Question about touch responsive amplifier design

[jammybstard]

As I read more about valves and amp design I keep coming across references to touch responsive valves and amps but I haven't found a decent explanation of what technically makes an amp more or less responsive.
I seem to find with my single ended 5w kit amp (Ampmaker SE-5A) that if I crank the gain but keep the master down at bedroom levels that the amp sounds great but is very unresponsive. where as clean its very responsive. At least thats my perception.
I'm reading the Megantz tube design book at the moment which is answering alot of questions but I'd like to know what makes an amp responsive.?

[wyatt]

In your case, you've overdriven the preamp to the point of compression.

On most non-master volume amp, the Volume pot is placed right after the 1st gain stage. On master volume amps, this is often simply re-labeled as Gain. You need to find the sweet spot of the amp by cranking the Master Volume and treating it as if it did not exist. Then use the Gain pot as a volume pot and find that sweet spot where you can control how much it overdrives by touch dynamics or with the volume pot on your guitar. Then set the Master Volume for the desired output.

But lots of things affect how touch responsive an amp is...preamp plate voltages, power supply design, etc. But, ultimately, it takes a lower-gain preamp to allow for enough voltage swing to stay touch responsive. Hi-gain amps are designed for lots of clipping and compression.

[jammybstard]

So Its about maintaining headroom in the preamp stage.
So is tube distortion, low volume and response a possibility? perhaps by adding more preamp stages that are just starting to simultaniously hit the sweet spot

[sepulchre]

Having more preamp stages heads toward high gain design. One of the Most touch sensitive amps ever made, the Trainwreck Express, has merely three preamp stages - the initial stage, then the tone stack and then only two more stages before the phase inverter. This design actually leaves half of one of the tubes unused but that's how Ken Fischer designed it and it works very, very, very well. You can look at the schematics here in the TrainWreck Files Forum, or you can check out the AX84 4-4-0 which is a low wattage version here: http://ax84.com/4-4-0.html

[Stevem]

If you want a schematic of a great touch responsive amp, and the osciliscope shots to show you why, do a scearch for guitarnuts.com/ pig nose gv40 mods, you`ll be quite glad you did.
It far too bad these amps around made any more, I was such a jerk for not buying 3 more when there price was 220 bucks!!

[martin manning]

If you want a schematic of a great touch responsive amp, and the osciliscope shots to show you why, do a scearch for guitarnuts.com/ pig nose gv40 mods, you`ll be quite glad you did.
It far too bad these amps around made any more, I was such a jerk for not buying 3 more when there price was 220 bucks!!

Disclaimer from the above site:

"Note, the following plots were generated by simulations that are reasonably accurate. However, in the real world things are never quite so cut and dried. The plots are provided for illustrative purposes only, you shouldn't expect to see exactly the same thing on an oscilloscope measuring a real amplifier, so don't knock yourself out trying..."

[rdjones]

The single ended Class A amp is biased such that the zero signal idle current and the maximum output current drawn from the high voltage supply is nearly the same.
There's no "sag" when going from low volume to high volume playing.
The amp is basically operating at full current draw when at idle.

In a properly biased push-pull amp (as in not class A) the high voltage supply will drop a bit, or a lot depending on the design, when played hard.
This creates the compression or sag that contributes to 'feel' or touch sensitivity.
A tube rectifier will add to the voltage sag (in a class AB amp), some types more than others.
A power resistor in series with the B+ will also add sag.

The compression in the preamp stages is more about saturation and nonlinearities in the individual tubes near the extremes of signal swing.
The peaks of the waveform will distort differently at cutoff compared to saturation.
This also varies with tube type and circuit design, and is why different amp designs have their own sound character and feel.

So it boils down to this: you won't have power amp sag at bedroom levels unless you use an attenuator or an isobox.
In a master volume situation you get all the preamp compression that the particular design allows, but it doesn't have the same touch sensitivity or feel as playing loud does.

reddog Steve

[teemuk]

The compression in the preamp stages is more about saturation and nonlinearities in the individual tubes near the extremes of signal swing.

Indeed. The compression mechanisms from either B+ dropping or plate voltage swing approaching B+ do not really have a great impact. For a typical 12AX7 gain stage the gain compression from B+ dropping about 200V (!!!) is only about 1 - 2 dB. Hardly even perceivable. And that's a lot of sag for almost zero audible effect.

The greatest amount of compression comes from either variation of bias or screen voltage of a pentode tube. This practically means you can only get notable compression from pentode power tubes and only when...

