I am breadboarding Thorsten's parallel 6SN7 line stage before committing to a chassis and building it up. When I power it up, it sounds like it is oscillating. I put a 0.1uF across the feedback resistor (circled in red below) and the noise went away. Obviously I don't want to add a cap in the feedback path but does anyone have any tips to keep this node from oscillating?
It does not appear to be a ground loop issue as I have connected an alligator clip to the star ground at the left and touched it to all input, output and tube grounds and no change in the sound. Likewise it is not heaters or wiring since I can touch and move wiring and no change. It is NOT volume dependent and stays the same magnitude regardless of volume setting.
It does not appear to be a ground loop issue as I have connected an alligator clip to the star ground at the left and touched it to all input, output and tube grounds and no change in the sound. Likewise it is not heaters or wiring since I can touch and move wiring and no change. It is NOT volume dependent and stays the same magnitude regardless of volume setting.
Attachments
craigtone, Same part values? Same wiring lead dress? Same grounding techniques? Same load at the output? i.e. a long shielded cable (lots of capacitive load to ground). It could be the circuit is doing Bursted RF oscillations. Acts and sounds like audio frequency, but is actually RF bursts. Do you have a high frequency scope? Try a new old stock 6SN7. Tung-sol did not used to be Russian. And yes, I have used lots of good Russian tubes.
Last edited:
Everything else the same. I have a 3' BJC RCA to RCA running to a Tubelab SE amplifier. I have an old Kenwood 40MHz scope.
Interesting comment about the NOS 6SN7. I tried some old GE's and the first time they were quiet but the sound was very rolled-off compared to the Tung-Sols. I have swapped back and forth a few times now and now the same set of GE's that were quiet are doing the same thing.
Interesting comment about the NOS 6SN7. I tried some old GE's and the first time they were quiet but the sound was very rolled-off compared to the Tung-Sols. I have swapped back and forth a few times now and now the same set of GE's that were quiet are doing the same thing.
craigtone,
Perhaps you moved some wires around. Or perhaps a ground connection is intermittent.
Also the loop gain is dependent on the setting of the 100k potentiometer. And so is the stability of that loop. It may have been on the verge of oscillation and now is beyond that all the time. Once you turn the pot up and down causing the oscillation to start, the oscillation usually swings full scale no matter where the pot is turned.
These are tough to troubleshoot (I am just preaching to the choir).
Having nothing to do with the oscillation: The 4.7uF cap can drive a 3400 Ohm load and only be -1dB at 20 Hz (and 26 degrees phase shift versus at 1 kHz with the 3400 Ohm load). The capacitance seems like it is larger than necessary, since a parallel 6SN7 driving a 3400 Ohm load will be very distorted, and very low gain.
Perhaps you moved some wires around. Or perhaps a ground connection is intermittent.
Also the loop gain is dependent on the setting of the 100k potentiometer. And so is the stability of that loop. It may have been on the verge of oscillation and now is beyond that all the time. Once you turn the pot up and down causing the oscillation to start, the oscillation usually swings full scale no matter where the pot is turned.
These are tough to troubleshoot (I am just preaching to the choir).
Having nothing to do with the oscillation: The 4.7uF cap can drive a 3400 Ohm load and only be -1dB at 20 Hz (and 26 degrees phase shift versus at 1 kHz with the 3400 Ohm load). The capacitance seems like it is larger than necessary, since a parallel 6SN7 driving a 3400 Ohm load will be very distorted, and very low gain.
Last edited:
Hello where are you taking the input signal off?
Maybe the input signal is a little bit noisy, almost all mobile and common output audio outputs develop high frecuency noise because all this things have class D amplifier outputs and they can be poorly decoupled, also the filament supply are you using DC?, some switchers develop lots of noise and end throwing high frequency bursts with some loads.
Maybe the input signal is a little bit noisy, almost all mobile and common output audio outputs develop high frecuency noise because all this things have class D amplifier outputs and they can be poorly decoupled, also the filament supply are you using DC?, some switchers develop lots of noise and end throwing high frequency bursts with some loads.
Input signal is coming direct from my Rega Apollo-R. It is not noisy!
AC heaters. It is not the heaters as it is not a 60Hz or 120Hz hum.
The 22uF caps are 145 Ohms capacitive reactance at 50 Hz. The cathodes are 'grounded' through 145 Ohms capacitive reactance. The heater cathode interface is rated for 100V and 200V. I hope there is not enough (bad tube) leakage from filament to cathode, to cause noise driving into 145 Ohms to ground.
But yes, in spite of the above, it is correct to ground the 6.3V either through the center tap, or through a two resistor pseudo center tap. Some power transformers wind the 6.3V right over the B+ secondary, or over the power line primary. Those high voltages (with noise) can capacitively couple to the filament winding.
But yes, in spite of the above, it is correct to ground the 6.3V either through the center tap, or through a two resistor pseudo center tap. Some power transformers wind the 6.3V right over the B+ secondary, or over the power line primary. Those high voltages (with noise) can capacitively couple to the filament winding.
Last edited:
Removed cathode caps ... still had noise.
Added a 100-ohm x 2 psuedo CT to the heaters ... VIOLA! Still a small amount of hum but low enough that I can now listen to music. The preamp sounds incredible!
Thinking I will go full DC heaters now.
I did also add 4.7K grid stoppers but I read that he omitted them since 6SN7's aren't very prone to oscillation. Since the 100K input is part of the feedback circuit, is there any issue with 2 4.7K grid stoppers on each grid? I'd prefer to build the circuit as posted in the schematic since Thorsten designed it that way for a reason and I rarely see grid stoppers on 6SN7 circuits. The less components the signal travels through the better IMHO. I know resistor noise is very low but I run 102dB horns also!
Added a 100-ohm x 2 psuedo CT to the heaters ... VIOLA! Still a small amount of hum but low enough that I can now listen to music. The preamp sounds incredible!
Thinking I will go full DC heaters now.
I did also add 4.7K grid stoppers but I read that he omitted them since 6SN7's aren't very prone to oscillation. Since the 100K input is part of the feedback circuit, is there any issue with 2 4.7K grid stoppers on each grid? I'd prefer to build the circuit as posted in the schematic since Thorsten designed it that way for a reason and I rarely see grid stoppers on 6SN7 circuits. The less components the signal travels through the better IMHO. I know resistor noise is very low but I run 102dB horns also!
Last edited:
So I just tried DC heaters with GND tied to star ground and noise level did not change vs AC heaters with psuedo center-tap. I then elevated the heater GND (still DC) to 60V and no change in noise level.
I am thinking it is not a heater issue now.
The B+ power supply is pretty lousy ... just a simple 100uF-6K-100uF CRC filter I whipped up for prototyping. I just got my Salas HV shunt regulator in the mail yesterday so will install that next.
I am thinking it is not a heater issue now.
The B+ power supply is pretty lousy ... just a simple 100uF-6K-100uF CRC filter I whipped up for prototyping. I just got my Salas HV shunt regulator in the mail yesterday so will install that next.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.