And that article particularly discusses the 4S preamp (without output buffer). Which is why I also pointed out the contents for the 12AU7 Color preamp which does include a buffer for lower input impedance solid state amplifiers. The topology decisions are nearly identical, differing only in the capability to drive output impedances.
Did you even read the writeup of the 12AU7 Color Preamp? The entire intention was driving a solid state amplifiers. Which is the primary reason for the selected design point of the buffer CF.
I really fail to understand the confusion surrounding this subject. Typical vacuum tube coloration is caused by even order harmonic distortion in geometrically decreasing magnitude with order. In a triode stage, the even order harmonic distortion is roughly proportional to drive level. Therefore, if one wants to preserve that "color" then the drive should be managed to produce that coloration and the signal level should be set subsequent to the harmonic generation. Hence, the volume control following the gain stage.
If minimum distortion is the goal, then the signal attenuation (volume control) should come BEFORE the gain stage to minimize its voltage swing and therefore (because the even order harmonic distortion is roughly proportional to drive level) it's total harmonic coloration.
Two different topological decisions based on two different (and some might say opposed) design requirements.
From reading these posts, I don't think 3/4 of the respondents understand what the goal even is, nor did they read the linked webpage to try to understand it. If we want a super low distortion pre-amp, CLEARLY this wouldn't be how anyone would design it. I know some of you think I'm a total idiot, but I'm not that stupid.
I was simply trying to wrap my head around why the pot was on the output vs the input like I normally see it, so I can explain this to my viewers, and I now 100% understand his reasoning for doing that on this "Color" Pre-amp: he is looking to preserve the even order harmonics created by the driven gain tube to add this coloring to the sound at any volume level. 99% of the people using this will use unity gain or less, they don't need a pre-amp or any extra gain. They are just looking to warm up the sound of their system and warm sound is not the same as EQing some extra bass. And just tossing a cathode follower/tube buffer won't get you there either. If they just wanted super clean sound, why even bother with tubes to start with?
I also want to apologize to Matt, I should have just emailed him this question instead of posting this here for people to attack his design, without even understanding what the goal is.
I was simply trying to wrap my head around why the pot was on the output vs the input like I normally see it, so I can explain this to my viewers, and I now 100% understand his reasoning for doing that on this "Color" Pre-amp: he is looking to preserve the even order harmonics created by the driven gain tube to add this coloring to the sound at any volume level. 99% of the people using this will use unity gain or less, they don't need a pre-amp or any extra gain. They are just looking to warm up the sound of their system and warm sound is not the same as EQing some extra bass. And just tossing a cathode follower/tube buffer won't get you there either. If they just wanted super clean sound, why even bother with tubes to start with?
I also want to apologize to Matt, I should have just emailed him this question instead of posting this here for people to attack his design, without even understanding what the goal is.
About the concept of "color": I have zero explanation for it, but a little tube amp driving a classic solid state amp does sound good...🤓
Observing that class-D amps can sound detailed - but sterile, a nice old-school tube pre-amp with Allen-Bradley carbon comp resistors does indeed balance things out. It does add "color"...
Observing that class-D amps can sound detailed - but sterile, a nice old-school tube pre-amp with Allen-Bradley carbon comp resistors does indeed balance things out. It does add "color"...
i will use the 6ew7 in this preamp, if you wanted higher currents in the cathode followers,
i will use single shaft volume control in the input side, and 10k volume pots in the output for balance controls..
i will use single shaft volume control in the input side, and 10k volume pots in the output for balance controls..
Least distortion is a normal design parameter in HiFi regardless of topology. It is different for guitar amplifiers and effect generators in musicians gear.
This is generally a HiFi DIY audio site so it is a bit swimming against the stream. Not many will deliberately design for distortion as that is not High Fidelity.
Last edited:
I would argue most people listening to tube gear is because of the 2nd (good) harmonic distortion is has. If it all had 0.001% distortion, why would anyone bother with hot, inefficient, limited lifespan tubes?
I would argue that valves are more linear than (conventional, cold biased) semi-con amps in the region where it matters, around zero crossing. We're very sensitive to distortions that are not monotonic with signal level, meaning decreasing continuously, even below measured noise floor, at smaller and smaller signals. A classic example is loudspeakers (or many other transducers), huge distortions at high levels but decreasing monotonically. We forgive their trespasses and can listen past them. I suspect that this is a large part of the draw of SET amplifiers - they're inherently monotonic with level, even the OPT's primary inductance - despite their limitations.
In this respect vacuum valves have a great advantage in being able to operate safely at high temperatures, unlike semi-cons which much be coddled and protected with large heat radiators, and are still largely limited in performance by their thermal fragility.
Monotonic operation in amplifiers comes with thermal inefficiency, not even a problem for vacuum valves. If we measured amplifiers at the levels that actually mattered, instead of the 1950s paradigm that we do, we'd see superior performance. Better measurements are now becoming widely available, with excellent modern ADCs and software spectrum analyzers, to see down below the conventional noise floor. We're in a new golden age.
All good fortune,
Chris
In this respect vacuum valves have a great advantage in being able to operate safely at high temperatures, unlike semi-cons which much be coddled and protected with large heat radiators, and are still largely limited in performance by their thermal fragility.
Monotonic operation in amplifiers comes with thermal inefficiency, not even a problem for vacuum valves. If we measured amplifiers at the levels that actually mattered, instead of the 1950s paradigm that we do, we'd see superior performance. Better measurements are now becoming widely available, with excellent modern ADCs and software spectrum analyzers, to see down below the conventional noise floor. We're in a new golden age.
