Push-pull microphone preamp with UTC LS and Tamura transformers

[aazaa]

"commonly held beliefs"

Interesting, this contradicts commonly held beliefs about 6SN7 versus 12AU7. Most in fact seem to think of 12AU7 as horribly distorted.

During decades people happily used the ecc82/12au7 tubes.
Came in Morgan Jones. Since the publication of his comparative test on harmonic distortion the tube is condemned.

Let it be clear, the winning 6SN7 is a superior tube. Evenly spaced lines in the graphs. Low distortion at all operation points.

Morgan Jones tested all tubes at 8mA anode current, same high tension.
It seems logical and fair to test all tubes under the same conditions... Well, it is not. Unfortunately he had not the idea for an extra test to compare tubes at an optimal operation point for each tube. He jumped to conclusions.

Conclusion about the ecc82/12au7 tube:
"Later generation valves such as the ECC82 (also intended for use as a field scan oscillator) benefited from improved production techniques and distortion is extremely consistent from sample to sample; it is consistently poor."

It is tempting to change it in:
... were consistently tested at an inappropriate operation point...

also:
"The results in Table 3.1 speak for themselves. All of the alternatives are inferior to the 6SN7GT/12SN7GT and produce significantly more H3. The Loctal 7AF7 dual triode and B9A ECC82 dual triode are particularly ghastly."

- Morgan Jones' test: H3 distortion of the ecc82 is 29dB higher than the distortion of 6sn7.
- In my preamp:
6sn7: H3 is approximately -85 dBu at all operation points. An amazing result!
ecc82/12au7 at the best operation point: H3 is below -100dBu in all cases. The ecc82/12au7 tubes are indeed very consistent.

All forums adopted Morgan Jones' conclusions and the 'commonly held belief' of the ecc82/12au7 being horrible was installed.

On this forum:
"... 6SN7 and friends beats the pant off the 12AU7 for distortion..."
"...the Russian 6N1P and that is almost as bad as the 12AU7..."
"... Definitely not a 12AU7... a tube with the worst ever intrinsic distortion..."
etcetera...
In this thread I got a friendly suggestion to consider a better tube in post #49.

For the second build I am faced with the dilemma: Use the 6sl7/6sn7 octal combination - go for a more 'vintage feel' - as was the first intention, or go for the lowest possible distortion (inaudible difference) and continue with ecc83 and ecc82/12au7 ???


Paul

 

[aazaa]

6SN7
I don't know yet, but maybe there is a possibility to improve the performance of the 6SN7 with a higher B+ : 390V à 400V???
Maybe its best operation point needs more voltage across the tube???

The tests were done at B+ 288V. I can't test higher now.
Will try later.

Paul

PS Of course I know. It is ridiculous to try to improve its performance. Distortion is more than low enough! Although the ecc82/12au7 does better, it makes no sense to improve distortion peaks at -85dBu... (inaudible)

 

[MidnightArrakis]

Is there a schematic of the B+ supply for this "Hybrid Microphone Preamp"? Or, have I somehow missed it within this thread? I have gone through all seven (7) pages of this thread and I have only seen the schematics for the vacuum-tube heaters and the mic pre-amplifier itself, but nothing for the B+ power-supply. Can you direct me to that, please? THANKS!!!

 

[Sorr]

Surprise! With 150V across the test tube and a current of about 8mA/triode, the third harmonic distortion peak drops below -100dBu. Idem for the 12kHz and 15kHz IMD peaks. Once over that point, the distortion increases.
Dissipation is a very reasonable 1.2W/triode.

When operated at those values. Did you get a chance to measure the gain of the ECC82?

 

[aazaa]

All tests were done with the tube in the preamp, so connected in push-pull. The gain was not determined.

@ MidnightArrakis:
It is not difficult to find a stabilised PSU-circuit online. With some minor adjustments of resistor values any stabilised voltage between 270V and 330V should work fine.
My power supply is not a good example. I had to use voltage doublers to avoid some mechanical noise of the transformers. (posts #101 and 106)

 

[aazaa]

Update

Many have the tubes inside.
I decided to change the variable current source to the classic approach with a pnp-npn combination. It will give better temperature stabilisation (BC560C / BC550C).

Circuits are updated in posts #101 and 110.

Always better to keep the heat out the case, separate PSU and tubes on top.

Paul

 

[aazaa]

Octal - UTC version

All components - resistors, caps, trimmers, transistors, test points - are now on two boards on the bottom plate. You can turn the amp upside down, flip open the bottom plate and have full access to everything. Only wiring inside.

Paul

 

[aazaa]

The circuit

UTC ls12(1:14) and ls51(7:1)
LSK489
RCA 6sl7gt - Sylvania 6sn7gtb

B+ 350V

Not obvious to go for a 1:14 transformer.
Simple, it was pure curiosity to explore the possibilities of the UTC ls12 - LSK489 combination. (A 1:4 input transformer would have been the logical choice.)

