Presence control

[lowell]

does a presence control w/ feedback ADD presence or only taketh away? I have an amp with no feedback that could use some brightness. I don't wanna try presence if it doesn't ADD highs.

[MWJB]

The cap to ground, in parallel with the NFB load resistor, adds presence to a circuit with a negative feedback loop, the presence control turns the cap off.

However, a lack of a negative feedback loop should sound tighter & brighter than the same amp with a negative feedback loop, presence control or not.

[Steve Conner]

A presence control shunts away some of the negative feedback at high frequencies. This gives the amp's output a high frequency boost. The frequency response is almost flat with the presence knob at 0, and considerably treble boosted with it at 10. I measured about 6-8dB boost on the last amp I worked on.

As you can guess from the above, if you have no NFB, a presence control wouldn't do anything. In fact there's nothing to connect it to.

But as MWJB says, the tone of a NFB-less power amp is usually pretty bright and rich (wouldn't say "tight"), because of speaker interaction. Without NFB, the speaker gets a current drive that overcomes voice coil inductance, giving considerably more treble. It also lets the speaker resonate more at its bass resonance frequency, so you get a bass boost too. You really hear the character of the speaker, more so than if the amp had NFB. The presence knob also releases some of this character, because it boosts treble by taking NFB away.

Probably the best thing to do in this case is reduce the bass somehow. You can't add highs, so try taking lows away, and the amp will seem brighter. You could rig up a NFB loop that only functions at bass frequencies, just like the normal circuit but with a really big presence cap and the presence control turned to 10.

[Enzo]

Well, you can add a brightness cap across the volume control, that is what the bright switch on a fender amp does. Unless the control is maxed, you get relatively more highs. And in places where there is voltage division of the signal, a small value cap parallel the upper leg resistor will also increase relative brightness.

[MWJB]

Also worth having a look at the preamp voltages, raising these will brighten the amp. What do you have on the preamp plates now? (In fact knowing what amp we're discussing, or seeing a schm might help).

[Merlinb]

Originally Posted by MWJB

Also worth having a look at the preamp voltages, raising these will brighten the amp.

Why's that then?

[MWJB]

Higher preamp voltages give better high end (assuming all other factors remain the same), lower preamp voltages will give a darker, browner tone. Of course, it's useful to know what voltages someone has to start with, as well as cathode & plate resistor values, but if you're generally happy with the tone of the amp & want a very similar sound, just a bit brighter, then bumping up the preamp plate volts might do the trick. You can still apply Enzo's bright cap suggestions.

You might have to see 20 or more volts difference to hear an audible change. Preamp tubes don't draw much current so you might need to make relatively aggressive changes to the preamp dropping resistor in the power supply. Be aware of any other parts of the amp that are fed from common nodes, that might not benefit from having their voltages raised.

[Merlinb]

Originally Posted by MWJB

Higher preamp voltages give better high end

But why? The theory seems to disagree with that observation. I'm not saying you're wrong, I just wondered if you could explain why it happens.

[MWJB]

"The theory seems to disagree with that observation." Which theory is that?

The more voltage you run through a tube, the less it distorts & the tighter & brighter it sounds. You could up the value of the cathode resistor but this will have a more pronounced effect on the character of the tone, tweaking the voltage (for a given plate/cathode combination) will give a more consistent change in the tone...IF that is what you are looking for.

[Merlinb]

Originally Posted by MWJB

"The theory seems to disagree with that observation." Which theory is that?

Well I was trying to think what factors might cause bass cut/treble boost when you increase the supply voltage.

Increasing the voltages will increase the headroom, so the amp will distort less for a given input, meaning less generation of harmonics.

Each stage becomes more linear before clipping, so again, less generation of harmonics.

Surely fewer harmonics can only mean a less bright sound?

The internal anode resistance will decrease slightly, so the output coupling cap's pole -and the cathode bypass cap's pole- will rise slightly, but the change is so microscopic that I suspect it is too small to be significant.

The gain and output signal swing will increase, however, so maybe the PA gets overdriven more which causes more harmonics? But then we wouldn't expect to hear any difference when playing clean.

Higher voltages increases susceptibility to microphonics (again, a microscopic amount), which might increase intermodulation distortion. I'm clutching at straws here...

