Help with cathode follower circuit

[gottlieb]

This is a question for the tube gurus out there.

I hope you could help me with a problem I'm having with a guitar tube preamp I'm building. It's a distortion circuit for guitar with two 12AX7s, here's the schematic.

http://img142.imageshack.us/img142/3327/schematic.gif

The problem is the output impedance.

I break the circuit at the 470K/470K mixer/attenuator after the volume pot. I like the effect the two resistors have in the sound, they lower the signal level and make the overall volume of the unit more controllable.

It sounds great with no hum or noise at all with the output at the 470K resistor junction if I connect it to an amp's #1 input (1M input impedance), and the volume control works as it should (no volume on "0"). The distortion part of the circuit also works as it should.

But when I connect it to an amp's #2 input (68K input impedance) or any other low input impedance device the sound becomes weak and thin.

The solution I found was to wire the spare triode as a cathode follower so that I could connect the preamp to any other device, but every CF configuration I've tried is noisy, leaving a constant buzzing on top of every note, mosquito-like.

Even with the volume control on "0" the buzzing is present at the output.

I've tried shielding the CF tube, locating the transformer away from the circuit (1 meter away!), fixed bias and cathode bias CF configurations, lowering the grid leak resistor value from 1M to 100K, 47K and 56K load resistors, 1K up to 2K7 bias resistors, bypassing the bias resistor with a capacitor, etc.

I'm obviously doing something wrong, and I really would appreciate your opinions on the subject, other places in the circuit to locate the CF, or a schematic of a CF configuration that could work here. HT is 290V, I have a spare 12AX7 triode left.

Many thanks for yor help!

[bacon665]

Idk what the cause of it is but if you can isolate the frequencies of the "misquito" you can make filters for it.
For instance if its 300hz-360hz you would have two filters parralel to the output
a lowpass and a high pass. design the lowpass for some number idk around 285-290 because passive stuff is never exact and the high pass you can design for 360 and round up a little. that will cut out the band of noise. But you also have the problem of cutting out [art of your audio

[gottlieb]

Thanks bacon665 for your suggestion, but I'd rather cure the problem not the symptoms!!!

The preamp sounds great and performs at it should without the CF, but it's very impedance-dependent...

I've already run out of ideas of what may cause this issue, maybe the distortion circuit messing with the CF?

Any help is greatly appreciated.

[leadbelly]

OK, #1, you're making some basic mistakes with your output caps. Your caps are way too small for the impedances you are trying to drive. If you are trying to drive a 68K input you need caps > 1uF as a bare minimum.
f=1/(2*Pi*R*C)
The reason your 68K input with the 12AX7 gain stage sounded weak and thin might be as simple as that.

#2 you don't have a output terminating resistor on your CF. What does your grounding look like have you troubleshooted for ground loops?

#3 what does your power supply look like? How much ripple? A simple 56K cathode loaded CF is not going to have wonderful PSRR.

[Merlinb]

Originally posted by leadbelly
OK, #1, you're making some basic mistakes with your output caps. Your caps are way too small for the impedances you are trying to drive. If you are trying to drive a 68K input you need caps > 1uF as a bare minimum.
f=1/(2*Pi*R*C)
The reason your 68K input with the 12AX7 gain stage sounded weak and thin might be as simple as that.

He is using 0.1uF caps. Into 68k that gives you a roll off of 23Hz, which is waaaay low for guitar. More than low enough.

#2 What does your grounding look like have you troubleshooted for ground loops?

Grounding certainly sounds like the next likely source of noise, since the transformer isn't it.

#3 what does your power supply look like? How much ripple? A simple 56K cathode loaded CF is not going to have wonderful PSRR.

The PSRR will still be many times better than the previous gain stage, so is unlikely to be the problem. Possible though.

[dsavitsk]

grid stopper?

[jjman]

I'm curious as to which guitar amp has a 1meg imp. on #1 and 68k imp. on #2 ?

For example, a common Fender has 1meg on #1 with 2 parallel 68ks as grid stoppers.

When plugging into #2, one of the 68ks "moves" becoming in line with the 1meg, resulting in 1.068k imp and 68k grid stopper.

e.g.:

http://www.schematicheaven.com/fende...b763_schem.pdf

[Bitrex]

The 12AX7 makes a lousy cathode follower; it has too high a dynamic plate resistance even in a cathode follower configuration. It's also going to have trouble on the negative swings as it has to sink current through that 47k cathode ressitor. The sound you're hearing might be the result of asymmetrical distortion from the cathode follower - but the fact that you hear it when the volume is at 0 seems to point away from that conclusion. Maybe we need some better adjectives to describe the noise - is it like power line hum? What kind of circuit are you using for the heaters?

In any event, I don't think a you're going to ever get good results with a 12AX7 as an output cathode follower, particularly if you're using a long cable with lots of capacitance. A tube with a lower RP like a 5687 would probably be better in this application, or something like a SRPP: http://www.freewebs.com/valvewizard1/srpp.html - but then you have that spare 12AX7 hanging there. Hopefully there will be further suggestions posted!

[gottlieb]

Originally posted by jjman
I'm curious as to which guitar amp has a 1meg imp. on #1 and 68k imp. on #2 ?

For example, a common Fender has 1meg on #1 with 2 parallel 68ks as grid stoppers.

When plugging into #2, one of the 68ks "moves" becoming in line with the 1meg, resulting in 1.068k imp and 68k grid stopper.

e.g.:

http://www.schematicheaven.com/fende...b763_schem.pdf

Sorry, impedance on #2 input is 136K, not 68K. It's still a lot lower than 1M and causes volume loss without the CF:

http://www.el34world.com/charts/fenderservice2.htm

Originally posted by Bitrex
The 12AX7 makes a lousy cathode follower; it has too high a dynamic plate resistance even in a cathode follower configuration. It's also going to have trouble on the negative swings as it has to sink current through that 47k cathode ressitor. The sound you're hearing might be the result of asymmetrical distortion from the cathode follower - but the fact that you hear it when the volume is at 0 seems to point away from that conclusion. Maybe we need some better adjectives to describe the noise - is it like power line hum? What kind of circuit are you using for the heaters?

In any event, I don't think a you're going to ever get good results with a 12AX7 as an output cathode follower, particularly if you're using a long cable with lots of capacitance. A tube with a lower RP like a 5687 would probably be better in this application, or something like a SRPP: http://www.freewebs.com/valvewizard1/srpp.html - but then you have that spare 12AX7 hanging there. Hopefully there will be further suggestions posted!

I would prefer using the spare triode and socket I already have. What about using a 12DW7 (half 12AX7, half 12AU7)?

The noise sounds like there's two signals: the original with another very distorted signal on top with lots of treble. With volume on "0", when I play only the distorted signal can be heard.

Sorry for the explanation but English is not my mother tongue...

I'm using regular 6.3V heaters, elevated 60V. Without the CF there's no noise at all, so I don't think that's the culprit.

[Merlinb]

Sounds obvious, but have you tried plugging in a different valve?

In any case, the fact that you've tried different versions of cathode follower and the noise is there even when not playing would seem to imply a ground problem.

I think we may need photos of the interior...