Why no plate R B4 cathode follower?

[iknowjohnny]

In typical marshall style amps why is it the plate from the stage before the cathode follower goes directly to the CF's grid without any resistor or cap? And if you put a r or cap there it sounds horrible like bad distortion. Please explain.

[Bob-I]

The plate load resistor is there on a common cathode gain stage to change a current modulation into a voltage modulation. As the grid varies, the current through the tube varies. V=IR so vary the current across a fixed resistor and the voltage must change.

The cathode follower stage uses the cathode as the output point so the cathode resistor is used in this formula. Change the plate voltage through a resistor and it all goes to hell (maybe someone more technical can explain this better)

[Jana]

you can't put a cap between the plate of the first stage of the cathode follower and the second. The grid of the second stage connects directly to the plate of the first. The voltage on the plate of the first stage sets the bias point of the second stage. Do some checking of voltages and it will make sense. You should find that the voltage of the grid of the second stage is a few volts lower than the voltage on the cathode of that stage, hence --- bias.

Naturally you can't put a bypass cap on the cathode of the second stage since that is where the signal is tapped off of. You can bypass the first stage cathode, early plexi's had a .68uf there. I call it "The Jimi Cap."

[iknowjohnny]

thanks. No, i wasn't going to try a cap on the cathode where the tone stack is connected. i meant between the plate and grid to the CF. I already have a cap across that stage's 820R cathode resistor. Don't know why more designs don't have that because it reacts really nice with a certain spongy feel that you don't feel when bypassing first stages.

if you don't mind, instead of starting another thread i'd like to ask another question thats not related. In a 2204 type circuit why is it that the post tone stack master causes the tone to get progressively treblier as you turn it up which causes too much high end and gets harsh? I have found using a couple caps at various points in the circuit to ground before the master helps eliminate that pretty much altogether, but i think if there was a way to implement a master that doesn't cause that would be a better way that would inhibit great tone less. I can't use a post PI master because when i try the the PI distorts horribly. can anyone explain to me why it acts like this?

[Jana]

well, if you look at the tone stack and the master volume as a unit, then what is happening? The 470pf treble cap in conjunction with the master volume is like a tone control in a way. With the master down the treble frequencies which come through the 470p have a lot of resistance to get through before getting to the PI. As you turn up the master, there is less resistance (naturally, that's the whole point of the master). You could either, A. add a bright cap on the master to make the response more even across its range, then cut more highs out in your circuit to compensate, or B. turn the treble down as you turn the master up.

[iknowjohnny]

thing is, the point at where the treble is heading to the master isn't just treble. It should be all frequecies as set my the rest of the controls. So the way it seems to me is that not only treble should see less resistance to ground, but all other frequencies too. so it should remain balanced. If thats not the case, what happens to the other frequencies coming thru that one point? Why are they too not equally attenuated or boosted with the setting of the master?

[Jana]

what is the capacitance of the following stage? (the phase inverter) I don't know off hand but I am guessing it is a couple hundred pf. The tube itself has I think 150 pf due to Miller effect. Then there are the other components in the circuit, wires, pixie dust, etc. 100 to 200 pf to ground means nothing to bass or mid frequencies. For treble and above it is significant. So, you have a resistor (the part of the master volume that is between the wiper and the "hot" terminal) going to a capacitor to ground. A hi cut filter! At low volume settings this resistance could be close to 1 meg with an Audio taper pot. That's why they invented bright caps, to compensate for this effect. A cap of about 100 to 250 pf as a bright cap on a 1 meg pot gives a balanced response across its range. At some point things got out of hand with the bright cap and it got larger and larger to give a treble boost. Treble sells amps, but that is another story.

[iknowjohnny]

Ok, so what you're saying is that when the master is turned up, thats the tone's natural state and when down the coupling cap is grounding highs? If this is the case then i think it says my pre has far too much high end, but i have endlessly tried to remove highs to no avail. it just makes it muddy. A .001 capto ground at the master helps a lot, but the tone just isn't quite right in a way thats hard to explain. There has got to be a simple way around this and i'm going to keep trying.

