6L6GC / 6BG6 x4 Ultralinear Guitar Amp

[kevinkr]

This is clearly a musical instrument amplifier discussion, please read the headers on both Tubes / Valves and I & A forums as that makes it clear where this thread belongs. Thanks! It's moving..(Moved!)

[DF96]

As a trend, older amps have lower value grid stoppers, or none at all.

Older valves tend to be bigger, and have poorer gain at UHF so there is less risk of parasitic oscillation. It could also be that older designers are more likely to understand RF as well as audio, so less likely to make their audio amplifier stages also look like a UHF oscillator.

Grid stoppers (on the first stage) can reduce RF coming in from the outside, but mostly they stop the stages from oscillating.

On a separate issue, ultimately IM and harmonics have the same origin in non-linearity.

[nazaroo]

Originally Posted by DF96

Older valves tend to be bigger, and have poorer gain at UHF so there is less risk of parasitic oscillation.

So, is this directly related to older tubes having higher Miller capacitance?

It could also be that older designers are more likely to understand RF as well as audio, so less likely to make their audio amplifier stages also look like a UHF oscillator.

Yes I'd tend to believe that the older engineers were more RF-savvy, although they couldn't have predicted the modern RF-polluted environment.

I still haven't got much input on modern RF issues vs. the 1960s when there wasn't so much transmission at every band.

Grid stoppers (on the first stage) can reduce RF coming in from the outside, but mostly they stop the stages from oscillating.

This is interesting, the distinction you are drawing here.
Obviously oscillation can go all the way down to sub-audio (motor-boating).

What are the low-frequency issues and considerations here?

On a separate issue, ultimately IM and harmonics have the same origin in non-linearity.

I am intrigued:

Since some designs have different ratios of IM vs. HD,
what are the controlling parameters here?

I am tending to believe at the moment that IM comes mainly from poor interaction between tubes and powersupply hum.

I might believe there is a small component of IM coming from mismatches or imbalances in a Push-Pull topology, (or rather IM falls into visibility and consideration when HD is cancelled out by Push-Pull arrangments!).

But do you think IM is directly related to operating a tube in a non-linear zone? How so?

[DF96]

Originally Posted by nazaroo

So, is this directly related to older tubes having higher Miller capacitance?

No. Wider spacing means lower UHF gain because of transit time effects. Old valve bases and sockets can be made from materials which are lossy at RF. New valves are mainly of all-glass construction.

What are the low-frequency issues and considerations here?

None. Grid stoppers prevent VHF/UHF parasitic oscillation. They have no effect on lower frequencies, unless too big in which case they will affect HF loop instability if there is global NFB.

But do you think IM is directly related to operating a tube in a non-linear zone? How so?

If an output contains Fourier components not present in the input (including all inputs such as PSU) then they arise from nonlinearity - this is simple algebra and trigonometry. Valves are not the only source of nonlinearity, but they are a major source. Sometimes harmonics can be cancelled, sometimes IM can be caused by re-entrant effects, so levels are not necessarily equal but they do arise from exactly the same source.

Hum IM is merely one type of IM, and that comes mainly from valve nonlinearity as the PSU hum sweeps the bias point up and down the valve characteristic. All valves work in a "non-linear zone" - the only issue is how much nonlinearity is there?

If you haven't yet read it, try to get hold of a copy of the Radiotron Designers Handbook. That will cover lots of these issues. It is available online, but it is so large that I much prefer the paper version.

[Printer2]

I hope you do not mind my asking, what do you want out of this amp? You seem to be coming at it from a hifi perspective and other than a jazz amp that is not really the goal. You say you want to keep the output clean and use the preamp to generate your color, most want the output to do that job if possible. Are there any amps out there that you like the sound of? The focus on RF with the grid stoppers concerns me as the values chosen for a guitar amp has more to do with driving the tube into overload than with RF.

[nazaroo]

Originally Posted by DF96

No. Wider spacing means lower UHF gain because of transit time effects. Old valve bases and sockets can be made from materials which are lossy at RF. New valves are mainly of all-glass construction.


None. Grid stoppers prevent VHF/UHF parasitic oscillation. They have no effect on lower frequencies, unless too big in which case they will affect HF loop instability if there is global NFB.


If an output contains Fourier components not present in the input (including all inputs such as PSU) then they arise from nonlinearity - this is simple algebra and trigonometry. Valves are not the only source of nonlinearity, but they are a major source. Sometimes harmonics can be cancelled, sometimes IM can be caused by re-entrant effects, so levels are not necessarily equal but they do arise from exactly the same source.

Thanks for your input.
I guess what I mean is that any "IM" which is harmonically (musically) related to the music being amplified is a lot more tolerable than non-related IM (e.g. frequencies having no relation to the main musical material).

So, IM which is a result of new harmonics and/or interaction of legitimate harmonics, or which can be tolerated because of 'harmony' with the music is of no real concern.

It is more like the fact that for example, Power Supply hum, which is a fixed frequency and so almost NEVER musically related to the program material is the kind of IM which actually grossly interferes with musical enjoyment.

I have certainly listened to many 'poor quality' stereos with much enjoyment, because the distortion was mainly harmonic distortion and/or IM that was musically related to the music.

But I am constantly reminded of the annoyance of powersupply hum (60, 120, + harmonics + IM with program) even when a lot of 'masking' is occurring via loud playing.

