• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Tube preamp questions? DIY?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
So just curious? So most line level sources (cd player, DAC, tape)are around 2 volts which will drive must any amplifier into clipping... so why is there a gain stage in a preamp? Why not just a resistor type volume control in the audio path?

Also any recommendations for a simple home brew hifi tube preamp
 
Rarely do I ever need/want additional gain from a preamp, so I usually use a simple cathode follower after the volume control to ensure a low source impedance.

OTOH, a little bit of gain is sometimes nice if you swap out equipment frequently (especially phono stuff).
 
So just curious? So most line level sources (cd player, DAC, tape)are around 2 volts which will drive must any amplifier into clipping... so why is there a gain stage in a preamp?

One potential technical reason to have gain is that some CD's are mastered well below full scale of the medium. So, while a full scale CD signal may produce an 2VRMS output, those CDs mastered below full scale will not.

Now, for an perhaps controversial answer. Certain preamps with gain may add musical realism to the subjective reproduction. Yes, such added musicality is an objective distortion of the signal, but, paradoxically, can subjectively sound less distorted, more real and less fatiguing on the the ear.
 
So just curious? So most line level sources (cd player, DAC, tape)are around 2 volts which will drive must any amplifier into clipping.


Many line outputs are much lower than 2Vrms, for instance modern battery powered phones and mp3 players(*). Conventional line level is -10dBV or 0dBV for domestic, +4dBV for professional, although these are very nominal values.


(*) these use earbud headphone levels typically
 
I'm surprised there aren't more simple two-stage tube line stages with negative feedback to reduce the gain to something like 2X to 4X (+6 to +12dB).

Why not take a 6DJ8 (ECC88), first stage common cathode, second stage cathode follower, with local shunt feedback taken from the cathode follower output to the input grid of the first common cathode stage?

Maybe...?
 

Attachments

  • _6DJ8_LineStage_NFB_rongon.png
    _6DJ8_LineStage_NFB_rongon.png
    17.3 KB · Views: 484
Why not just a resistor type volume control in the audio path?


There is no obvious technical reason for this. Over the years i have made numerous attempts to live with a passive attenuator but it never lasted long, even in very impedance-sympathetic systems. It is one of the many mysteries of audio.

I could probably live happily with a passive attenuator in an analogue system, but digital, no matter how good the dac seems to demand a separate active stage in order to obtain sound with substance, texture, weight and rhythm.

It is certainly not a universal preference as some listeners place more importance on the unsurpassed transparency of passive attenuators.

At the same time getting an active pre to sound right is among the hardest things in audio.
 
I'm surprised there aren't more simple two-stage tube line stages with negative feedback to reduce the gain to something like 2X to 4X (+6 to +12dB).

Why not take a 6DJ8 (ECC88), first stage common cathode, second stage cathode follower, with local shunt feedback taken from the cathode follower output to the input grid of the first common cathode stage?

Maybe...?

I like this idea, and it's been on my project list for a while. Only modification I might make is to put some Jensen transformers on the inputs and outputs, but I get that mode people don't feel a need to have balanced inputs and outputs on all their equipment. I like it for compatibility reasons, but my stuff often has to interface with pro gear, and my reference amp (MC650 from MC^2) has differential inputs.

The idea I've been dying to try, but haven't had time, is to do a fully differential vacuum tube preamplifier. A while back I designed a balanced discrete solid-state microphone preamplifier. As a means of testing the flexibility of the design I configured a pair of modules for a gain of about 3dB and built them into a preamp with Jensen transformers on the ins and outs. The preamp module itself has distortion around .004%. It gets hard to measure it accurately at those levels. The transformers add just enough "mojo" (distortion) to make it fun to listen to, but it is still very much a "reference system", so to speak.

I've been thinking of trying something somewhat like this using tubes, only I would likely eliminate the second gain stage. No need for that much open-loop gain. I was thinking of using 6BQ7s for the inputs and 12CA5s for the output (little beefier tube to drive long cable runs). Oh, and of course, transformer balancing.

I like to have around 3-5 dB of gain in a line stage. That way it can take the output of a consumer DACs and CD players and drive most professional gear to full power. Some of the power amps I've designed have relatively low gain and require a pretty hot preamp output to drive as well. For most people, 3dB of gain in a line stage is more than enough, and a buffer would likely be sufficient.
 

Attachments

  • H713 Microphone Preamplifier V4.pdf
    53.3 KB · Views: 94
Last edited:
An excellent illustration why 99% of commercial preamps sound really bad:

1 NFB

2 CF

3 Resistive loads

Given such options i would rather take a simple TVC and be done. Greta would be happy too.

An excellent example of sweeping generalisations and of damning often-appropriate solutions in favour of a fad-ish devotion to a specific technology, all without reference to external requirements.
 
I've said this before, I know it's unpopular, and I don't care. Prove me wrong.

The reason that people like single-ended amplifiers with no negative feedback is that they add harmonic distortion that people find pleasing. This is especially true when you have a gapped transformer with DC flowing through the primary.

It's okay to like this distortion, but recognize that when asked which is a better designed amplifier, most engineers will choose a design with a CF and a bunch of negative feedback with 0.005% THD rather than the fad design with 4% THD. I know I certainly would.
 
An excellent example of sweeping generalisations and of damning often-appropriate solutions in favour of a fad-ish devotion to a specific technology, all without reference to external requirements.

