Push-pull microphone preamp with UTC LS and Tamura transformers
- Thread starter aazaa
- Start date
Help Support GroupDIY Audio Forum:
ruffrecords
Well-known member
Interesting. The tube I normally use is the Electroharmonix 6922EH and its data sheet says says Vk plus or minus 200V.aazaa said:
Looks like a 6922 is not always a 6922 and is not equal to a E88CC.
Cheers
Ian
At the start it was not my intention to post the heater psu.
But... in the official G9 help thread (also a circuit with ECC82) I see people struggling over and over with overheated regulators, and that since 2006.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Conversation between andow and giraf on feb 25-27 2018:
andow: The heatsink already is pretty big, but maybe it has to be even bigger...
Is this some kind of behaviour one would expect when the heater voltage regulator is getting too hot?
???
I think I will also get a bigger heatsink anyway - just to be sure...
giraf: If heat sink is not over 80degrees C, you should rather look at thermal coupling between regulator and sink. Heat transfer paste and/or silicone-type isolation..
andow: I already have silicone isolation between the regulator and the heat sink and heat transfer paste between the heat sink and the case. BUT the heat sink is inside the case, which doesn't have too much ventilation....
So I guess I will try to mount the regulator directly to the case and then the heat sink on the outside of the case.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Maybe it's time to redraw the PSU. Rather than going for a 4x bigger heatsink and heat transfer paste, there is a possibility to produce almost no heat.
So here it is:
Revision of the PSU when the extra transformer came in: addition of 12V zeners.
In the first version without zeners, voltages across the filaments were between 11.6V and 12.4V. With the combination of U1, U2 and 12V zeners, voltages are between 12V and 12.2V. No traces of AC visible on scope.
Although the zeners draw only very low current they are 5W. In experiments with one channel there is a possibility to leave out the 2 tubes of the other channel.
U2 is a current limiter, providing a soft start. One can choose the smallest available heatsink.
(No heatsink for U2 in my preamp.)
Current is 1.25/7.5=0.166666A or 166mA
Tension over U1 is 63-56= 7V. Dissipation: 7x0.166666=1.16W
Very easy to cool with a normal heatsink...
It is possible to reduce the residual noise by adding caps. Unnecessary as it doesn't improve the performance of the tubes.
Though I never saw a heater PSU designed this way... my oscilloscope seems to like it.
And no, I did not forget to connect it!
Paul
But... in the official G9 help thread (also a circuit with ECC82) I see people struggling over and over with overheated regulators, and that since 2006.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Conversation between andow and giraf on feb 25-27 2018:
andow: The heatsink already is pretty big, but maybe it has to be even bigger...
Is this some kind of behaviour one would expect when the heater voltage regulator is getting too hot?
???
I think I will also get a bigger heatsink anyway - just to be sure...
giraf: If heat sink is not over 80degrees C, you should rather look at thermal coupling between regulator and sink. Heat transfer paste and/or silicone-type isolation..
andow: I already have silicone isolation between the regulator and the heat sink and heat transfer paste between the heat sink and the case. BUT the heat sink is inside the case, which doesn't have too much ventilation....
So I guess I will try to mount the regulator directly to the case and then the heat sink on the outside of the case.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Maybe it's time to redraw the PSU. Rather than going for a 4x bigger heatsink and heat transfer paste, there is a possibility to produce almost no heat.
So here it is:
Revision of the PSU when the extra transformer came in: addition of 12V zeners.
In the first version without zeners, voltages across the filaments were between 11.6V and 12.4V. With the combination of U1, U2 and 12V zeners, voltages are between 12V and 12.2V. No traces of AC visible on scope.
Although the zeners draw only very low current they are 5W. In experiments with one channel there is a possibility to leave out the 2 tubes of the other channel.
U2 is a current limiter, providing a soft start. One can choose the smallest available heatsink.
(No heatsink for U2 in my preamp.)
Current is 1.25/7.5=0.166666A or 166mA
Tension over U1 is 63-56= 7V. Dissipation: 7x0.166666=1.16W
Very easy to cool with a normal heatsink...
It is possible to reduce the residual noise by adding caps. Unnecessary as it doesn't improve the performance of the tubes.
Though I never saw a heater PSU designed this way... my oscilloscope seems to like it.
And no, I did not forget to connect it!
Paul
Last edited:
I like your setup. I did a similar thing for a tube microphone power supply recently. A linear voltage regulator followed by a currenct source, to achieve lowest possible noise since I'm using a directly heated tube. Advantage is there is no high in-rush current at startup when the filament is cold. Also no adjustment necessary to compensate for the loss through the cable.