The more voltage you run through a tube, the less it distorts & the tighter & brighter it sounds.

I don't think we can take that as a fact,

[MWJB]

"Surely fewer harmonics can only mean a less bright sound?"

Why? What if the harmonics you are losing are part of the initial problem?

"The internal anode resistance will decrease slightly, so the output coupling cap's pole -and the cathode bypass cap's pole- will rise slightly, but the change is so microscopic that I suspect it is too small to be significant."

Are you suggesting that a triode with a given cathode & plate resistor combination will sound the same at 150v on the plate as at 250v on the plate? I hope not. This is one of those scenarios where it is quicker to change a resistor & see than it is to speculate. You don't give any basic perameters for what you might consider "high/low", or what kind of difference might be "significant". A 12AX7 triode running at 150vdc with a 100K/1.5K combination will run at approx 0.7mA, with 250vdc on the plate you'll see more like 1.2mA, or nearly twice the current, strikes me as significant.

"Higher voltages increases susceptibility to microphonics (again, a microscopic amount), which might increase intermodulation distortion. I'm clutching at straws here..."

I'm only suggesting running at voltages that are already in use in other amps, say mid/late 200's tops.

"Quote:
The more voltage you run through a tube, the less it distorts & the tighter & brighter it sounds. - I don't think we can take that as a fact" No really, please do.

[Merlinb]

Originally Posted by MWJB

"Surely fewer harmonics can only mean a less bright sound?" Why? What if the harmonics you are losing are part of the initial problem?

Because nearly all harmonics are integer multiples of the fundamental. i.e., they are always higher in frequency than your audio signal. So by adding more harmonics you're adding more high frequencies.

Are you suggesting that a triode with a given cathode & plate resistor combination will sound the same at 150v on the plate as at 250v on the plate?

No. Rather I'm saying that at lower voltages I would expect it to sound brighter, due to an increase in THD.

You don't give any basic perameters for what you might consider "high/low", or what kind of difference might be "significant".

I suppose a relative increase of, say, 0.5dB over the mid/treble range, relative to the bass range, would look reasonable. But in this case even going from 150V to 400V will only move the poles up by a factor of less than 1.1

A 12AX7 triode running at 150vdc with a 100K/1.5K combination will run at approx 0.7mA, with 250vdc on the plate you'll see more like 1.2mA, or nearly twice the current, strikes me as significant.

Yeah, but you can't hear current.

The more voltage you run through a tube, the less it distorts & the tighter & brighter it sounds. - I don't think we can take that as a fact" No really, please do.

Well, um, facts, kinda, need evidence.

[Steve Conner]

I believe the reason is that when the voltage is increased, the harmonics generated are higher in order, though they may be less in amount. And human hearing weights the audibility of harmonics according to the square of their order.

My own (patchy) experiments seem to vaguely support this observation.

[Merlinb]

Originally Posted by Steve Conner

I believe the reason is that when the voltage is increased, the harmonics generated are higher in order, though they may be less in amount.

Hmm, that's plausible, I hadn't though of that- thanks Steve, I'll look into it.

[Steve Conner]

I should probably look into it too. A while back I promised to measure the distortion characteristics of the Techtube ECC83 alongside an original Mullard, and I could get some results on distortion spectrum vs. voltage as part of the same experiment. Any ideas on how to do it appreciated :-)

[MWJB]

"Well, um, facts, kinda, need evidence."

Then rig up a SPST in say a Super Reverb, one way it feeds the stock 10K preamp dropper, the other feeds a 100K preamp dropper instead, tell me what you hear...you'll hear it at all frequencies I promise you.

I don't know what kind of evidence you expect from me, other than testimony. This is a web forum, all people see here are words/opinions. What "evidence have you put forward that is in any way more tangible than what I have. We're looking at the same scenario from different perspectives, you from the theoretical, me from the practical, as result of personal experience.

Again, it takes longer to read this thread than it does to change a power supply resistor, then folks can decide for themselves.

[Merlinb]

Originally Posted by Steve Conner

I should probably look into it too. A while back I promised to measure the distortion characteristics of the Techtube ECC83 alongside an original Mullard, and I could get some results on distortion spectrum vs. voltage as part of the same experiment. Any ideas on how to do it appreciated :-)

Got a spectrum analyser? Surely that's all you need, and a variable power supply of course.

