Cathode Follower?

[dwhutchens]

Hey guys,
I wanted to try driving a tone stack with a cathode follower to tame some of the gain. This circuit started life as a 5e3 then morphed into something else and then a princeton reverb minus the reverb and tremolo but that was unstable until I cut the gain way down with a voltage divider. Then I though about a cathode follower. It works and sounds pretty good about half way up the volume then the distortion sounds constipated. I put the scope on it and it appears the cathode follower is only amplifing half of the signal. What should I adjust or change in the circuit to balance it out?
Thanks,
David

[tubeswell]

Your CF setup seems unusual. I note you are driving it from a normal Voltage Amplifier stage. In that case is it really a follower? - it looks more like the lower half of a PI. I thought a typical CF stage was different. I'll post some pics

[tubeswell]

Here they are (Courtesy of Mr Valve Wizard) all set up to operate slightly differently.

The 5F6a direct coupled pair is different again. (The grid voltage is quite high)

[Enzo]

A cathode follower is a cathode follower, regardless of where its input signal comes from.

If you replace a gain stage with a cathode follower, then yes, the overall gain of the amp will be reduced. But since a CF has a gain of just slightly under unity, adding one into a circuit won't reduce the overall gain.

What David's circuit lacks is what all of Tubes' examples included - a bias arrangement for the grid. The way it is, the poor tube is biased almost to cutoff, so that is why you are losing waveform.

Take the first example on the left in Tubes' post and convert your CF to that. SO add Cin, Rg, and Rb. Cin goes between the volume pot wiper and the grid - That way the volume pot won't drag the grid DC voltage down. For the resistors, try the common 1M for Rg, and make Rb 1.5k. You can adjust them at some other time, but that should make it work. Cout is not needed because the three caps in the tone stack will serve the purpose.

[dwhutchens]

Thanks guys, great replies. I learned this stuff in high school some 30 years ago and obviously I wasn't paying a log of attention, maybe that's why I was getting Cs and Ds on my report card I'll give this a try as soon as I get a chance.
David

[dwhutchens]

OK, I did what Enzo suggested and made the changes to my circuit and now the waveform looks normal. The amp sounds very good with the volume half way up. I tried to crank it all the way but the wife thought her phone call was more important. I'll have to wait till she's out of the house to give it a good test but it looks promising. I've got about 415v B+ with a GZ34 or about 360v with a 5Y3. I think I'm liking the sound with the GZ34 a little better but not sure yet. Thanks again guys.
David

[dwhutchens]

I just thought of another question. Would there be any advantage or disadvantage to moving the volume pot to after the tone stack? I thought about trying that to see if it made a difference, it might help neaten up my wiring a little.

[Mal88]

From the way that I see it, you now have a 100 k ohm cathode follower (V1b) off a 100 k ohm pre-amplifier (V1a), and you have two high pass circuits crippling your bass response (0.02 uF and the 1 M ohm volume pot), and the feed into the cathode follower (0.02 uF and probably 2 M ohm); when you already have the cathode of V1b raised significantly above the heater voltage.

How about you simply connect the plate of V1a to the grid of v1b and remove the 0.02 uF, 1 M ohm vol pot, 0.02 uF coupling capacitor, 1 M grid bias, short out the 1.5 k cathode self bias; now put the 0.02 uF and volume conrtol off the V1b cathode and feed this directly to your tome controls.

I would also be inclined to drop the HT feed to the pre-amp (V1) with say a 10 K 2 W resistor and decouple this with another electro.

That ought to keep the wife outside and on the phone!!

[dwhutchens]

What effect will dropping the plate voltage and decoupling V1 from V2 have?
Thanks,
David

[Gtr_tech]

You're feeding all the pre stages with the same node. Isolating them a little will keep the supply rail from being modulated. Look at the Mar$hall JCM800 schemo for one way to do it. Kevin O'Connor suggests doing it with a resistance in series with the plate load and a lytic to gnd at the junction of the 2 resistors. It works well and I use that method on PCB amps where its prohibitave (well, not really, but a PITA) to alter the layout.

In short, you just want to give the supply wiggle somewhere to go so the rail stays steady.

[dwhutchens]

5e3 modded into oblivion!

Being one that just can't leave well enough alone I ripped this thing apart again. I decided I wanted a little more gain than I was getting with the cathode follower. I put it back to a 5e3 layout with V1 cathodes split and tried revoiceing the normal channel. I couldn't find a tone I liked so I decided to go back to the drawing board on the TMB tone stack. I found some info on using a dual gang pot to turn up the gain on the first and second stage at the same time. I had allready tried two seperate pots but found that I was turning them up about the same amount anyway. I started with a 1meg dual pot but had way too much gain, blocking distortion set in. I switched to a 500k dual pot with a 470k in front of each one and now I'm getting maximum gain without the blocking distortion or getting unstable. If I'm right I'm getting about 11~12 watts clean power out. The amp is fairly clean with the volume about half way, a few degrees more and it gets warm and crunchy, maxed out and you can get a note to feed back indefinitly I think I've used a half a pound of solder on this thing Here's the diagram, I think I have all the values right, its starting to all run together.