You're right about the qualities of a current source following a linear voltage regulator, Volker. Have done that before. It is the first time this current source is feeding a string of 5 zeners.
Finally the extra 30VA transformer on top of the PSU got its cover. I painted a Ø 8 cm Rösle jar to do the job.
https://www.roesle.com/en/kitchen/c...2/jar-with-silicone-glass-lid-oe-8-cm-3.2-in.
Paul
Finally the extra 30VA transformer on top of the PSU got its cover. I painted a Ø 8 cm Rösle jar to do the job.
https://www.roesle.com/en/kitchen/c...2/jar-with-silicone-glass-lid-oe-8-cm-3.2-in.
Paul
Last edited:
I like the push pull long tail concept but I wonder if using 12AY7's (6072's) or 6DJ8 (6922) is more suitable, especially for "long tail".... The heater supplies need to be floated in order to stay within heater/cathode voltage specs...I never liked 12AX7's....Altec 1567A mixers, some Ampex equipment, etc. used 12AX7's but most 12AX7's I had were noisy and/or microphonic..and the 7025 wasn't much better....
Both rmburrow and ruffrecords (Ian) seem to prefer the 6922. As said before, it is possible to use other tubes.
If you don't have this tubes at home I would suggest to buy the Electroharmonix 6922EH as advised by Ian. (max. voltage between heater and cathode = +-200V, see post #62). The values differ depending on the brand. You can see on the data sheet of the manufacturer if elevation of the heaters is necessary.
Below the heater PSU adapted to 4x 6922 tubes. A 24V transformer is resulting in 24x√2=34VDC after rectification.
Current is 1.25/3.9=0.320A=320mA.
The heat is distributed over U1 and U2.
Without the pilot lamp the dissipation of U1 is 5x0.320=1.6W. With the pilot lamp connected it is even less. Dissipation of U2 is (5-1.25)x0.320=1.2W.
Both U1 and U2 can be cooled easily with a normal heatsink.
Paul
If you don't have this tubes at home I would suggest to buy the Electroharmonix 6922EH as advised by Ian. (max. voltage between heater and cathode = +-200V, see post #62). The values differ depending on the brand. You can see on the data sheet of the manufacturer if elevation of the heaters is necessary.
Below the heater PSU adapted to 4x 6922 tubes. A 24V transformer is resulting in 24x√2=34VDC after rectification.
Current is 1.25/3.9=0.320A=320mA.
The heat is distributed over U1 and U2.
Without the pilot lamp the dissipation of U1 is 5x0.320=1.6W. With the pilot lamp connected it is even less. Dissipation of U2 is (5-1.25)x0.320=1.2W.
Both U1 and U2 can be cooled easily with a normal heatsink.
Paul
Last edited:
@ rmburrow
I ordered two new EH 12AY7 tubes in an attempt to improve my preamp.
Same output level. (There is enough NFB to cancel the differences in gain between the ECC83 and the 12AY7.) The output of the RME ADI-96 PRO AD converter can be read very precisely on computer.
To my surprise, no improvement in noise. Same output level, same noise level. For now I keep the old Philips ECC83. I have a lot of them, some nos some used, but nearly all high quality.
With the Neumann M49 (6S6B-V tube) or the Shure SM81-LC connected, the level of noise is determined by the microphone noise, not by the noise in the preamp. (Remember, there is a high gain of the UTC LS12 transformer at the input of the preamp.)
And why not replace the ECC83 by two (Philips) EF86's wired as triodes?
Paul
Last edited:
Hybrid version with UTC LS-30 and LSK389.
Experiment in the left channel with a low noise monolithic matched dual JFET.
The UTC LS-12 (1:14) transformer is replaced by the UTC LS-30.
Happy with the result!
advantages:
- a 1:1 or 1:2 transformer can be used at the input
- very low noise when the Sennheiser MD421 is recording a low volume source
- the input Z is not reduced by the NFB
- no 'Miller' roll-off
- frequencies flat 10Hz - 40kHz at the variable output
drawbacks:
- alas, no ability to hear frequencies of 40kHz
- my microphone cannot record these frequencies
- no need to record frequencies below 25Hz
Paul
Edit: For lower noise a 1:4 transformer at the input and the LSK489 as monolithic matched dual J-fet is probably the best choice. See post #86.
Experiment in the left channel with a low noise monolithic matched dual JFET.
The UTC LS-12 (1:14) transformer is replaced by the UTC LS-30.
Happy with the result!
advantages:
- a 1:1 or 1:2 transformer can be used at the input
- very low noise when the Sennheiser MD421 is recording a low volume source
- the input Z is not reduced by the NFB
- no 'Miller' roll-off
- frequencies flat 10Hz - 40kHz at the variable output
drawbacks:
- alas, no ability to hear frequencies of 40kHz
- my microphone cannot record these frequencies
- no need to record frequencies below 25Hz
Paul
Edit: For lower noise a 1:4 transformer at the input and the LSK489 as monolithic matched dual J-fet is probably the best choice. See post #86.
