This has probably already been discussed many times but the forum search engine refuses to search for sag :-(
I'm working on an experimental guitar amp for a friend, and I plan to do this in stages, i.e. first build a pre-amp and then start work on the power stage. As a result I will need a power supply for the pre-amp before I even know what the final B+ will be for the whole, so for a while at least I will be working with a power supply which is only designed to drive the pre. And that leads me indirectly to some speculations ...
One of the things I plan to play in this amp with is controlled power sag, and this leads me to ask: when sag occurs, does this primarily affect the power amp or is there also also a sonic effect on the earlier stages? My guess is that all stages are affected, so for example a pre-amp stage might go into clipping earlier when the plate voltage falls. Is this so, and is the result more likely to be "interesting", or just crappy?
I'm working on an experimental guitar amp for a friend, and I plan to do this in stages, i.e. first build a pre-amp and then start work on the power stage. As a result I will need a power supply for the pre-amp before I even know what the final B+ will be for the whole, so for a while at least I will be working with a power supply which is only designed to drive the pre. And that leads me indirectly to some speculations ...
One of the things I plan to play in this amp with is controlled power sag, and this leads me to ask: when sag occurs, does this primarily affect the power amp or is there also also a sonic effect on the earlier stages? My guess is that all stages are affected, so for example a pre-amp stage might go into clipping earlier when the plate voltage falls. Is this so, and is the result more likely to be "interesting", or just crappy?
With irony which is both thick and deep, the “whole point” of guitar amplifiers seems to be to string together … limp power supplies … over-strong gain stages … starved cathode effects … nonlinear elements (germanium PN junctions, for example), all in an attempt to creatively distort the incoming signal in sonically “interesting” ways.
And it works!
So, keep on trucking. You don't really have to worry about whether 'this' or 'that' would be sonically useful in the abstract sense. Just build, try, listen carefully, think on it.
⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅
And it works!
So, keep on trucking. You don't really have to worry about whether 'this' or 'that' would be sonically useful in the abstract sense. Just build, try, listen carefully, think on it.
⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅
Thanks Mooly, the google search works a lot better. And now I know that people worry about sagging downward-facing drivers too 🙂
GoatGuy of course the only definition of what is sonically interesting is what sounds interesting to me and (especially) to the guy that's going to be using it. I do know however from experience that a lot of things that look as if they should sound interesting end up sounding like just "blat" or "fart" so I was ready to receive some sage advice. When I said "experimental" amp I meant just that - I have some ideas (and a lot of them are in your list, lol) but above all my plan is to try things out as I go along.
Time to stop theorising and start stringing stuff together ...
Time to stop theorising and start stringing stuff together ...
Marshall made a big name and a fortune from none linear guitar amps since the 60's.
Most tube amps do not use voltage regulation... When designing a circuit, you design to the full load voltage, ie the sag voltage at full power output. That is the real design voltage to use... The voltages at idle are useless number for design...same goes for the pre-amp stages...
There, fixed in bold. For audiophile amplifiers and preamps, exactly the opposite is important, and is designed in. I just want future readers to remember where this thread is heading…
⋅-=≡ GoatGuy ✓ ≡=-⋅
I would rather say that idle and full power are two extremes, and most audio amplifiers will spend most of their time somewhere in between. For most purposes full power output is the "worst case" condition, and so a competent engineer will pay close attention to that case.
For the particular case of tube amps for guitars, the effect of voltage sag on the amp's performance results in a compression-like effect which some guitarists find useful some of the time. For these amps the designer may choose to consciously allow the performance of the amp to vary across its power range, taking care of course to avoid overheating or component failure.
GoatGuy I'm intrigued by the "germanium PN junctions" in your list. What makes germanium more interesting than silicon in this respect? Germanium devices do present a smaller voltage drop, but the main other thing I remember about germanium is its tendency to conduct better as the temperature rises, which might sometimes be useful but can also lead to auto-destruction.
For the particular case of tube amps for guitars, the effect of voltage sag on the amp's performance results in a compression-like effect which some guitarists find useful some of the time. For these amps the designer may choose to consciously allow the performance of the amp to vary across its power range, taking care of course to avoid overheating or component failure.
GoatGuy I'm intrigued by the "germanium PN junctions" in your list. What makes germanium more interesting than silicon in this respect? Germanium devices do present a smaller voltage drop, but the main other thing I remember about germanium is its tendency to conduct better as the temperature rises, which might sometimes be useful but can also lead to auto-destruction.
The VF (forward voltage drop) of germanium, depending on doping, ranges from 0.3 to 0.45 volts. So, there's that. Technically, one can achieve similar VF (if not lower) with metal-on-silicon Schottky diodes. And of course WAY higher heat tolerance.
However, 'way back when' (in the 1960s, but mostly early 1970s when Ge diodes were still cheap and common), it was found that in practice Ge PN was more 'interesting' acoustically. There is a softness to the VF which is musically competent.
This was also the era of the bar-of-soap effects-pedal boxes. Oh, there are still zillions of them, but the first Golden Era was centered on 1972. Well past the tube era (in compactness, voltage requirements), one could purchase pretty inexpensively germanium diodes, lil' JFETS, remarkably high performance low-noise NPN silicon transistors (and yah, Germanium ones too…) and little tough-as-hêll hard-cast aluminum enclosures into which all could be assembled with a couple of 9 volt batteries. That'd last for quite a few hours without going kaput.