All good fortune,
Chris
Well said @Chris Hornbeck. I would only add that : 1) loudspeakers are unable to generate enough high order harmonics due to their mechanical nature; 2) lower damping factor of tube amps is more often an advantage, regardless of common hype, from the dynamic point of view because it is much less affected by the non-linear behaviour of transducer response in time domain. Despite the non-linear part of transducer's suspension/motor is generally small-to-very small, its effect on impulse response can be significant with high damping factor amps. Especially if that series of impulses (music) will be perceived with an emotional meaning attached to it. That also explains why 2 apparently extremely similar (on paper) solid state amplifiers can sound very different.....
Last edited:
There is another aspect
The CD out is nominal 2 volts and in some cases more.
With the pot on output we have always the first stage that amplify full the signal.
The THD is related to the level in input ( and a proper working point) so, in this case, we have the same shape of FFT for every step of the position of the volume installed on ouput.
This is not smart.
Walter
The CD out is nominal 2 volts and in some cases more.
With the pot on output we have always the first stage that amplify full the signal.
The THD is related to the level in input ( and a proper working point) so, in this case, we have the same shape of FFT for every step of the position of the volume installed on ouput.
This is not smart.
Walter
I don't think anyone is thinking that bad of you but the assumption that this distorting preamplifier that feeds a sterile SS amp is going to emulate a good tube amp is simply not true. If you drag in this discussion the tube power amp and assume that people use it because it's distorting then you get such reactions. You minimize (or stay below a threshold) yourself the distortion of your tube amps, so we all know what we are talking about.
You want to add "color" fine, no problem. What color this is going to be is fairly random if the deviation from an ideal HiFi preamp is too much.
What I defined "best spot" in post n.3 (that you discarded straightaway) is not my cup tea either and is already affected by fairly high distortion for being a preamplifier because the ECC82 is not a champion of linearity even at its best bias point. It will add "color" in a less random fashion, IMHO.
The ECC82 is actually one of best mainstream tubes for achieving 2nd harmonic cancellation in power amps thanks to its good predisposition to generate good amount of 2nd harmonic distortion and very little else if properly tuned.
If one can't hear a difference between high distortion preamp feeding a SS amp and a true good tube amp, I think he has a much bigger issue elsewhere. More likely the room and the speakers.
If I where to add "color" in wild (and tunable) manner I would just buy a $50 guitar tube screamer pedal and put it just upstream the power amp.
It's actually brilliant given that is the whole point of this pre-amp and why the pot is on the output.
Last edited:
I apologize for my confusion.
I didn't realize that your analysis of the volume control on the output was specific to the "universal / 4S" design. I hadn't even looked at that one since this thread is about the "color" version, so I didn't realize that you also put the volume control on the output of that one too.
In your earlier post you posted the direct link to the volume control analysis but you didn't mention that your very detailed analysis applied to a different preamp. Since this thread is about the "color" version, not the "universal", I assumed that it applied to the the "color" version.
I fully understand that sending an unattenuated signal to the gain stage will produce more "color" and I understand why there is interest in adding "color".
Apparently, your placement of the volume control is intended to add "color" to both the "universal" and "color" versions. The difference seems to be that one is designed for owners of tube amps who want to try the various 12A*7 types and the other is designed for people who are using SS and Class D amps who are content to stick with a single type.
As I mentioned earlier, the preamp I built uses pots on both input and output. In effect, its "color" is adjustable. Paul Joppa, the designer and owner of Bottlehead, described this arrangement earlier in the thread as "the ultimate . . . enabling the end user to experiment with varying levels of "color"."
It seems that adopting such an arrangement would make your designs more versatile and totally eliminate the "controversy" over having a single volume control on the output.
Your analysis of the pot on the output addresses the concern about the effects on high frequencies but makes no mention of how the various volume level settings affect low frequencies.
Earlier, I also posted concerns about how the combination of the 1uf cap and the 10k volume pot affects the high pass rolloff point when used with an amp with 10k input impedance. Perhaps my concerns were unfounded but it seemed to me that using higher cap and pot values would be better.
As I said in that post, if my analysis was wrong I was hoping that someone would correct me. I'd appreciate any comment you might have on that.
Last edited:
If the only purpose of this device is to introduce a nice cascading SE profile of tube mojo. Which I fully understand and support. Then Id say design it as unity gain stage with no volume control at all. Volume can be controlled in the digital domain of the source, or at the input of the amp it is feeding. Just make sure it doesn't start clipping below 2v. Simple SE mojo box, no controls. Better still make it a tube roller, something that can take a variety of tubes with 12AU7 base.
Last edited:
Here's another possible approach, inspired by Broskie, but depends on if the cathode follower doesn't draw grid current. Otherwise you'll get scratchy noise when you turn the pot. Certain bias point can also create grid current too. You still need an extra cap to ground at the low point of the point.
Here are couple Broskie's takes from his website on a "color" unity gain stage he called "harmonic restorer."
Last edited:
I am the designer of many of the Bottlehead circuits, but Dan Schmalle, a.k.a. Doc B, is the founder and owner of the company. I am compensated as a consultant only.
That said, Bottlehead has in recent years moved to putting the pot at the output of our preamps, for the reasons I cited. Paul Birkeland, another designer of Bottlehead circuits, originated this move and it has proved to be quite successful, both audibly and commercially.
That might be a great future project! The only issue would be: changing to tubes with more or less gain would ruin the "unity gain" aspect.
Some people seem to like the glow. 🙂
A friend of mine who tinkers a LOT with Nelson Pass amplifiers (which sound superb) still admits that his go-to reference is his PP 2a3 amp.
A large number of tube front-ends will distort tremendously when fed a full 2V. Once the stew is spoiled, it takes acrobatics to make it taste good.
But maybe this has been accounted for in this design?