Last edited:
Winston OBoogie
Well-known member
J-Fetcode! Nice 
Michael Tibes
Well-known member
I' m obviously late to the party, but this one came to mind:
The output triodes are some of the best ever made IMO.
What do you mean by "correcting itself"?
Winston OBoogie
Well-known member
If I were to hazard a guess, I'd say the correcting that needs to be done
is to someone's reading glasses.
The cathodes of the output valve share a 330R resistor.
The 680K resistors are grid leaks.
is to someone's reading glasses.
The cathodes of the output valve share a 330R resistor.
The 680K resistors are grid leaks.
I was referring to the design of the original post, not the DESA one.
Having a 16.5K cathode resistor, not bypassed, is going to raise the total resistance of the output tubes, in relation to the transformer.
The 16.5K resistor will compensate this through negative feedback, which is a form of correction (that is what I meant), and the amount of correction needed because of the high resistance driving the reactive load, is probably more than it should be, without affecting the sound, even though THD may be low enough.
Also, such a high value for a common resistor between cathodes, appears to be doing good things for reducing distortion in a push-pull class A circuit, but since the tubes used are usually not matched perfectly, and don't have perfectly equal curves, the phase relationship between the tubes, will also be mismatched to some degree, and instead of cancelling out harmonics, you get positive feedback of odd harmonics, between the tubes.
There's a paper about this, relative to power tube output, but I think that to some degree it applies to push-pull tube class-A in general, especially when the tubes are loaded heavily.
The argument being that there should be no interaction between the two output tubes, via the cathode resistor, because of the cross-modulation between the tubes, and that it generally increases Intermodulation Distortion, while it can lower THD. So the idea is to bypass the cathode resistor and just use more local feedback on either side, to reduce distortion.
Having a 16.5K cathode resistor, not bypassed, is going to raise the total resistance of the output tubes, in relation to the transformer.
The 16.5K resistor will compensate this through negative feedback, which is a form of correction (that is what I meant), and the amount of correction needed because of the high resistance driving the reactive load, is probably more than it should be, without affecting the sound, even though THD may be low enough.
Also, such a high value for a common resistor between cathodes, appears to be doing good things for reducing distortion in a push-pull class A circuit, but since the tubes used are usually not matched perfectly, and don't have perfectly equal curves, the phase relationship between the tubes, will also be mismatched to some degree, and instead of cancelling out harmonics, you get positive feedback of odd harmonics, between the tubes.
There's a paper about this, relative to power tube output, but I think that to some degree it applies to push-pull tube class-A in general, especially when the tubes are loaded heavily.
The argument being that there should be no interaction between the two output tubes, via the cathode resistor, because of the cross-modulation between the tubes, and that it generally increases Intermodulation Distortion, while it can lower THD. So the idea is to bypass the cathode resistor and just use more local feedback on either side, to reduce distortion.
Last edited:
Indeed with such a high value for the common Rk, the output stage has morphed into a LTP.
Indeed, I mentioned this already in post #8:
"Moreover I like the idea of using opposite triodes correcting each other when wired in a long tail circuit. Idem for the 'crossed feedback' connecting the opposite sides."
In my opinion this is one of the qualities of the design...
Interesting paper!
Conclusion of the paper: " In a class A amplifier the use of a bypass capacitor generally reduces the intermodulation distortion, although it may either increase or decrease the harmonic distortion.
The decision as to whether or not to use such a capacitor depends mainly on the magnitude of the distortion. If it is very small, then it may be safely left off, with no possible audible difference."
The 2 charts (5kHz and 10kHz) of fig.9 show that when the output is low, - 6V with a charge of 15 ohm - the intermodulation distortion without bypass capacitor is lower in both cases. A low output of 2.4 Watt has lower intermodulation distortion without cap. Most of the music information played in a normal living room with reasonable high rendement speakers is within the first couple of watts.
In the microphone preamp the maximum output level used to record is low. Average level very low.
Not only the unbypassed common cathode resistor is playing a correcting role. The interplay with the cross-coupled NFB is important. This NFB is also influencing the behavior of the double j-fet. The voltage on the drain on both sides is constantly changing with the signal.
Whatever, the output connected to the RME AD converter is measuring flat on scope. I'm pleased with the result when recording ...
And... of course it is easy to bypass the 16k5 resistor or to connect 2 separate 33k resistors on the cathodes and bypass both.
Paul
Last edited:
