caps across plate resistors - questions...

[HTH]

I've used caps across plate resistors to tame extreme top end in higher gain amps (and to control oscillations).

I've always just soldered them directly across the plate resistors, but I notice on a Boogie Dual Recto, they've got a 4700pF cap from the B+ end of the plate resistor with the other end connected straight to ground.

What difference is there doing it this way?

[dai h.]

my understanding was that they are the same but just that the cap across the plate shunts highs (from the output signal) to *AC* ground (through the local filter cap, while the other is to DC ground.

[HTH]

thanks for that Dai,

my next question would be - what difference does that make in practice?

I guess I should just get some croc-clips out and see for myself, heh.

[dai h.]

you mean is there is a difference in sound between the two? I don't think I've tried comparing myself but I faintly recall Bruce saying they are the same. I'd guess putting across the plate R is mostly about convenience (i.e. you don't have to get long leads to stretch out).

[Bruce / Mission Amps]

you mean is there is a difference in sound between the two? I don't think I've tried comparing myself but I faintly recall Bruce saying they are the same. I'd guess putting across the plate R is mostly about convenience (i.e. you don't have to get long leads to stretch out).

Yes the small snubbing cap from plate to ground will be the same as across the plate resistor ... as long as the plate resistor is fed from a capacitor filtered B+ node because the B+ filter cap is an extremely low (AC) impedance to ground.
Any AC (audio) coupled by the tiny snubbing cap across the plate load resistor...at the "plate to plate load resistor junction" will see the + end of the filter cap as ground too... but the bypassing 4700pF cap is so small, only the higher freqs will see the cap as a low impedance path to the grounded filter cap.
As dai H said, just grounding the .0047pF cap will do the same but you have to find a place to ground it, so just jumping around the plate load resistor is much easier.

By the way, a grounded .0047uF cap from the plate is pretty large value.
I didn't look to see what the actual coupling cap value was but the two will also act as a from of an AC voltage splitter (yes frequency sensitive) and limit the gain too... but because it using just a little cap, it will just start limiting the gain from some frequencies and beyond.
In other words at some frequencies and beyond, the plate of the tube will be completely AC grounded and have no output... like ultrasonics that can result in instability in later higher gain circuitry... but the lower frequencies from that plate will see the little .0047pF cap as a much larger impedance and just shoot right through the actual coupling cap.

Kinda wordy but you all get it, right?

[martin manning]

Yes the small snubbing cap from plate to ground will be the same as across the plate resistor ... as long as the plate resistor is fed from a capacitor filtered B+ node because the B+ filter cap is an extremely low (AC) impedance to ground.
Any AC (audio) coupled by the tiny snubbing cap across the plate load resistor...at the "plate to plate load resistor junction" will see the + end of the filter cap as ground too... but the bypassing 4700pF cap is so small, only the higher freqs will see the cap as a low impedance path to the grounded filter cap.
As dai H said, just grounding the .0047pF cap will do the same but you have to find a place to ground it, so just jumping around the plate load resistor is much easier.
---snip---

This had me confused in the previous posts because I didn't see how a cap bypassing the plate load resistor could act the same as a cap from the B+ side of the load to ground. Bruce, you mention a cap from the plate side of the load resistor to ground, and that seems like it would have the same effect- dumping HF so it won't produce any signal voltage on the plate side.

The Mesa Dual rectifier's pre-amp in fact has both a 1n cap bypassing a 100K plate load resistor, and then on the adjacent (previous) gain stage, a 5n cap from the b+ side of its 100K load to ground. Those two stages are not decoupled, so they both see the 5n to ground.

The only thing I can figure is that the 5n to ground on the B+ keeps these two stages from talking to each other, HF-wise.

Am I makin' any sense here?

MPM

Schematc:

http://www.freeinfosociety.com/elect...iew.php?id=507

[Steve Conner]

A cap from the B+ rail to ground is just a bypass capacitor the same as any other. It doesn't (shouldn't!) have any tone shaping effect.

