As long as the ratio between the plate resistor and cathode resistor is correct, I've seen values between 100K and 220K.
Is there a simple way to explain what the difference in sound is, between a high plate resistor value and a low value (with corresponding cathode resistor).
Tommy
plate resistor values..
Moderators: pompeiisneaks, Colossal
-
Darkbluemurder
- Posts: 585
- Joined: Tue Jan 03, 2006 7:28 pm
Re: plate resistor values..
All things being equal the higher values sound like they have more headroom to me and take pedals a bit better.
Re: plate resistor values..
> As long as the ratio between the plate resistor and cathode resistor is correct
Ratio of tube plate resistance to DC resistor matters too. Imagine a 12AX7 with 1K plate and 1.5 ohm cathode resistors. Or try it. Kinda sucks.
> Is there a simple way to explain what the difference in sound is
No.
I'm not sure there is a complicated way to explain it.
Certainly nothing so quick as try-it-and-see(hear).
How do you write a song? With only D note and G note? (Yeah, I know some people do...) No, you try this note and that note, and a third note, and then after hearing it you go back and change the second note....
Cooking up an amp is actually simpler than cooking up a song. However it does take a bit longer to swap a resistor than to change a note. And it does cost 12 cents everything you ruin a resistor with excessive swapping; notes are free. But it is just a different way to noodle-around and come up with something you like.
Ratio of tube plate resistance to DC resistor matters too. Imagine a 12AX7 with 1K plate and 1.5 ohm cathode resistors. Or try it. Kinda sucks.
> Is there a simple way to explain what the difference in sound is
No.
I'm not sure there is a complicated way to explain it.
Certainly nothing so quick as try-it-and-see(hear).
How do you write a song? With only D note and G note? (Yeah, I know some people do...) No, you try this note and that note, and a third note, and then after hearing it you go back and change the second note....
Cooking up an amp is actually simpler than cooking up a song. However it does take a bit longer to swap a resistor than to change a note. And it does cost 12 cents everything you ruin a resistor with excessive swapping; notes are free. But it is just a different way to noodle-around and come up with something you like.
Re: plate resistor values..
Ok, så there is basically no difference in sound between a 100K and a 200K plate resistor. If there were, it would be easy to describe it.
Tommy
Tommy
Re: plate resistor values..
> basically no difference
No, it could go either way.
Darkblue's experience is as he says. But in other situations it could be different. What's good for one situation is bad for another. Depends a LOT on what comes after the stage, also what you put in.
No, it could go either way.
Darkblue's experience is as he says. But in other situations it could be different. What's good for one situation is bad for another. Depends a LOT on what comes after the stage, also what you put in.
-
Darkbluemurder
- Posts: 585
- Joined: Tue Jan 03, 2006 7:28 pm
Re: plate resistor values..
Exactly!PRR wrote:> basically no difference
No, it could go either way.
Darkblue's experience is as he says. But in other situations it could be different. What's good for one situation is bad for another. Depends a LOT on what comes after the stage, also what you put in.
I liked the 220k in the first stage of two Fenders I tried it in because it gave a more robust sound to my ears. I didn't like it in my modded Marshall (100k had less buzz than 220k), and I didn't like it in the Soldano Atomic 16 I once had. Reducing the 220k to 110k (I simply added a 220k in parallel) took out some of the spikiness in the sound. I still believe that the only way to find out is to try, listen and trust your ears.
Re: plate resistor values..
As PRR hints I don't know if thinking ratios is all that helpful in the long run. The two impedances have different functions in a voltage amp. Plate cct resistance interacts with rp and load and tube amplication factor to determine AC gain. The cathode restistance (in common cathode config) is there primarily to determine the operating point. A bit like 3D positional triangulation. The time constant of a bypassed cathode leg is more important from a tonal point of view than the size of the plate resistance as long as it all operates within limitations determined by B+.
Happy to hear contrary viewpoints.
Richard
Happy to hear contrary viewpoints.
Richard
Re: plate resistor values..
> As PRR hints I don't know if thinking ratios is all that helpful in the long run.
Hmmmmmm.... maybe I have not run long enough yet, but I do most of my thinking in ratios.
> A bit like 3D positional triangulation.
Right. Or even 10-D navigation. All the ratios have to be reasonable.