Consider it a beleave of the poster, so do I anyway. Then there's genuine experience with actual circuits, but again who's talking? Then there's measurements of course... to complicate things :)
My best volume control is with a 10K autoformer by Tribute, straight at the analogue output of a highly modified Marantz CD-player. No outboard DAC solution surpassed this digital source to date. But I've only heard a couple of dozens so far.... It's all relative and when is good good enough?
 
So just curious? So most line level sources (cd player, DAC, tape)are around 2 volts which will drive must any amplifier into clipping... so why is there a gain stage in a preamp? Why not just a resistor type volume control in the audio path?
A preamp easily can be designed to yield a very much smaller output impedance, compared with a plain potentiometer or other resistive divider. This allows to chose among much more interconnects.
An excellent example of sweeping generalisations and of damning often-appropriate solutions in favour of a fad-ish devotion to a specific technology, all without reference to external requirements.
+1!!!
Best regards!
 
As I see it, the reason this interminable argument about line stage/preamps vs. 'passive preamps' lives on is that:

1) Nobody needs the gain, and there is no commonly available triode (or triode-wired pentode) that when configured as a common cathode amplifier stage has:
- an open loop gain of only 2X to 4X
- plate resistance <2k ohms
- requires <10mA plate current for most linear operation
- is readily available and not too expensive, and
- 'sounds good'.

Maybe triode-wired EL86 running at 20mA plate current comes close? Gain of about 5X, Zout of about 800 ohms.

Triode-wired 6V6 would also come close (Ip of 20mA or so), but the gain without feedback would be more like 8X, and the Zout would be approx. 1.5k ohms.

2) It may be that in most home audio playback systems, low impedance/high-enough current drive of downstream cabling makes a noticeable improvement.

The output impedance of a 50k ohm potentiometer at 50/50 resistance position is 12.5k ohms (about 2 or 3 o'clock for a typical log pot). Maybe that's high enough to cause some psycho-acoustically discernible loss of 'presence' or other desirable qualities of audio reproduction?

The output impedance of a 10k ohm pot is 2.5k ohms at 50/50 resistance, which should be fine for anything. Maybe that 10k input load is low enough to load down/muddy up some sources?

A simple cathode follower would cure all these problems and (if competently done) should introduce no new ones. But I know there are people who claim they can hear when a CF is in the signal chain and that they all 'sound bad'.

So what do we do?
_______________________________

I have been very happy with a Dave Slagle autoformer volume control (thanks Dave!). I've used it for something like 10 years and it still makes me happy. This is not one of the very early ones that added some transient overshoot to signal peaks (easily discernible on a 'scope). This is one of the later ones that has extremely good behavior, even when fed a 10kHz square wave. Just adding that Slagle autoformer as a 'passive preamp' makes me like my playback system better. (I don't know why. I just like it, so it stays.)

Whether using the Slagle autoformer or not, I enjoy the added 'body' that an added line stage gives driving my current 21st century apartment listening setup with active speakers (JBL LSR305, believe it or not!). I can't explain what it is that I like about it. Add the 6FQ7 > 6DJ8 cathodyne line stage and the sound gains 'heft' and 'weight' (sorry about the Stereophile-style subjective lingo). Is that a bit of 2nd harmonic added? I don't know. BUT...

Take away the Slagle autorformer, put in a 10k or 50k (Noble) potentiometer, or even a 50k stepped attenuator (using SMD resistors) as a 'passive preamp' and the sound gets thin. Again, I don't know why. The resistive attenuator should be far superior to any other attenuator.
____________________________________

Anyhow, I haven't built that anode follower line stage I posted earlier, but if a little gain and a good low impedance drive are the necessary ingredients to make a system sound better then that circuit should do it. But if what I like is the whatever-it-is one gets from a feedback-less common cathode triode in the chain then that circuit will not help.

In the end, one has to build it and listen.
--

PS - Using LTspice, I designed a phono stage for a friend, with the goal of making it sound euphonic. I designed in a slight bass boost and a predicted THD higher than I usually aim for, =/>0.1% at 1V rms output (1kHz). He built it and everybody who has heard it loves it.

That reinforced my belief that most listeners (including myself) prefer a little 'good' harmonic distortion with their playback, but not too much. However I don't have a statistically large enough pool of examples for certainty.
--
 
Last edited:
My solution is not very dissimilar. I use a resistive attenuator followed by a differential transformer loaded valve stage with a measured thd around 0.02% @1khz. Not sure if this qualifies as "good" thd.

Otoh my many attempts to find/build a valve poweramp which does not turn half my music collection into caricatures have all been unsuccessful.
 
--

PS - Using LTspice, I designed a phono stage for a friend, with the goal of making it sound euphonic. I designed in a slight bass boost and a predicted THD higher than I usually aim for, =/>0.1% at 1V rms output (1kHz). He built it and everybody who has heard it loves it.

That reinforced my belief that most listeners (including myself) prefer a little 'good' harmonic distortion with their playback, but not too much. However I don't have a statistically large enough pool of examples for certainty.
--

I'd probably tend to agree with you. Looking at the success of the old Neve 80 series consoles (and the 1073 mic preamp in particular) is interesting in this regard. Rupert Neve used a gapped output transformer (LO1166) with DC flowing through it. It used a 2N3055 running in class A rather than a tube, but the distortion it introduces clearly isn't making people unhappy.
 
An excellent example of sweeping generalisations and of damning often-appropriate solutions in favour of a fad-ish devotion to a specific technology, all without reference to external requirements.

+1

A strong cathode follower preamp can help drive tricky cables, and buffer the effects of a pot. Done right they sound fine.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.