I think Right Mark's free audio analyzer could work: Latest News. Audio Rightmark

[Merlinb]

Originally Posted by MWJB

"Well, um, facts, kinda, need evidence." Then rig up a SPST in say a Super Reverb, one way it feeds the stock 10K preamp dropper, the other feeds a 100K preamp dropper instead, tell me what you hear...you'll hear it at all frequencies I promise you.

That's observation, granted, but I think it takes a bit more than someone's own fallible hearing for something to become a fact. It needs a theory to explain why it happens, that what I was doing, throwing around some possibilities. I mean, can we be sure that raising the voltages always has this effect, every single time? And how much? What other factors come into play?
I mean, raising the PSU voltages affects the whole amp, and that's a big change. Can we be sure that it is simply the higher voltages which cause this, or is it some knock-on effect which could be attacked more directly next time we want more brightness, without having to mess with the PSU.

As you can see, I don't like sweeping statements about audio, mainly because they're nearly always wrong. They belong to the subjectivists, not us audio engineers.

[Steve Conner]

Originally Posted by Merlinb

Got a spectrum analyser? Surely that's all you need, and a variable power supply of course.

Yes, I have an audio spectrum analyser and a variable power supply. The question is, what exactly should I do with them? Of course I have some ideas of my own, but a second opinion would be nice.

MWJB: A while back I had to turn down the HT regulator on my old Ninja Toaster amp. I noticed it didn't have enough headroom and was hardly regulating at all. To restore the headroom I had to drop the main HT from 475V to 400, which would be the equivalent of EVH dropping hs legendary variac from 115V to 97. When I did this, it seemed to me that the dirty channel got darker and mushier. I ended up fitting a presence boost to restore the lost high end.

Of course I could have imagined all of this by Douglas Self's experimenter expectancy effect

[Merlinb]

Originally Posted by Steve Conner

Yes, I have an audio spectrum analyser and a variable power supply. The question is, what exactly should I do with them? Of course I have some ideas of my own, but a second opinion would be nice.

Hmm, well I guess you would need to try out a number of topologies, to see what results are consistent. These might include:

1. Standard stage w. resistor load and Rk. Nothing changed but the supply voltage.

2. Repeat but with large bypass caps added.

(Maybe repeat the two above, but with different Rk values?)

3. Resistor loaded but with constant voltage biasing, such as an LED.

4. CCS loaded, constant voltage bias.

5. CCS loaded, resistor biased.

6. Same as above but using some other triode types, other than ECC83 / 12AX7, to see if it is common to all triodes or just that one.

I guess you'd want to observe the harmonics at low level, high level, and also when clipping.

[MWJB]

Hearing is how amps are evaluated. People play & buy amps that they like the sound of. Of course, you have no way of knowing how fallible my hearing is, or isn't (so equally you have no reason not to trust my judgement - you could consider that I am actually credible & offer my suggestions through goodwill), but again why would anyone reading this, looking for tips to brighten their amp, believe your theorising over my practical observation, or the other way around? A fact is a fact, whether you know the theory or not...if it's raining, then it's raining...you don't have to be a meteorologist, or how clouds form to know you're getting wet.

"I mean, raising the PSU voltages affects the whole amp," changing anything has a knock on effect...but changing the preamp voltages mostly just changes the preamp voltages (as close as you can get to "all things being equal"), add another node if you need to. On a 2 channel amp with a shared input tube you can run both halves at whatever voltage you want and hear the differences.

You make "messing with the PSU" sound like a big deal? Changing a resistor is not a big job, you don't HAVE to write a thesis on it first, you can just heat up the iron and do it.

"They belong to the subjectivists, not us audio engineers". Tone is subjective, you need to balance the theory with what the player wants to hear, most players aren't audio engineers either but they can still decide on what kind of tone they like. That's the job, to build/service/mod an amp that sounds better to the user than when it did before/what they used before.