- Screen voltage sags drastically
- Bias varies drastically

Usually this means that those conditions are met only when power amp is push-pull and specifically class-AB -type (and drawing varying amount of current from the supply depending on loading), driven into heavy clipping and screen circuit is deliberately designed to have great amounts of voltage sag. (Most screen circuits are quite well filtered and therefore do not introduce a lot of voltage sag). So, while compression is a real phenomenon -in certain types of designs - it seldomly happens under normal operation and usually requires heavy loading and some deliberate design for accentuated effects of power supply sag.

All in all, the great "tube compression" is once again more akin to those nice catchphrases that people like to throw around than a phenomenon of significant impact.

Touch sensitivity isn't neccessarily about compression. Take for example a typical class-AB tube power amp: When overdriven slightly it may have very soft asymmetric clipping from PI inbalance, drive harder and it usually turns to symmetric hard clipping of the power tubes, which is often modulated by power supply sag and ripple, drive harder and the power tube grids usually begin to conduct too and crossover distortion appears and gradually increases depending on loading, namely on envelope of the driving signal. So, in a certain type of circuit architecture you can have many different distortion mechanisms piling up and adding more and more harmonics when you overdrive more and more. It's a way different response to touch than what you get from a single stage that just clips the signal on a fixed threshold and does nothing else.

I think at least one important key to touch sensitivity is introducing that kind of time and amplitude -variant distortion rather than distortion just at fixed amplitude level.

Cascaded preamp stages designed properly usually introduce time and amplitude variant clipping too: One stage may clip asymmetrically which gradually skews signal's DC offset point having significant effect on created harmonics. Add another gain stage and it piles up another time and amplitude variant effect when the changing DC offset of an symmetrically clipped signal clipped again introduces duty cycle modulation and varying harmonic patterns. This again proves much different touch sensitivity to overdriving than just clipping the signal at fixed amplitude threshold and nothing more.

[teemuk]

The amp is basically operating at full current draw when at idle.

No. A Class-A amp draws about half of the full current at idle. Full current is drawn at the peak of one halfwave and lowest current at the peak of the other. The average current draw, however, remains pretty constant under normal operation.

Therefore no sag.

In a class-B or AB amp the idling point is close to point of minimum current draw. Current draw increases towards waveform peaks and therefore the average current draw varies much much more.

On that note, overdriving a class-A amp hard may temporarily misbias it (ever seen output of an overdriven SE amp on scope screen?) and actually cause similar varying current draw that takes place in class-AB amps. However, current draw even in this condition still isn't usually drastically higher than current draw at idling and therefore the effect of sag still tends to be quite insignificant. But technically under right conditions even class-A amps may portray some tiny amount of power supply sag too.

[Reeltarded]

Welp, my take is there are two things that people cross more often than not.

Touch responsive is a late tweed. Kerrang kerrang kerrang.

Good feel is not raw like that and is a high gain amp with moderate control settings. It plays itself and sounds like a horn if you are decent at hiding attack.

[boots]

I was thinking the same thing Reel just said. To me, touch sensitivity is like a tweed Champ driven just to the point of breaking up. Maybe about half volume. Or maybe full volume but driven with the vloume control on your guitar turned down to 4 or 5. When you dig in to the strings, it growls and breaks up, and when you ease off the string attack, it sounds pretty clean. Thus, responsive to your attack with the axe. You can get pop, twang, and growl all without any adjustments to the amp or volume control.

Some of SRV's recordings are perfect examples of my idea of touch sensitivity.


I guess sag is a different type of touch sensitivity, but maybe more of a feel than a sound.

[Structo]

I think there is a difference in what some people call "touch sensitivity" and what others mean by it.

I actually don't know the true definition, but to me it means pick dynamics, where you can control the sound by how hard you attack the string.
A light attack would give a somewhat clean low volume tone while hitting the string hard would give the full signal overdrive type tone.

Some amps do this better than others.

I really don't know how compression figures into the debate.

Some amps are definitely more squishy on the attack, like an old tweed sometimes is.
I think most likely that is a power supply sag that causes that.

Just my 1 1/2 cents.

[Tillydog]

And then there's the effect of the negative feedback loop collapsing...

[sepulchre]

Touch sensitivity isn't neccessarily about compression.

I really don't know how compression figures into the debate.

Thank you. I know compression possibly adds to the effect I know as touch sensitivity (well described by Tom), but it's certainly not the only component.

[VacuumVoodoo]

I've once built an over the top touch sensitive amp. It was enough to just touch it and it would blow all 7 fuses. Every time. Even when not powered up. The Fringe Division got very interested in it.

Seriously: touch sensitivity thing resides in 2 vs 4 gain stages preamp topology. Doesn't need to have a massive total gain but a smart gain - high mid emphasis - attenuation & de-emphasis sequencing between the stages.