In the experiment with ls12 - LSK389, 20kHz was a few dB's down. Now, with LSK489, a Zobel network is needed to correct the +3.5 and +4 dB at 20kHz...

Adjustment of the output level by a 1k log stereo pot.

Adapted amplification in the tube section.
15dB additional NFB by 1k8 resistors at the sources of the LSK489.

Surprising linearity of the ls12(1:14) - ls51(7:1) combination.

The influence of the extra NFB on distortion:

Further improvement of the previously very low noise level.

So, happy...

Paul

 

[aazaa]

gain control

Very interesting schematics! What is the best place for the "gain" potentiometer?

With the 1k8 source resistors there is an additional possibility for gain control.
Placing a resistor across the sources will increase the gain. (+15db with connected sources.)
Gain can be increased and lowered if the 1k8 resistors are replaced by a larger value.

Of course there is still the option to change the 51k resistor value,
or to combine both using a two-circuit rotary switch.

Paul

 

[abbey road d enfer]

gain control

With the 1k8 source resistors there is an additional possibility for gain control.
Placing a resistor across the sources will increase the gain. (+15db with connected sources.)

I would think it results in significant increase in distortion; have you measured that?

Of course there is still the option to change the 51k resistor value,

Have you actually tried that? The FET drains are loaded by the cathodes, which are low impedance (1/Gm). I reckon this resistor should be drastically reduced to noticeably decrease gain.

 

[aazaa]

I would think it results in significant increase in distortion; have you measured that?

Of course distortion increases with a reduced NFB - post 128: "The influence of the extra NFB on distortion".

Measurement +15dB (sources of the LSK489 connected):
THD: 0.0088%
IMD: 0.0073%
Increased distortion yes, but still very low. So, who cares?

With+15dB more gain THD is still slightly better than when the 6SN7 was plugged in the previous version of the preamp (was 0.01%). There is some more NFB in the tube section.

I'm changing my mind about the choice of the input transformer. If you want +15dB gain, a quality 1:10 or even better 1:14 transformer is the best choice!

Have you actually tried that? The FET drains are loaded by the cathodes, which are low impedance (1/Gm). I reckon this resistor should be drastically reduced to noticeably decrease gain.

Replacing the 51k resistor with 13k results in -10dB reduced gain.

Paul

 

[abbey road d enfer]

Replacing the 51k resistor with 13k results in -10dB reduced gain.

That's interesting. It appears like the cathode impedance would be very large. I need to think about it.

 

[aazaa]

I'm afraid calculating 1/Gm has put you on the wrong track.

Both grids of the 6SL7 are connected to the opposite side. The received signal on the grid is in phase with the signal on the cathode of the opposite triode. This affects the impedance between the cathodes.

Paul

 

[abbey road d enfer]

I'm afraid calculating 1/Gm has put you on the wrong track.

Both grids of the 6SL7 are connected to the opposite side. The received signal on the grid is in phase with the signal on the cathode of the opposite triode. This affects the impedance between the cathodes.

Paul

Correct. I neglected the effect of voltage NFB on the grids.

 

[aazaa]

gain control, continued

Conclusion of further tests:
It is best to keep the value of the 51k resistor below 82k. At higher values the preamp starts to lose its excellent linearity.
Better only change the value of this resistor to decrease gain. A lower value gives perfect results.

The gain control of my choice:
A two-circuit rotary switch with 3 steps: -10dB / neutral / +10dB

-10dB: replace the 51k resistor with 13k ( connect 17k5 (18k) in parallel with the 51k resistor )
THD: 0.0014%
neutral: circuit as posted in post 128
THD: 0.0014%
+10dB: 750Ω across the sources of the LSK489
THD: 0.0046%

Same excellent linearity at all positions.

Paul

 

[aazaa]

Thanks

Although the basic concept remained, it took some time to evolve the preamp into the final circuit (post 128).

We can forget all previous versions. I consider them steps necessary to arrive at the final version.
The build with ecc83/82 will be modified to the final circuit with UTC ls10 transformers at the input.

Many thanks for all comments and support!
Special thanks to the first posters: 'Winston', 'emrr' and 'abbey'. With their initial comments the thread came to life.

MaxDM's influence was invaluable. Though unintentionally, he inspired me to correct NFB with the 1M trimpot. Control of H2 distortion and IMD was taken to the next level.

We all miss 'Winston OBoogie'.
He was the very first to reply. It would have been nice to read his analysis of the final circuit.

I really hope he's doing well, wish him all the best.

Paul

 

[emrr]

Thanks for continuing the updates!