DIY Tube Preamp

[fdegrove]

Hi,

A plot of the 5687:





Cheers,

[Joel]

Originally posted by Brett
...I agree with your sentiment about freedom of speech.
However there is a difference between, "I think you are wrong, here's why...", or "My experience is different...." and shouting <b>you are all deluded</b> without being able to back it up.

Except when I talked about a "mass delusion", it was quite obviously in a joking manner. If you have no sense of humor, I feel sorry for you.

When you talk the talk, better be able to walk the walk. Now when someone like Thorsten makes a big claim, I stop and listen

What 'big claim' did I exactly make? It seems to me that you're the only one making sweeping generalizations. And yes, I have no doubt that when any fashionable "guru" speaks, you diligently listen.

because to my thinking audio is probably 50% art, 50% science.

And this is your whole problem. Art is painting a picture. Your painting is as valid as mine. Audio amplifiers are not art. They are machines. Either they work, or they don't. Everything they do can be measured, either now, or in the future with better tools. There is no mystery involved.
If you are truly serious about learning something on this forum, as you say you are - you will stop your childish posturing.

[planet10]

Originally posted by Joel
Everything they do can be measured, either now, or in the future with better tools. There is no mystery involved.

What can be measured only in the future, is a mystery now.

dave

[fdegrove]

Hi Joel,

Welcome back to the scene.


Hey Dave,

What can be measured only in the future, is a mystery now.

I'll have to ask my guru about that.

Cherio,

[dhaen]

Joel said:

Everything they do can be measured, either now, or in the future with better tools. There is no mystery involved.

Unforunately, there is mystery because I suspect that either:
1) We miss some important measurements,
or
2) We put our measurement data in a form that does not prioritise the elements correctly Eg THD.

I agree that it should be a science more than an art, but until the measurement data is interpreted more knowledgeably, the art factor remains.
Whether things will get better; science (repeatable), or worse; art (unique), remains to be seen. It will depend on whether people use science or voodoo to make judgements.

Cheers,

[6BG6GA]

Hi

A few questions if I may:

If given a choice between a 5687 based design and the two designs posted here using the 6SN7 which would be better in terms of sonic quality ? What would be the good points and bad with respect to design ?

http://www.netaxs.com/~vkalia/diy.html (5687 design)

Your comments will be highly appreciated as it will be a learning experience for me.

Many thanks,
Joe

[dice45]

FBLEAGH,

1st, i fully second what Brett told you here technically, re-read his posts, please.

I feel uneasy about the SRPP thing posted on the 1st page of this thread. Let me tell you why, please keep in mind the message comes from a zero feedback amp high efficiency speaker guy.

A 6SN7-based SRPP linestage would make about µ/2; as the 6SN7 has a µ of 20; this would make gain 10. But there is a conglomerate of huge cathode bypass capacitors used which boosts the gain to, say, 14. A linestage with gain 14 is just like shooting with cannons on sparrows IMO; nowaday line sources come along with 2Vrms and the poweramp sensitivity is also around 1 or 2Vrms with most amps. If an ancient tuner output comes along with 200mV rms, well, as the attenuator is never fully open in real life, you probably have that 10dB gain left.
I would not use more gain than needed.
Shooting with cannons on sparrows is ok if we talk about low output impedance and i mean not calculatory one, i mean one usable in real life. Where cable capacities and possible input inductivities can occur.

Then: i find 100k for the stepped attenuator. 100k with stray capacities of a stepped attenuator makes a nice upper bandwidth limiting already. Fellow Munich Triode Mafia members disagree if 10k or 50k is the maximum tolerable if the attenuator is located at the input of the line stage (BTW all are striving to have a 1200R or 600R attenuator at the output). 100k is just too high. Yeah, i know, there are well-reputed commercial preamps out there with 500k attenuators. OK for those believing in famous amplifier brands. Back then when i worked in hifi stores, those units soon lost their fame to me.

Then:
I once had a preamp running at my place made completely from SRPP stages. Nice sound, better than anthing i knew from the store, nevertheless somewhat blurry. My suspicion grew that an SRPP is great for gain stages (phono) but unusable as output with low impedance.
Finally i tried out different topologies for the line stage: high-G_m / low µ common cathode stage (12B4 ? so i faintly remember), cathode follower (not yet a transformer-coupled output stage, i was not yet mentally ready for that.)
Even with the cathode follower the sound was no longer blurry. Ok, it was a bit rough and nasty but had a lot of punch.
With the 12B4? common cathode stage It was neat and clean and fast. I loved it.
Back to SRPP linestage and

I am reporting listening experiences here but they seem to be well aligned with what Brett posted.
I have seen a 5687-based SRPP here and i would expect it to work much better than a 6SN7-based one. But then, remembering my "fancy" SRPP experiences with the ECC81, a tube having twice the transconductance of the 6SN7, i permit to myself the prejudice that there are better uses for a 5687 than an SRPP.
The line stage from vkalia (link posted by burnedfingers) for instance. Although i am not too fond of paralleling tube sections.
Or, probably better, the cheapskate's transformer-coupled line stage on my good friend Manfred's homepage which is a parafeed circuit.
Or, very probably better, the thing i am using currently: 50k attenuator looking DC at the grid of a battery-biased 5687 which is driving a step-down output transformer, a Simon Shilton pa102. No coupling capacitor. Manfred built that too.
Just outstanding sonically. Okok, nowadays one has to pay thru the nose for the Shilton trannie, therefore i hinted to the cheapskate thing which is close in performance.

