Safe to put V1a coupling capacitor on a switch for variety?

[Mr. Bill]

I'm building another prototype amp at the moment and want to make lots of things switchable just to explore options. I'm basing the circuit around a JCM 800 single channel. One option I'd like to have is to make the coupling cap following V1a switchable between 3 different caps; .001, .0022, and .022. Is it safe to just rout those to the front panel and put them on a 3 way selector? There's roughly 260 volts passing through it.

Thanks for thoughts and ideas.

[Dave H]

I'd like to have is to make the coupling cap following V1a switchable between 3 different caps; .001, .0022, and .022. Is it safe to just rout those to the front panel and put them on a 3 way selector? There's roughly 260 volts passing through it.

Thanks for thoughts and ideas.

You could put the 0.022u followed by a 10M resistor to ground on the circuit board then run a wire from there to the switch with the 0.001u and 0.0022u mounted on the switch. That way all the switching is at ground and there shouldn't be any pops when switching.

[Chuck H]

What Dave suggests will work fine. You get .002u (rather than .0022u) and .00095u (rather than.001u) and that should be close enough. More complicated (but more accurate) would be to decouple the circuit from DC with something like a .47u film cap and follow with the switching for the .022u, .0022u, .001u caps. You'll still need to reference the caps to 0V with a resistor to avoid popping. Another option would be to series the whole arrangement. A .022u cap followed by a .0022u cap (paralleled with a 220p cap) followed by another .0022u cap. Then you tap from the junctions. And you STILL need the 0V reference resistor at the end of the string. This would give you .022u, .00218u (pretty damn close!), .00109u (also pretty damn close).

Lost of ways to approach this without DC on the switch

[Mr. Bill]

Thanks for the help guys. While I've come pretty far tinkering with these circuits, and can throw together superleads and 800's like legos at this point, I think your suggestions are a little over my head. Being a coupler cap, it has no direct reference to ground but feeds the next stage. If I tapped each one to a 1M to ground, wouldn't I end up creating a voltage divider with the 470k feeding the gain pot and drastically alter the gain feeding the next stage? I'll attach a picture of what I mean. Maybe I'm just not getting it.

[Dave H]

I was thinking you could do it like this.



I've done it like this to switch in octaves. It gives roughly 10n, 4n7, 2n2, 1n, 470p

[Chuck H]

Plus, a rotary switch such as might be used in diagram two would allow the use of a standard knob that matches the others, or not (maybe a pointer type knob) for a nice visual appeal

[Mick Bailey]

When switching caps in an audio circuit you need to ensure that there is no potential difference between the caps, otherwise this difference will cause a pop when switching. The way I do this is to use 10M Ohm equalizing resistors between the lugs of the switch.

Edit: Take a look at the EF86 channel of a Matchless DC30

[Chuck H]

Positively! (pun intended)

But IMHE when actually building or modding things can get crowded very quickly. I've burnt up more than one switch over soldering clunky. crowded lugs. Sometimes it's easier to simply make sure there is a 0VDC reference resistor at either end of the caps as Dave has done. I haven't had a popping problem doing it this way.

[Chuck H]

Here's the arrangement I was talking about. It allows for the stock tone one position, but skips the 470k/500p circuit for the other positions. I don't think keeping that in the circuit with the smaller coupling caps would be to anyone's liking. You will need an on/on/on toggle (can be hard to find) or a rotary switch for this.

[Mr. Bill]

Ah! I see now. I was planning to also follow that switch with another bypass switch that would take the 470k/ 500p out of circuit then put two different bright cap values on a push pull 1M pot as well. This would be used in conjunction with another switchable cap or two in the NFB loop so I could experiment with varying degrees of thinning out the front end and beefing up the NFB to compensate. In theory, I'd also like to add switchable smoothing caps on the plate resistors of the first two stages, but that might add noise if I'm not careful.

I have worried about the idea that juicing up the front end too much might also require grid stoppers on following stages for stability, and making those switchable wouldn't make too much sense to me.

I'm guessing the switchable cap in the feedback loop could be put on a direct switch without the 0 volt reference, correc

[ric]

Ah! I see now. I was planning to also follow that switch with another bypass switch that would take the 470k/ 500p out of circuit then put two different bright cap values on a push pull 1M pot as well. This would be used in conjunction with another switchable cap or two in the NFB loop so I could experiment with varying degrees of thinning out the front end and beefing up the NFB to compensate. In theory, I'd also like to add switchable smoothing caps on the plate resistors of the first two stages, but that might add noise if I'm not careful.

I have worried about the idea that juicing up the front end too much might also require grid stoppers on following stages for stability, and making those switchable wouldn't make too much sense to me.

I'm guessing the switchable cap in the feedback loop could be put on a direct switch without the 0 volt reference, correc

I like your switch stuff idea, I'm into that too. I don't like the sound of switching plate resistor bypass caps, you're putting plate high voltage on a switch you touch, not good.

Also grid resistors like to be right on the socket lugs as close as possible to the tube. By bringing the circuit out to the faceplate you may have instability problems.

There are a lot of techs here and I'm not one of them, but I believe these are valid cautions.

[Dave H]

But IMHE when actually building or modding things can get crowded very quickly.

Yes, that's why I only used one 10M resistor. It was difficult enough fitting xicon caps on a little rotary switch. I've made three now and they work well enough without popping.

[Chuck H]

Changing coupling caps and bright caps doesn't beef up the gain any. You will get a little more gain in the low mids (and a little less in the bass) by bypassing the 470k/500p circuit and switching to the smaller coupling caps. A smaller bright cap on the volume control will actually reduce gain in the upper mids at volume pot settings below max. If by the cap in the NFB loop that you mean the presence cap, that cap does isolate signal from DC. That's why the old style Marshall circuit can have a scratchy presence pot. So you do need to arrange or design that circuit specifically to avoid switch noises. Increasing that cap value will increase mid gain and distortion in the power amp. Decreasing the value would reduce gain and distortion in the mids. Adding "smoothing caps" to the plate resistors will reduce gain in the treble. Nothing you propose so far will increase gain and cause blocking distortion. In fact most of what you propose will actually decrease gain in one frequency or another except for bypassing the 470k/500p circuit. Doing this reduces the voltage division of low mid and bass frequencies with the volume controls load, BUT, since the coupling cap value is reduced the low bass frequencies get cut, having the effect of increasing mids.

[Mick Bailey]

Just thinking about amps that use switched coupling caps and another classic is the Orange OR120/OR80 that works nicely.

[Mr. Bill]

The NFB cap would go right before the OT tap as a resonance switch. I've also found I like upping the plate resistor on V1a in SPL's. This tends to at least sound to me like it adds gain and needs to be controlled downstream to prevent oscillation. Again, I'm still learning, and most of these tweaks are by ear with safety checks as I learn. Here's a schematic of my last build which sound great but is a one trick pony. (I'm a studio guy, so I like one trick ponies).

Pardon the crude nature of my drawings, they're just for my reference so I can remember what I like and don't like.