You seem also to have failed, up to this point, to make any suggestions yourself as to how the OP might brighten their amp. This is where people come for suggestions/help/tips ...they're free to use them/try them/ignore them, whatever, but I'm not keen on this fad for hijacking threads purely for grandstanding. If you want to carry out my suggestions and make some audio clips to "prove me wrong" then do so, but you might, as a matter of courtesey, withhold "judgement" until you have something tangible to display. FWIW I have nothing to prove to you and I am not asking your opinion on the subject at hand...if I can help you get your head around the scenario then great, perhaps I may in turn learn something, but I'm not prepared to listen to you tell me that such & such does not happen, just because you don't see how it could...without trying it.

[MWJB]

"To restore the headroom I had to drop the main HT from 475V to 400" interesting but, just to be clear, I am not suggesting changing any dc voltage other than that supplied specifically to the preamp tube. This is being achieved in my scenario by reducing the value of the preamp dropper resistor (not the plate resistors, not the cathode resistor).

[Merlinb]

Originally Posted by MWJB

This is being achieved in my scenario by reducing the value of the preamp dropper resistor .

Aha! We may have found another cause...
By lowering the dropping resistor you raise the pole of the PSU smoothing cap (in many amps these are pretty puny to begin with). So what may be happening is that you're reducing the bass response of the PSU, rather than increasing the treble response of the preamp tube! If that's the case then leaving the voltages the same but reducing the size of the smoothing cap would have the same effect...

Oh, and if I tell you it's raining its and it is indeed raining then it is a fact.

But if I tell you that clapping my hands causes it to rain, and it does happen to start raining when I clap my hands, is what I said a fact?

I think you can see how the the two are different. This is an electronics forum, rafter all, not just any old here-say forum.

but I'm not prepared to listen to you tell me that such & such does not happen, just because you don't see how it could...without trying it.

I think youneed to re-read the thread. I never once said you were wrong.

[MWJB]

"If that's the case then leaving the voltages the same but reducing the size of the smoothing cap would have the same effect..." It has an effect, and is a valid suggestion, but it's not "the same" effect and does not fully explain the difference in tone. You're also relatively limited in useful cap values, given their sonic impact. I am not closed to the idea that there is more than one factor at work, however the obvious one is the change in plate voltage, there will undoubtedly be other contributory factors.

Why is what I say "hear say" & what you say something else?

[Old Tele man]

Originally Posted by MWJB

Why is what I say "hear say" & what you say something else?

...as it was once bluntly put to me: "...because my SIX is not the same as your HALF-DOZEN until you can show me how you got SIX doubled to TWELVE, equated to DOZEN, then HALVED back down to equal my SIX."

...just my worthless 2¢.

[loudthud]

When I started this experiment, I thought the results would clearly show a bandwidth reduction in step with B+ reduction. I didn't notice until the end that the heater voltage was 7.01VAC coming from my Heathkit bench supply. I used a GE 12AX7A/7025 that I got a big batch back in the 80's. The plate resistor was 100K and cathode resistor 1.5K bypassed with 100uF. The output was taken on an oscilloscope with a 10X 10Meg probe AC coupled. The input to the tube is from a function generator through a .022 with a 1Meg to ground on the grid to eliminate the DC coming out the function generator. The gain numbers are taken with an input of 100mV pk-pk. F(3db) is taken by setting the output to 6Vp-p at 1kHz and running the frequency up until the scope indicates 4.2Vp-p and reading the function generator dial. Note that at 150V I had to take the reading on a different frequency range so it could be off by a couple of kHz.

B+....V(k).......V(p).....Gain.....F(3db)

350...1.538.....245.1......58......102kHz
300...1.296.....211.5......56.8...102kHz
250...1.064.....177.8......55.2...101kHz
200...0.824.....143.3......52.8...100kHz
150...0.582.....109.3......46.6...98kHz

Conclusion: B+ reduction BY ITSELF does not significantly effect bandwidth.

[Merlinb]

Originally Posted by MWJB

Why is what I say "hear say" & what you say something else?

Well, ok, hear-say might be too strong, but what I said was basic electronic theory- and I can provide references to the effect, on request. You made a statement based only on observation (and I had to press you for that), but you said it as though it was a universal truth. It might be universally true of course, I just wondered if you could explain to me why it happens, but I guess it needs looking into.

[Old Tele man]

Originally Posted by loudthud

Conclusion: B+ reduction BY ITSELF does not significantly effect bandwidth.

