plate R value and PSU voltage question

[daz]

This is really just out of curiosity, but i was wondering about supply voltage to the preamp in a hi gain pre as it relates to plate resistor value. Specifically i wanted to know if a 100K plate r wants to see more voltage than a 220k does in order to get the best tone. The reason i ask is this. A while back Chuck gave me some good info that i started playing around with and i went back to a 100K on V1A's plate. I had a 100 there at first, but switched to a 220k i dunno, maybe 5 6 months or more ago. anyways, the 100K just didn't cut it so i want back to the 220k. But then it hit me that when i went to a 220k the first time my preamp voltage was way lower. I had selected some huge dropping resistors to get the preamp at V1 A and B something like 100v and 150v respectively ala bad cat. anyways, long story short i eventually thought about it abd wondered it the supply voltage would make a 100K work better and i changed the PSU resistors till i had voltages in the mid 200 range or more and tried the 100K again. It seemed to sound good and i then ended up biasing V1 with more traditional cathode r values. I've been playing it like this for several weeks now and it sounds better than ever. I believe i will leave it, but like i said i am really curious if in general a 100K plate resistor works best with higher supply voltages than a 220k does. When i upped the supply voltage with the 220 it always seemed to get harsh. But with the 100k i and higher voltage it got cleaner and all around better in all ways. with lower voltages the 100 never sounded quite right.

[Chuck H]

In the context of your post the 100k plate R souded more "correct" with the higher voltages because the higher voltage produces more gain. So it follows that the 220k plate R sounded more correct with the lower voltage because the higher plate R also provides more gain.

This could be why you'll usually see the 220k plate R on lower voltage preamps. At least vintage amps. Many modern high gainers use a high plate R with highish voltages to build intentionally excessive gain.

Lower plate volts will create a browner tone. Less defined with the edges smoothed out. Raising the gain at this point with a higher plate R does not have as much dynamic response as higher plate volts with a lower plate R. It can even sound mushy and buzzy in the wrong amp. But for some amps it seems to balance with other characteristics to create a good tone. The Vox amps are a good example. As well as some of the lower powered vintage Fender amps.

Higher plate volts sounds brighter and tighter. Tighter is the key here. Because if you get a fizzy mess with the lower plate volts and higher plate R it's probably partly due to the bottom end distortion.

Other than gain I can't say much about how other tube characteristics may be effected by the plate R. Like internal capacitances and mu factor. I know the plates output impedance is raised quite a bit with the higher plate R. This would make it more susceptible to loading and damping by following circuits. An effect that increases with frequency. So thats another difference. And there must be others, and I hope someone chimes in on this, because even with equal bias properties there is a notable difference in tone, not just gain.

Chuck

[Merlinb]

This is really just out of curiosity, but i was wondering about supply voltage to the preamp in a hi gain pre as it relates to plate resistor value. Specifically i wanted to know if a 100K plate r wants to see more voltage than a 220k does in order to get the best tone.

Broadly speaking, the supply voltage has very little to do with the tonal result, within reason.
Supply voltage v. load does affect output signal swing of course, which itself affects how much the following stage can be overdriven. However, even with a low-ish supply voltage, using a large anode resistor can give you a lot of swing (and most amps use interstage attenuators anyway, since they have too much swing available!).
What matters is load and biasing. If you simply increase the load, you will end up hotter biased with a lower anode voltage (which does increase grid current). The anode voltage is irrelevant, its the change in bias that caused 99% of the tonal change. But if you the bias back towards the original operating point you can vary the tone back towards the original sound (within reason). As long as your supply voltage is sensible (200V or more), you can always duplicate the tone of the another amp even if that amp has a different supply voltage, by altering the load and biasing.

[MWJB]

"The anode voltage is irrelevant, its the change in bias that caused 99% of the tonal change." I don't think irrelevant is the quite the right word, the tube bias in most amps is the secondary result of power supply voltage, e.g. Take a tolex Fender with 200v on the plate and a tolex Fender with 300v on the plate (both 100K plate R, 1500ohm cathode), there's quite a difference in tone...rebiasing with a different cathode & plate resistor in the case of the lower voltage amp isn't going to make both sound the same. Reducing the value of the power supply dropping resistor and keeping original values for plate & cathode, will get you closer.

Daz - look at plate load resistor as being the key factor in "fidelity", 100K is most common, 220k (with the same voltage supply & cathode resistor) will sound a shade darker/grindier due to both the lowered voltage and the rebiasing of the tube (they go hand in hand). That's often why plate resistors higher than 100K are used in the early stages of high gain designs (sometimes with a smaller cathode resistor than 1500/triode too), to get the "flavour" early on, later stages will often be set up cleaner so as not to compound the issue & keep everything from turning to mush.