Tdale omitted the load... could be a 1Meg pot or a 50K trem-roach network. As a wild-guess ratio, I'd want the DC plate resistor 2 to 5 times lower than the load, and the tube's plate resistance 2 to 5 times lower than that resistor. But in maximum-gain/tube systems we may cheat both ratios lower.
In triodes, the RP/RK ratio does tend to be happy near RK= RP/(Mu*0.6). For 12AX7 with 100K in the plate, 100K/(100*0.6)= 100/60= 1.67K, use 1.5K. But a smaller RK will give a bit more gain with a lower input overload level.
It is fairly simple to work out on paper what happens on "small" signals. Signals so small that nothing distorts. But electric guitar is boring that way. The "sound" of an amp is a LOT about how it overloads, and how it transitions from mild warpage to extra color to gross fuzz. That's a VERY complicated question. Just calculating the waveforms tends to choke computers. Knowing the shape of the waveform does not tell you how the EAR will perceive it. Computations (napkin or SPICE) tend to use steady waves because changing waves are infinitely more work; guitar waves are always changing.
For initial clean-sheet rough design, use the tube manual R-C Amplifier tables. They represent hundreds of man-hours of junior engineer time, working for companies who wanted to sell more tubes, trying to make it easy for production designers to make stuff that would make the final customers happy so the market would expand.
Or steal existing designs, but if you take early Fender you will see that most of his 1954-1960's (and later) preamp and mix stages are EXACTLY from the GE tube manual. (RCA published slightly different tables, so we know Leo read GE.)
Read the fine print. For small-signal work, as plate resistor goes up the treble response limit goes down. You should do an exact analysis after you pencil a value, but what value to start with? For many typical systems, 100K will give response flat to 40KHz, 470K will be rolling off at 10KHz. Sure, geetars don't do much 10KHz, but we like the 3KHz-6KHz band and don't want any loss here. Also cascaded stages have more roll-off: four "-3dB at 40KHz" stages will obviously be down 12dB at 40KHz, 4dB at 20KHz, and a good (bad!) dB at 5KHz where we may (or may not) want all we can get.
And that's just for "small" signals. Overdrive adds harmonics and makes the sound bright. We may then actually want a rolloff to tame the harshness. Fender used 100K and invented hard rock. Gibson tended to 270K and 470K and was favored by some mellow (or less raspy) country and jazz players.
Of course even that is too simple. You can build an overbright LM558/LM377 chip-amp, use a broad pick and tame pickups, run it through a cone made from soggy toilet-paper, and get a fine mellow sound. A lot of transistor guitar amps have real pulpy cones.... wonder why.
Hmmmmmm.... maybe I have not run long enough yet, but I do most of my thinking in ratios.
> A bit like 3D positional triangulation.
Right. Or even 10-D navigation. All the ratios have to be reasonable.
Tdale omitted the load... could be a 1Meg pot or a 50K trem-roach network. As a wild-guess ratio, I'd want the DC plate resistor 2 to 5 times lower than the load, and the tube's plate resistance 2 to 5 times lower than that resistor. But in maximum-gain/tube systems we may cheat both ratios lower.
In triodes, the RP/RK ratio does tend to be happy near RK= RP/(Mu*0.6). For 12AX7 with 100K in the plate, 100K/(100*0.6)= 100/60= 1.67K, use 1.5K. But a smaller RK will give a bit more gain with a lower input overload level.
It is fairly simple to work out on paper what happens on "small" signals. Signals so small that nothing distorts. But electric guitar is boring that way. The "sound" of an amp is a LOT about how it overloads, and how it transitions from mild warpage to extra color to gross fuzz. That's a VERY complicated question. Just calculating the waveforms tends to choke computers. Knowing the shape of the waveform does not tell you how the EAR will perceive it. Computations (napkin or SPICE) tend to use steady waves because changing waves are infinitely more work; guitar waves are always changing.
For initial clean-sheet rough design, use the tube manual R-C Amplifier tables. They represent hundreds of man-hours of junior engineer time, working for companies who wanted to sell more tubes, trying to make it easy for production designers to make stuff that would make the final customers happy so the market would expand.
Or steal existing designs, but if you take early Fender you will see that most of his 1954-1960's (and later) preamp and mix stages are EXACTLY from the GE tube manual. (RCA published slightly different tables, so we know Leo read GE.)