If i was right in accessing what you said, then i wonder if i recreated that same scenario BEFORE the master if that would help. in other words, measure how far from ground the coupling cap is when the volume is low where it sounds good. Then put a resistor of that value in series with a cap of the value of the coupling cap from the input lug of the master to ground. I guess that would kill the volume tho even if it worked, eh? And at low levels it would mud, tho i could probably make it come in gradually with a dual pot. Well, just thinking out loud here....

[Jana]

can you post a schematic of what you have now?

there is a very fine balance between removing the razor blades and turning to mud.

[iknowjohnny]

Ok, but this is NOT how the amp is now. this is just one point in time a week or so ago. the voltage dividers on the first 2 stages are gone and 100k resistors to ground in thier place. (trust me, it sounds better like that) The first stage and CF stage PSU node resistors have gone thru a million value changes and i thing V1 now has a 2.2k and V2 a 8.2k if i recall. The cut control is now gone. (it worked nice, but killed the amp's volume, and a smaller cap did nothing worthwhile) Theres probably a few things i'm missing, but it's pretty close to what you see plus the things i told you have changed. By the way, the V1 cathode and plate resistors and bypass caps have been thru every configuration imaginable. It's somewhere after the master....just gotta be.

[img603]http://i24.photobucket.com/albums/c39/dazco/el34amp.jpg[/img]

[Jana]

Are you saying that the 470k/470p and 330k after the 1M pot are gone and is a 100K resistor to ground? Same after the V1b coupling cap? If so, a .022 cap with a 100k to ground would certainly cut low end.

One thing that stands out is that the first stage is unbypassed and then the second two are, with small value caps to boost the highs. I understand the desire to boost highs to get a better distorted sound, but usually you boost highs, then distort instead of boosting the highs *IN* the stages of distortion.

To try to trim some highs, try a 100 to 250 pf cap or so from the plate to cathode of v1b and maybe v2a. Also a similar cap on the first stage wouldn't hurt.

[iknowjohnny]

Are you saying that the 470k/470p and 330k after the 1M pot are gone and is a 100K resistor to ground? Same after the V1b coupling cap? If so, a .022 cap with a 100k to ground would certainly cut low end.

Yes, but low end isn't the issue here. I would however like to cut it a bit so i may try that. but i mainly need to figure out the master volume issue.

One thing that stands out is that the first stage is unbypassed and then the second two are, with small value caps to boost the highs. I understand the desire to boost highs to get a better distorted sound, but usually you boost highs, then distort instead of boosting the highs *IN* the stages of distortion.

But where and how can i do that? With the 470pf/470k filters i had there before the amp sounded much less complex then now, and the lows and highs were about the same as now. As for the places i bypassed cthodes, i did it that way because thats the way it sounded best. other wise it got too sterile and high endy. And using larger caps only made it muddy.

I'll try those hi cut caps from plate to cathode. I've already tried then across the plates R's and across the PI plates but to no avail. The best so far was the cheesy cap to ground in the signal path in various places. Anyways, i'll try plate to cathode.
Thanks

[iknowjohnny]

Can no one here (mods) do anything about this freakin "debug mode" issue? I'm getting tierd of double posts. This time when i gor the debug message i actually closed all browser windows then reopened firefox and came back to see if it posted and saw nothing. So i copy my post back and submit and now theres 2 posts ! This is wacked.

[FUCHSAUDIO]

Cathode followers with gain sound like ass....lol. A standard CF is unity gain (actually a half db or so less than unity), and by adding a plate resistor, you're eventually going to get some gain. Depending on the cathode resistors, the amount of gain with vary, but in a normal FX loop or tone stack driver, you don't want gain.

The only place a CF with gain seems to work okay is in a driver (often called an accordian style phase splitter) where the plate is one phase output, and the cathode is the other. More than a few people did this (Traynor, Ampeg and others). I think the feedback around the power stages using this circuit made it more listenable.

[iknowjohnny]

Tried 500pf and 1000pf (all i had) from plate to cathode (pin 1 to pin 3 and pin 6 to pin on V1 and tried it also on V2A. i was shocked that i heard no change that i could be sure of ! I figured those values would be even more obvious.