. Two things are happening when you switch to a lower impedance. First the effect is halved because the load the cap bypasses is halved and second the frequency knee of the cap shifts up. The series resistance of NP caps sucks. If you can get cap that specs better for this it may help a little and perhaps add to the impedance switch a circuit that parallels a second cap into the circuit to double the uf in the 4 ohm position.

Chuck

[Regis]

Thanks Chuck. I thought it might be something like that. I think I tried another cap on top of the one there and didn't hear any difference but I'll try it again. Do you recommend another type of cap other than a non polarized?

[rockin roy]

It's been a while since I took an L-Pad apart but if I remember right there was an large value resistor, maybe 60 ohms and a low value one, about 16 ohms, As you adjust your volume the amp would always see 8 ohms. Now you have hooked two of these in paralel and thrown in a cap and it doesn't work? go figure.
Does the actual input resistance track true? Always 8 or 4 ohms.
Do you get an output with the cap removed?
I'd use 47 mfd caps for motor starting or making a 220 volt fan run on 115. Maybe .047mfd?

[Chuck H]

Now you have hooked two of these in paralel and thrown in a cap and it doesn't work? go figure.

Why shouldn't it work??? Have you looked at the diagram. Two parallel 8 ohm L-pad circuits absolutely do make a 4 ohm L-pad. Why wouldn't they? And the cap is situated in the same place relative to the circuit whether one or both L-pads are in use. I guess you haven't made your point well enough. Help me uderstand the problem you see here.

I'd use 47 mfd caps for motor starting or making a 220 volt fan run on 115. Maybe .047mfd?

What's your point about 47uf being an appropriate value for motor starting or filtering a voltage doubler??? Does this somehow arbirtarily make it a wrong value for an audio circuit? Your gonna need to explain this to me also I guess. I seem to be missing the point all over the place here.

FWIW a .047uf cap in the above circuit has a -3dB knee above the range of human hearing!!! How is that any good for anything in an audio circuit???

Sorry (or not) to get up in your grill but your post is really assumptive and glib. You don't have the knowledge to counsil or take that attitude with either Regis or myself.

[dai h.]

yeah I think what Regis has should work (factual basis) :

http://www.colomar.com/Shavano/lpad3.gif

so, for a dual L-Pad, the sections are paralleled, making each part 1/2 the R. IIRC, this is also in the instructions on dual L-pads(that impedance can be varied).

[Regis]

Why shouldn't it work??? Have you looked at the diagram. Two parallel 8 ohm L-pad circuits absolutely do make a 4 ohm L-pad. Why wouldn't they? And the cap is situated in the same place relative to the circuit whether one or both L-pads are in use. I guess you haven't made your point well enough. Help me uderstand the problem you see here.

What's your point about 47uf being an appropriate value for motor starting or filtering a voltage doubler??? Does this somehow arbirtarily make it a wrong value for an audio circuit? Your gonna need to explain this to me also I guess. I seem to be missing the point all over the place here.

FWIW a .047uf cap in the above circuit has a -3dB knee above the range of human hearing!!! How is that any good for anything in an audio circuit???

Sorry (or not) to get up in your grill but your post is really assumptive and glib. You don't have the knowledge to counsil or take that attitude with either Regis or myself.

Yes I have measured the two sections paralleled, it does make 4 ohms. That was one of the first things I checked, thinking I had wired something wrong. In both the 4 and 8 oms settings it tracks very well, keeping the two impedance's correct. I did try a 47mf at the 4 ohm setting and there was very little difference. I haven't had time to parallel another 15mf on top of the one that is there, I'll try to get to it tomorrow or Friday. Thanks for the help fellas, Id still like to get this figured out.

[Regis]

When I switch the other L-Pad gang into get 4 ohms it disappears.

What I mean is that the treble boost disappears, the attenuation of the L-pad still works fine.

[Chuck H]

Regis, Part of the trouble with the treble boost effect dissapearing is that the ESR on bipole e caps is so bad. At least a couple of ohms most of the time. With this info it's easy to see that switching from the 8 ohm setting to the 4 ohm setting has a profound effect on the ability of the treble to bypass effectively. Sort of like an 8 ohm resistor jumpered with a series 3 ohm resistor and a cap. Now make it a 4 ohm resistor jumpered by a 3 ohm resistor and a cap. If you can use a "better" cap that switches in when using the 4 ohm setting that could fix the problem. Mouser has several film caps in 10uf/100V. A pair in parallel switched in parallel when going to 4 ohms would probably do it. If you just replace the cap thats in there with the better caps the effect may be to bright in the 8 ohm setting.

HTH

Chuck

[Regis]

Regis, Part of the trouble with the treble boost effect dissapearing is that the ESR on bipole e caps is so bad. At least a couple of ohms most of the time. With this info it's easy to see that switching from the 8 ohm setting to the 4 ohm setting has a profound effect on the ability of the treble to bypass effectively. Sort of like an 8 ohm resistor jumpered with a series 3 ohm resistor and a cap. Now make it a 4 ohm resistor jumpered by a 3 ohm resistor and a cap. If you can use a "better" cap that switches in when using the 4 ohm setting that could fix the problem. Mouser has several film caps in 10uf/100V. A pair in parallel switched in parallel when going to 4 ohms would probably do it. If you just replace the cap thats in there with the better caps the effect may be to bright in the 8 ohm setting.

HTH

Chuck

Thanks Chuck. I tried paralleling another cap but it made no difference at all. If you don't mind would you provide a link or a catalog page number to the caps you are talking about?

[Chuck H]

I tried paralleling another cap but it made no difference at all.

Hmmm... That should cut any resistance from the cap in half. I would think that would make some difference. But it can't hurt (other than a little $$$) to try a film cap anyhow.

ECQ-E1106KF Panasonic Electronic Components Suppression Film Capacitors

This is the cheapest one I could find (can't find my paper catalog??? Never been without it but did some organizing ) It should work fine.