Read the fine print. For small-signal work, as plate resistor goes up the treble response limit goes down. You should do an exact analysis after you pencil a value, but what value to start with? For many typical systems, 100K will give response flat to 40KHz, 470K will be rolling off at 10KHz. Sure, geetars don't do much 10KHz, but we like the 3KHz-6KHz band and don't want any loss here. Also cascaded stages have more roll-off: four "-3dB at 40KHz" stages will obviously be down 12dB at 40KHz, 4dB at 20KHz, and a good (bad!) dB at 5KHz where we may (or may not) want all we can get.
And that's just for "small" signals. Overdrive adds harmonics and makes the sound bright. We may then actually want a rolloff to tame the harshness. Fender used 100K and invented hard rock. Gibson tended to 270K and 470K and was favored by some mellow (or less raspy) country and jazz players.
Of course even that is too simple. You can build an overbright LM558/LM377 chip-amp, use a broad pick and tame pickups, run it through a cone made from soggy toilet-paper, and get a fine mellow sound. A lot of transistor guitar amps have real pulpy cones.... wonder why.
Re: plate resistor values..
Ok, så there is basically no difference in sound between a 100K and a 200K plate resistor. If there were, it would be easy to describe it.
It’s not there is no difference, it’s that there are a great many variables that come into play and as such the audible possibilities tend to be greater than most generally give initial thought to, as such you cannot easily widdle or funnel down those possibilities into specific adjectives without making a great number of assumptions. IE: change one thing and have introduced several variables, all of which have a direct impact on not only the tone but also the feel of an amp. In essence, making numerous assumptions (with regards to the variables) would be greatly over simplifying a very complex equation.
I’m not trying to be vague and I’m sure there are several here who will truly do a wonderful job trying to explain everything in as much detail as possible (God bless them) but in the end nothing takes the place of hands on experience. Sometimes it’s best to just try various plate resistor values so that your ears can tell you what effects those various changes have made and equally as important, how the amp feels when playing on it.
I liked the 220k in the first stage of two Fenders I tried it in because it gave a more robust sound to my ears. I didn't like it in my modded Marshall (100k had less buzz than 220k), and I didn't like it in the Soldano Atomic 16 I once had. Reducing the 220k to 110k (I simply added a 220k in parallel) took out some of the spikiness in the sound.
Something that may be worth documenting on the first stage is what the actual plate voltage was for each amp. This may give you more of an idea of why you like a specific value in one amp but not the other, etc and so on. It’s just food for thought as the ‘ol saying goes— ha, ha).
I still believe that the only way to find out is to try, listen and trust your ears.
I whole heartedly agree, you can crunch match until the day is long however the proof is in the pudding or the listening in this case (ha, ha).
Great board you gentleman have here! (Kudos)
Trace
It’s not there is no difference, it’s that there are a great many variables that come into play and as such the audible possibilities tend to be greater than most generally give initial thought to, as such you cannot easily widdle or funnel down those possibilities into specific adjectives without making a great number of assumptions. IE: change one thing and have introduced several variables, all of which have a direct impact on not only the tone but also the feel of an amp. In essence, making numerous assumptions (with regards to the variables) would be greatly over simplifying a very complex equation.
I’m not trying to be vague and I’m sure there are several here who will truly do a wonderful job trying to explain everything in as much detail as possible (God bless them) but in the end nothing takes the place of hands on experience. Sometimes it’s best to just try various plate resistor values so that your ears can tell you what effects those various changes have made and equally as important, how the amp feels when playing on it.
I liked the 220k in the first stage of two Fenders I tried it in because it gave a more robust sound to my ears. I didn't like it in my modded Marshall (100k had less buzz than 220k), and I didn't like it in the Soldano Atomic 16 I once had. Reducing the 220k to 110k (I simply added a 220k in parallel) took out some of the spikiness in the sound.
Something that may be worth documenting on the first stage is what the actual plate voltage was for each amp. This may give you more of an idea of why you like a specific value in one amp but not the other, etc and so on. It’s just food for thought as the ‘ol saying goes— ha, ha).
I still believe that the only way to find out is to try, listen and trust your ears.
I whole heartedly agree, you can crunch match until the day is long however the proof is in the pudding or the listening in this case (ha, ha).
Great board you gentleman have here! (Kudos)
Trace