[Merlinb]

Take a tolex Fender with 200v on the plate and a tolex Fender with 300v on the plate (both 100K plate R, 1500ohm cathode), there's quite a difference in tone...rebiasing with a different cathode & plate resistor in the case of the lower voltage amp isn't going to make both sound the same.

I would disagree. Witha little fiddling you can make the two sound substantially similar. Maybe not identical, but close enough that most people wouldn't notice. It's not the plate voltage that makes the two examples sound different, it's the relative bias points (and to some extent the output swing).

[MWJB]

In turn, I have to disagree with you. Most manufacturers stick to a plate resistor/cathode resistor combination accross the majority of the models, it's this (in part) that gives them that particular brand's characteristic sound. Yes, as the plate voltage rises, or falls, the tube is rebiased (so it is very difficult to say that one half of the equation has all the effect & the other none - because they are interlinked) but the overall character remains recognisable. Otherwise every amp from the same manufacturer, with a different power supply voltage, would have a different plate/cathode combination.

I have done an awful lot of testing with this perameter and replicating a 220K/820ohm triode with a 100K/1500 triode, at any voltage is a tall order.

[Merlinb]

Otherwise every amp from the same manufacturer, with a different power supply voltage, would have a different plate/cathode combination.

I must admit I doubt most non-boutique manafacturers would ever go to such effort. They just go with whatever generic design they've used before, add a few new bells and whistles using whatever PSU voltage happens to come off the power stage. The PSU voltage is never going to vary that much, since 99% of power stages operate from about 350V to 500V.

[daz]

The anode voltage is irrelevant, its the change in bias that caused 99% of the tonal change

But see, i DID bias it like that before with the lower supply and it sounded/felt quite different than it does now. Thats the point i was getting at. With the much higher voltage i could bias them like this and instead of sounding worse it now sounds good enough that i haven't touched anything for a few weeks and don't plan to. last time i tried these values with low voltage the tone was mediocre.

It's killing me because i developed a case of tennis elbow (a form of tendonitis) and it hurts to play, and of course this happens right when i get it sounding so good i don't want to put my guitar down !

[Merlinb]

But see, i DID bias it like that before with the lower supply and it sounded/felt quite different than it does now.

But do you mean you biased it to the same anode voltage as before, or the same grid-to-cathode voltage or something else?

[Chuck H]

I think what Daz means is that with either plate load he adjusted the bias so the tube would be at least at a similar operating point. Which was your point before. And all things being equal, though they never are, this still sounds different.

If you look at load lines and charts all you see is nice smooth lines. But in real world operation I don't thing the tube will ever operate as ideal as the charts drawn for it. There ARE differences in tone. Though it can be minimized with proper biasing as you said. But on the macro view there are large clumsy differences that cannot be biased out, like plate impedance, power supply resistance and voltage. If you really think that two tubes of different plate voltages can be made to sound the same with proper biasing, then I say nay. Tubes DO sound different at different voltages. Not just as a function of the voltage shifing the bias (assuming the bias is not adjusted). Higher plate volts will make the tube sound/feel tighter and more dynamic.

And then theres the plate impedance. It's ideal in audio circuitry to run low impedance outputs into high impedance inputs because it reduces damping, which is non-linear and increases with frequency. This is a very audible effect.

Also consider that the plate load IS another resistor in the power supply. So a higher plate load will also create more sag, Adding to the effect of signal damping.

Chuck

[MWJB]

With a different cathode or plate resistor, even if you arrived at the same plate voltage the bias (current draw) would be different, you would have to change dropping resistors in the power supply, but by the time you got the current through the tube to be the same as before, the plate voltage would now be different...anyone feeling dizzy yet? :-)

If we take a typical, middle of the road, guitar amp, preamp triode, running say 225vdc on the plate with 100K plate load and a 1.5K cathode, this would equate to a supply voltage to the plate resistor of roughly 330vdc and a plate current of 1mA.

Now if we just change the plate to 220K, the plate voltage will drop, the current through the tube will drop. If we want to get the tube to bias the same (draw the same current) we need to adjust the dropping supply so that the plate sees the same voltage ...this might mean applying 440-ish volts to the plate resistor (gets us back to 1mA)...fly in the ointment might be that the B+ doesn't generate 440vdc in the first place. It will still sound different to the original set up.