[Brett]

<b>...I agree with your sentiment about freedom of speech.
However there is a difference between, "I think you are wrong, here's why...", or "My experience is different...." and shouting you are all deluded without being able to back it up.</b>

<i>Except when I talked about a "mass delusion", it was quite obviously in a joking manner. If you have no sense of humor, I feel sorry for you.</i>

Good try, but I saw no sign of there being any humour involved, merely you trying to assert the correctness of your position by shouting the loudest.

<b>When you talk the talk, better be able to walk the walk. Now when someone like Thorsten makes a big claim, I stop and listen</b>

<i>What 'big claim' did I exactly make? It seems to me that you're the only one making sweeping generalizations. And yes, I have no doubt that when any fashionable "guru" speaks, you diligently listen.</i>

Ask asked before, please quote in context.
I have not made many generalisations at all. The things I've posted here are born of my direct personal experience and University and College studies.
The "fashionable guru" comment is really disingenuous as you didn't bother to quote the end as well <i>"...as he has enough 'runs on the board' with me to know he has looked at the topic under discussion and made a reasoned decision based upon analysis and experience. I don't often agree with Thorsten, but I almost invariably learn something from his posts." </i>
I'm not a supporter of Thorsten's beleifs in general, but his analyses are usually pretty good, and often come from a different way of viewing things from the generally held orthodoxy. Looking at things from a different angle is a great way to learn, even if you eventually dismiss them as not appropriate, because the benefit comes from the process. However, if you wish to analyse any of Thorsten's analyses on audio design, I would be very interested in reading them.

FTR, I have no audio guru's. Note it was Bernhard who used that word, not me, in his reference to V&W. I repect greatly the work of veterans such as Langdon-Smith, Norman Crowhurst, DEL Shorter, Valley and Wallman, the Engineers at HP, Tektronix, WE and RCA, as well as modern Engineers / Commentators such as Steve Bench, John Broskie, Lynn Olson, Allen Wright, Garry Pimms and a small number of others. The circuits I've designed and built are based on the work of these people and others as well as my own experimentation.

Quoted in full it clearly means a different thing, but that wouldn't help you make a cheap shot then, would it?

<b>because to my thinking audio is probably 50% art, 50% science.</b>

<i>Audio amplifiers are not art. They are machines. Either they work, or they don't. Everything they do can be measured, either now, or in the future with better tools. There is no mystery involved.</i>

Do you seriously beleive that the current sets of measurements entirely categorise the sonic attributes of any amplifier? I don't, though I agree that in the future forthcoming technologies may be able to do that, but <i>today</i> they can't IMO/IME. Then why not just use 5534's for all your pre and driver circuits? They measure better in the standard tests than tubes do, and are smaller, lighter, more energy efficient, produce less heat, much cheaper and don't have any HV supplies to concern yourself about. Because tubes <i>sound</i> better, when done well. Irrespective of how well a spec-an says an amp measures, there is still the element of human interraction. If inserting a "technically perfect" amp into a system removes the emotional interaction with the music for a listener, it fails as a product. Music is not an intellectual experience. As an engineer, everything you do will be a compromise in one way or another as perfection is not directly attainable. The "art" part of my comment way using your experience, knowledge, ears and heart to choose those compromises within the design that are musically congruent.

dave also said in response to you, <b>What can be measured only in the future, is a mystery now</b>. I agree, and so does history.

<i>If you are truly serious about learning something on this forum, as you say you are </i>

I learn something from nearly every poster, whether it is their intention or not. Yet again though, you have only quoted part of what I said. I mentioned Jocko and Phred before, and those above, but I've even learnt from you.

<i>you will stop your childish posturing.</i>

I didn't think I had been. But as you're telling me I am, I shall have to bow to the superior experience of an expert such as yourself.

[guitaristz]

Thanks, Bernhard. I discovered battery bias recently and am blown away by how much better it sounds. Too bad aaa's are so large. Your quest was a good read vs. the Insoluable.

I'd rather not have to use a transformer quite yet. It might be icing on the cake later though. You're right about pots too matching to lower values is best; something guitar/amp makers should realize!

It seems that the 5687 is better here than the more common 6sn7? Futhermore, just by adding the circuit's latter half, one can add a headphones-out to any tube amp. I also just discovered the 12ay7 now, and love it. If there is something better we need to hear about it!!

I think most users of this design rarely crank their cans anyways so we don't need mucho watts .