Sure, today, (or in my case, as a Christmas gift to my budding steel-string rocker son, 10 years back), EFX pedals have matured hugely. And with a surprising sophistication of integration of many features into single double-or-triple wide pedal boxes. And still there are the bars-of-soap. Endlessly, with loudly emblazoned catchy names. ZoomBoom and so on.
So, the 'camps' have divided into basically 3 or 4 quadrants. One is “few-to-no EFX boxes, just the wickedly cool old-school tube amp” (and hand wound pickups, no 'active' amps, either), perhaps as an extreme. To the right of that, are the “old school tube amp”, but a big head, and lots of power, with a nice assortment of soapbars mounted on a solid stage footplate. A lot of serious guitar people are found here. Then there is the relatively new “Pretty Clean amp, plus either lots-of-integrated, or a totally stoked external pedal-box-array” camp. Often solid state through all but the preamp input and 6× 6550 output stages. Lastly, are the jazz/acoustic bunch that eschew almost all the EFX, and go for really clean, amplification, along with custom-wound bifilar coil pickups, built-in-the-body (of the guitar) FET based preamps, and remarkably competent speaker enclosures. The mad lads of clean-rad.
The first № 1 group really does capitalize on amps that “fit my first description”. Intentionally sagging power supplies, long strings of R-C power supply preamp-and-integrated-amp stage power filters, with curiously small capacitors with RC time constants on the order of a few tenths of a second. No attempts to power-supply regulate. Intentionally under-volted cathodes, barely visible in a dark room. All sorts of hard-won-and-nearly-magical circuits to shape tone, attack, overdrive, and much, much more.
All in all, I'd just venture that “you gotta figure out what camp you really are in” and just go “whole hog” once that clarity impacts your drawing board. Circuits will flow accordingly. Just resist the urge to borrow too much from the other camp(s).
Anyway, my unsolicited opinion about this matter.
⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅
However, 'way back when' (in the 1960s, but mostly early 1970s when Ge diodes were still cheap and common), it was found that in practice Ge PN was more 'interesting' acoustically. There is a softness to the VF which is musically competent.
This was also the era of the bar-of-soap effects-pedal boxes. Oh, there are still zillions of them, but the first Golden Era was centered on 1972. Well past the tube era (in compactness, voltage requirements), one could purchase pretty inexpensively germanium diodes, lil' JFETS, remarkably high performance low-noise NPN silicon transistors (and yah, Germanium ones too…) and little tough-as-hêll hard-cast aluminum enclosures into which all could be assembled with a couple of 9 volt batteries. That'd last for quite a few hours without going kaput.
Sure, today, (or in my case, as a Christmas gift to my budding steel-string rocker son, 10 years back), EFX pedals have matured hugely. And with a surprising sophistication of integration of many features into single double-or-triple wide pedal boxes. And still there are the bars-of-soap. Endlessly, with loudly emblazoned catchy names. ZoomBoom and so on.
So, the 'camps' have divided into basically 3 or 4 quadrants. One is “few-to-no EFX boxes, just the wickedly cool old-school tube amp” (and hand wound pickups, no 'active' amps, either), perhaps as an extreme. To the right of that, are the “old school tube amp”, but a big head, and lots of power, with a nice assortment of soapbars mounted on a solid stage footplate. A lot of serious guitar people are found here. Then there is the relatively new “Pretty Clean amp, plus either lots-of-integrated, or a totally stoked external pedal-box-array” camp. Often solid state through all but the preamp input and 6× 6550 output stages. Lastly, are the jazz/acoustic bunch that eschew almost all the EFX, and go for really clean, amplification, along with custom-wound bifilar coil pickups, built-in-the-body (of the guitar) FET based preamps, and remarkably competent speaker enclosures. The mad lads of clean-rad.
The first № 1 group really does capitalize on amps that “fit my first description”. Intentionally sagging power supplies, long strings of R-C power supply preamp-and-integrated-amp stage power filters, with curiously small capacitors with RC time constants on the order of a few tenths of a second. No attempts to power-supply regulate. Intentionally under-volted cathodes, barely visible in a dark room. All sorts of hard-won-and-nearly-magical circuits to shape tone, attack, overdrive, and much, much more.
All in all, I'd just venture that “you gotta figure out what camp you really are in” and just go “whole hog” once that clarity impacts your drawing board. Circuits will flow accordingly. Just resist the urge to borrow too much from the other camp(s).
Anyway, my unsolicited opinion about this matter.
⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅
Apparently none of others know the answer to this.
Answer is ; mostly in the Power Amp stage, and mostly when the PA is running class AB ; and mostly when is a tube rectifier for the high voltage, and mostly when the amp is a guitar amp ; as this effect gives the amp a little bit of attack when the performer is playing hard...........
Thanks, mooreamps. That means that if I start by building (and experimenting with) the pre-amp, and I first build a power supply just for that, it may make sense to retain that separate power supply in the final design. Then again it may not 😀, in which case the power supply in question can doubtless be re-assigned to another project ...