On the other hand, we could stick to 330vdc supply & change the plate to 220K, we could then double the cathode, this might get us back nearer the original 225vdc on the plate & 1mA dissipation but now things will sound different (the bypass cap may now also need changing to restore original frequency response).

One always affects the other...the more typical approach is to select plate resistor for the degree of fidelity you want, cathode resistor for the balance of clean/distortion & then tweak B+ supply voltage (via the dropping resistor) by ear. You might, with a typical 100K/1.5K triode (running in early/mid 200's) need to adjust by 20v or more before hearing an audible difference in tone, but the overall flavour of the amp will remain...as Chuck previously posted...higher voltages will be subtly (at these small increments) cleaner with more high end, lower voltages grindier/browner. For significant changes in clean vs distortion it might be the case that you stick with the plate resistor & change the cathode value.

[daz]

I think one thing that should be mentioned is that a lot of what we, or at least I, tend to call tone is really more of a feel and dynamics things. If you wanna talk strictly tone then theres still a difference, but it's not as radical a difference as i FEEL in the amp. But lets face it, the dynamics from one circuit to another can be huge in the way the TONE ultimately comes out of the speaker because it allows your hands to manipulate the tone in ways you cannot with a less dynamic circuit. But in short, thats where the biggest change was.

I think what Daz means is that with either plate load he adjusted the bias so the tube would be at least at a similar operating point. Which was your point before. And all things being equal, though they never are, this still sounds different.

basically, yes, but much more than that i tried dozens of combinations with both anode resistors with the low supply voltage and with the 220k i could never use lower cathode resistors. And a 100k plate never sounded as good no matter what the cathode resistor values were. the cathodes had to be above 2.2k and the plate 220k or it just wasn't near as good. With a lot more voltage and a 100k anode i now have a 820R on V1A's cathode and a 1.5k on V1B. and the tone/feel/dynamics are all fabulous. but in any case it just seems like with higher voltages regardless of the cathode resistors and how the tube is biased, it works best with 100k vs 200k plate and visa versa.

[Steve Conner]

This is actually a fairly complicated question. All of the things everyone has said so far are true in one way or another, except Merlinb's assertion that increasing the plate load resistor makes the bias hotter. It's the opposite: the grid-cathode voltage might decrease, but the plate/cathode current decreases.

I think there are two main effects. The first comes from where the tube's operating point sits in its characteristic. If you sit it at a low plate current, you're near the curved portion of the characteristic and you get soft, "brown", dirty distortion. And if you overdo it, a funky, sputtery tone like a transistor radio with a low battery. Higher currents mean more linear operation, and the clipping is more abrupt and harsh when it eventually comes. Hence, a brighter, more aggressive tone.

The other effect is how the bias point determines the available voltage swing at the plate. If you try and dial up a low plate current with a small plate resistor and/or high supply voltage, the swing at the plate will be highly asymmetrical. And ditto for a high plate current with low B+ and/or a high plate resistor. It always seemed to me that the best tones were with a roughly symmetrical swing, and this is why the 220k needs a larger cathode resistor to sound good.

These two effects are orthogonal, so you can play with them independently. Unfortunately, changing any one resistor, or the B+, affects both of them. Effect 1 is controlled by changing the plate and cathode resistors in the same proportion (ie doubling them both, which doesn't change their ratio) while effect 2 is controlled by changing their ratio (ie doubling Rp while halving Rk)

For instance, 12AX7 datasheets give two standard sets of values, 100k/1.5k and 220k/2.2k, or whatever. They both give roughly symmetrical clipping at the usual voltages used in tube amps, but the 220k gives more harmonic distortion because the operating point is down closer to the knee. It also has somewhat more gain, though not 2.2 times more.

As to how all of this relates to tone, I must admit that I'm not too sure. I once played around with an old Selmer Corvette, and these bias the first preamp tube very cold with a large cathode resistor. I think the point is to generate lots of harmonic distortion to hide the fact that the poor Corvette, with its 8" speaker and minuscule OT, had no bass response whatsoever. It also made it a little dirt monster that was great fun to overdrive with a clean boost. But it was impossible to get anything like a clean tone out of it.

[daz]

It also made it a little dirt monster that was great fun to overdrive with a clean boost. But it was impossible to get anything like a clean tone out of it.

Thats something i forgot to mention, which is now my cleans are much better.

[Merlinb]

All of the things everyone has said so far are true in one way or another, except Merlinb's assertion that increasing the plate load resistor makes the bias hotter. It's the opposite: the grid-cathode voltage might decrease, but the plate/cathode current decreases.

Yes sorry, I use the term 'hotter' to indicate biasing closer to grid-current, rather than literal temperature increase.