Bruce said that "the B+ filter cap has an extremely low impedance to ground". The B+ rail is supposed to have zero signal voltage on it, and hence be the same as ground from the point of view of AC signals. So, the snubbing cap under discussion should have the same effect, no matter if it's connected from plate to ground, or plate to B+. Bypass capacitors are what make this happen.

In standard tube amp circuits, the electrolytic filter cans are good enough for bypassing. Modern high-gain designs are more sensitive and the designer might need to add extra capacitors in other places to stop unwanted coupling between stages.

So yeah, it probably does "stop the stages talking to one another HF-wise".

[HTH]

thanks for all the replies, I feel like I have a better understanding of whats going on now.

[Konny]

While talking about those caps . . .
what´s the deal with a small cap (around 100pF) from anode to the cathode such as Marshall did in some of their 2203 models?

Does this work as some sort of local negative feedback?

Cheers,
Swen

[martin manning]

While talking about those caps . . .
what´s the deal with a small cap (around 100pF) from anode to the cathode such as Marshall did in some of their 2203 models?

Does this work as some sort of local negative feedback?

Cheers,
Swen

After a little searching around the web I find two suggestions: These could prevent HF oscillations and/or RF pick-up, and also that the effect will be essentially the same as placing the cap from plate to ground, since the cathode resistor is bypassed to ground with another cap. It's easy and sound practice to place it directly on the tube socket, too.

The reference books I have don't seem to cover this (unless its in the RDH4 and I just can't find it!), but my library is still small.

MPM

[Guitarist]

At the risk of sounding incongruent to the amazing answers here, my memory/experience prefers the tone and sound of sending the glitch across a plate cap vs other methods. I remember A/B'ing that on a couple of amps. If you don't have a high voltage cap then use a lesser one off the treble-send perhaps.

680 pF is a good value by the way.

[Bruce / Mission Amps]

At the risk of sounding incongruent to the amazing answers here, my memory/experience prefers the tone and sound of sending the glitch across a plate cap vs other methods. I remember A/B'ing that on a couple of amps. If you don't have a high voltage cap then use a lesser one off the treble-send perhaps.

680 pF is a good value by the way.

That may be true but depending on the plate load resistor, the 680pF cap can have quite a different effect.
Example:
100K plate load and parallel 680pF has a -3dB point of about 2350Hz
The same 680pF cap across a 220K plate load resistor would have a -3dB of around 1064Hz.
If you happen to have a guitar, amp and speaker combination that has some unpleasant tone at around twice the 2400Hz, or 5000Hz, the 680pF cap across the 100K plate load would sound better to you as at that point it could be down -6dB.
These simple filters are not sharp, notchy or extreme in anyway, but are a bit more subtle, broad in nature, as they are rather low Q filters.
However, if the right RC combo(s) are installed in the right places, they work well enough without killing the tone too much.

[HTH]

At the risk of sounding incongruent to the amazing answers here, my memory/experience prefers the tone and sound of sending the glitch across a plate cap vs other methods. I remember A/B'ing that on a couple of amps. If you don't have a high voltage cap then use a lesser one off the treble-send perhaps.

680 pF is a good value by the way.

Bruce, how are you working the math on this? - I'm terrible at maths, but I should really make an effort as this is useful stuff.

Thanks, Ian

(edit: this was supposed to be in reply to Bruce's post with the 3dB rolloff points)

[Earl]

He's using the -3dB cut-off equation, Fc = 1/(2piRC). Althougm that plate resistor is effectively in parallel with (in the case of the 1st amp stage) the tone stack and the volume pot.

I made a mathematical model of a Champ, and will see if I can post a couple of plots that show the response with and without a 680pF cap across a plate resistor.

The 1st image is without the cap...sorry about the image size, haven't figgered out how to do the thumbnail thing yet.

[HTH]

Thanks for that Earl, I can see the steeper rolloff in the top end with the 680pF bypass cap.

To be honest, theory is great but most of what I know comes from trial and error - I would never self apply the title 'tech' my myself in a million years (lmao)