Hum IM is merely one type of IM, and that comes mainly from valve nonlinearity as the PSU hum sweeps the bias point up and down the valve characteristic. All valves work in a "non-linear zone" - the only issue is how much nonlinearity is there?

If you haven't yet read it, try to get hold of a copy of the Radiotron Designers Handbook. That will cover lots of these issues. It is available online, but it is so large that I much prefer the paper version.

I have a copy here, along with just about every published book on tubes circulating from 1940-1980.

Its probably more to do with my memory retention in my old age.

[nazaroo]

Originally Posted by Printer2

I hope you do not mind my asking, what do you want out of this amp? You seem to be coming at it from a hifi perspective and other than a jazz amp that is not really the goal.

Well, yep I am coming from a 'hifi perspective'. And I love Jazz.
The goal for me is a really musical guitar amp.
If that means it ends up "outside the box" as far as previous guitar amps and experience goes, oh well, I won't be heartbroken.

You say you want to keep the output clean and use the preamp to generate your color, most want the output to do that job if possible.

I'm sorry to hear that, because I know from my own experience that the kind of distortion that comes from overloading output transformers and crappy unregulated power supplies is not musical.

A Push-pull stage is not the place to introduce 'harmonic distortion'.
That should be left to single-ended (pre-amp) stages.

If you're talking about 'soft-clipping' that is another matter,
but one which can again be solved and better controlled in an input stage.

Are there any amps out there that you like the sound of?

Short answer: no.
I'm a perfectionist, and I can't help but mess around and improve things.

The focus on RF with the grid stoppers concerns me as the values chosen for a guitar amp has more to do with driving the tube into overload than with RF.

I think we agree here.
I actually AM more concerned with how the driver stage will be affected than by RF worries. What I see in popular designs is over-compensation for RF, and little concern for grid current effects during performance.

[DF96]

Originally Posted by nazaroo

I guess what I mean is that any "IM" which is harmonically (musically) related to the music being amplified is a lot more tolerable than non-related IM (e.g. frequencies having no relation to the main musical material).

So, IM which is a result of new harmonics and/or interaction of legitimate harmonics, or which can be tolerated because of 'harmony' with the music is of no real concern.

It is more like the fact that for example, Power Supply hum, which is a fixed frequency and so almost NEVER musically related to the program material is the kind of IM which actually grossly interferes with musical enjoyment.

IM is rarely harmonically related to the music. That is why IM is often more annoying than a similar amount of harmonics. However, some IM may be in rational ratios with the music so could sound like a chord.

Hum IM can be an issue with push-pull stages which cancel direct hum but not IM. I suppose some people might prefer some hum IM - fast vibrato?

Large grid stoppers may have a small effect on blocking distortion caused by grid current in an overdriven stage, but unless the grid stopper is large compared to the grid bias resistor it won't help much because the coupling cap still has a much larger resistance to discharge into than was present to charge it. The output impedance of the previous stage is relevant too.

[nazaroo]

Originally Posted by DF96

Large grid stoppers may have a small effect on blocking distortion caused by grid current in an overdriven stage, but unless the grid stopper is large compared to the grid bias resistor it won't help much because the coupling cap still has a much larger resistance to discharge into than was present to charge it. The output impedance of the previous stage is relevant too.

This is interesting, because its the opposite of what I have thought:

I supposed that lower resistance meant a higher driving voltage at a given preamp/driver level (and smaller accumulated distortion from previous stages) and less strain on the first half of the amp.

I assumed that high resistances would cause not only voltage loss requiring higher voltages in previous stages for a given loudness (and more distortion), but also a much less linear behaviour by the output tubes, because of the 'voltage-divider' effect of the resistor in the circuit with the grid current.

Of course I have never seen any proper detailed explanation of the interaction of the components as the output tubes are driven into grid-current conditions.

The little I could gather comes from the strange curves (when makers bother to provide them) as grids are driven positive.

[Printer2]

Originally Posted by nazaroo

Well, yep I am coming from a 'hifi perspective'. And I love Jazz.
The goal for me is a really musical guitar amp.
If that means it ends up "outside the box" as far as previous guitar amps and experience goes, oh well, I won't be heartbroken.

And a Fender, Marshall, Vox, or any of their offshoots are not musical?

I'm sorry to hear that, because I know from my own experience that the kind of distortion that comes from overloading output transformers and crappy unregulated power supplies is not musical.

Not talking about overloading output transformers, its the tubes where the action is. Don't like sag? Guess you fall in line with the high gain metal guys.

A Push-pull stage is not the place to introduce 'harmonic distortion'.
That should be left to single-ended (pre-amp) stages.

Well if you like thin preamp distortion.

If you're talking about 'soft-clipping' that is another matter,
but one which can again be solved and better controlled in an input stage.

So it is a clean amp you want. Why not go SS?

Short answer: no.
I'm a perfectionist, and I can't help but mess around and improve things.

As said above, maybe tubes are not for you.

I think we agree here.
I actually AM more concerned with how the driver stage will be affected than by RF worries. What I see in popular designs is over-compensation for RF, and little concern for grid current effects during performance.

Most designs now days take overloading the grids inputs in consideration. What may look like RF mitigation is more tone shaping and nonlinear operation control.