...but, what's the content/composition of that bandwidth? Same number of harmonics at differing amplitudes?

[MWJB]

Loudthud, what is your conclusion therefore on how the tube sounded at the differing B+ you tried. Are you saying it sounds the same irrespective of voltage. What were the sonic effects of the differing voltages, curent draws & gain? There are some flaws in the test as compared to real life scenario, particularly with respect to heater voltage - when you tweak preamp voltage in an amp, heater voltage remains constant, it does not rise or fall, this will affect tone.

[guitarmike2107]

Interesting thread, I think Merlin is just trying to get a proper understanding of what you are reporting. It certianly is usefull to know that raising the voltages increases the brightness but for the Engineers amoungst us its also good to have an understanding as to why.

After reading this far I want to know why too?

Mike

[pdf64]

My understanding was that the role of the engineer was to implement a solution to reach the desired result, generally by using previoulsy proven solutions to similar situations.
And it's the scientist's role to find out why a particular implementation works.
When that is known, engineers can then use this knowledge to inform as to why one solution might be preferable to another etc.
Obviously there's a lot of crossover, but still a basic distinction.
Therefore Enzo and MWJB have engineering solutions to the OPs problem, and Merlin, Steve et al are taking a scientist's perspective by wanting to know how/why MWJB's proposal works.
I feel that Enzo's solution has the best chance to solving the OPs problem, but if we could get something quantifiable on the pre-amp voltage / brightness thing, it would be a big step forward. Peter.

[Steve Conner]

I always believed that it worked as follows:

Tubes intended for audio have a portion of their characteristic over which they can be assumed linear. Outside of that, they have "cutoff" where the plate current stops. Cutoff is not abrupt, the characteristic starts to curve as it approaches zero, and the curved part is called the knee.

When the tube runs on higher voltage, the linear part of the characteristic gets bigger, but the knee stays the same size in absolute terms. This means that it's now smaller compared to the linear part. This means the tube has more headroom, and when it finally does distort, you can see from Fourier analysis that the harmonics generated will be higher order than if the voltage was low, and the knee took up a larger part of the loadline. The higher order harmonics lead to a brighter, more aggressive overdriven tone.

Therefore, you won't notice any difference in tone from supply voltage, until you actually start overdriving things. Loudthud's results agree with this (thanks Loudthud!)

[guitarmike2107]

Originally Posted by pdf64

My understanding was that the role of the engineer was to implement a solution to reach the desired result, generally by using previoulsy proven solutions to similar situations.
And it's the scientist's role to find out why a particular implementation works.
When that is known, engineers can then use this knowledge to inform as to why one solution might be preferable to another etc.

Not that his matters to this thread, unfortunately the definition of Engineer differs from country to country, for me an Engineer is “degree qualified” Applied Scientist, some would take that further and say that chartered engineers are the only true engineers, but whatever, And what many people here in the UK call Engineers is actually what I would call a technician. Hopefully that makes my thread make more sense then.

[MWJB]

"Therefore, you won't notice any difference in tone from supply voltage, until you actually start overdriving things." Maybe not at small intervals (20-30v), but if you say halved the plate voltage on V1 of a Super Reverb you will hear the difference at pretty well all useable volumes...likewise if you increased plate voltage from <150vdc to 250vdc or more. Agreed my examples are on the extreme side, but if there wasn't "any" difference, then there still wouldn't be "any" difference whatever the degree of rise/drop.

[Steve Conner]

OK, I should have perhaps qualified that by saying "you won't notice any difference provided all other factors are held constant." As Loudthud showed, decreasing the plate voltage decreases the gain of the tube. So, even though V1 doesn't distort, it'll not drive the following tubes as hard. And you will notice that. Even a "clean" channel is probably clipping on pick attacks, and the amount of clipping will change.

As Douglas Self observed (yes, I'm a hopeless Douglas Self fanboy) changing any one component in an amplifier usually has several effects. To hear just one of those effects in isolation, you might have to change several components in the right proportion. Or, as some ecologist said whose name I forget, "you can never do just one thing."

This is where Enzo and MWJB's engineering approach starts to show its value, since their wisdom consists of things that they've already tried out.