Gary H, the trick to reduce noise and microphonics is to lift every heater supply above their respective cathodes, therefore you need two lifters, one for the upper tubes, one for the lower tubes in the schematics.
Try 68K connected to ground + 150 K + 120 K connected to B+, total 338K. This will split your 330 V B+ three ways.
Connect the bottom tube's heater supply (-) terminal to the 68 K resistor, so that the bottom tube heaters are lifted ~68V, therefore positive in relation to the cathodes.
The upper tube cathodes are at~165 V DC, so you need about 220 V on the heaters to play it safe.
Connect the junction of 120 K and 150 K to the upper tube's heaters. They will be lifted to ~220 V, again positive re cathode.
Add a 10 uF cap from ground to each resistor. This will further reduce noise.
I am assuming that the board layout allows this hookup. You may have to cut move/remove jumpers or cut traces to make sure that the upper and lower heater supplies do not share the same (-) trace.
If your board uses the same tube envelope for the upper and lower triode on each gain cell, self-oscillation is difficult to work around.
You can try a 180V lifter: 180K + 150K, but 180 V is dangerously close to the maximum heater to cathode lift in the data sheet. The tubes may not last long.
I hope this helps
Try 68K connected to ground + 150 K + 120 K connected to B+, total 338K. This will split your 330 V B+ three ways.
Connect the bottom tube's heater supply (-) terminal to the 68 K resistor, so that the bottom tube heaters are lifted ~68V, therefore positive in relation to the cathodes.
The upper tube cathodes are at~165 V DC, so you need about 220 V on the heaters to play it safe.
Connect the junction of 120 K and 150 K to the upper tube's heaters. They will be lifted to ~220 V, again positive re cathode.
Add a 10 uF cap from ground to each resistor. This will further reduce noise.
I am assuming that the board layout allows this hookup. You may have to cut move/remove jumpers or cut traces to make sure that the upper and lower heater supplies do not share the same (-) trace.
If your board uses the same tube envelope for the upper and lower triode on each gain cell, self-oscillation is difficult to work around.
You can try a 180V lifter: 180K + 150K, but 180 V is dangerously close to the maximum heater to cathode lift in the data sheet. The tubes may not last long.
I hope this helps
Hi Casouza,
thanks for taking the time to explain the heater bias in such detail. I appreciate it. John Broskie recommends a similar application in his manual and that is the one I used. I included a diagram of it in post #6. What do you think? Does it look correctly implemented? Could the values be tweaked for better performance? Another poster suggested I go right up to the max heater to cathode specs. Bear in mind that I am getting closer to 295 V B+ after rectification. Also, not all tubes generate these microphonics. Right now I have some NOS Sylvanias and New Tung Sols in there. The microphonics are fleeting but those NOS tubes are noisy, nearly 1V rms of hiss coming through the mid and tweeter.
Still, even with that and the scary warm up distortion, I am listening to the finest sounding music I've ever had in my living room.
thanks,
gary
thanks for taking the time to explain the heater bias in such detail. I appreciate it. John Broskie recommends a similar application in his manual and that is the one I used. I included a diagram of it in post #6. What do you think? Does it look correctly implemented? Could the values be tweaked for better performance? Another poster suggested I go right up to the max heater to cathode specs. Bear in mind that I am getting closer to 295 V B+ after rectification. Also, not all tubes generate these microphonics. Right now I have some NOS Sylvanias and New Tung Sols in there. The microphonics are fleeting but those NOS tubes are noisy, nearly 1V rms of hiss coming through the mid and tweeter.
Still, even with that and the scary warm up distortion, I am listening to the finest sounding music I've ever had in my living room.
thanks,
gary
Hi Gary.
Tubes connected in cascode, SRPP or other vertical / series arrangements like the Aikido tend to amplify any small heater or cathode movements; the practical effect is increased microphonics as compared to regular cascade circuits.
The way out is to make the heater supply more positive than the cathode and choose well-made tubes with the internal elements well supported (vibration-free). That's why you have noticed that some tube brandss have less microphonics.
Looking at the schematic posted on #6, all the heaters are lifted to the same DC potential, about 100V.
Since the B+ is about 300V, the upper tube cathodes are at about 150 VDC. Heaters floating at 100 V DC will be positive in relation to the lower tubes cathodes (great) but negative in relation to the upper tubes cathode (bad, it increases microphonics and noise).
That's why I suggessted that you separate the heater PS (-) terminals:
one heater PS feeds the bottom tubes and floats about 100 V above ground;
the second heater power supply feeds the upper tubes and should float above the cathode; 200 V is fine.
So, your voltage divider could a very simple one: ground> 100 K *>100K **> 100K > B+
Connect * to the heater supply feeding the bottom tubes;
Connect ** to the heater supply feeding the upper tubes.
Increase the capacitance size, 0.1uF is good only for radio frequencies.
IME 10-100uF is fine. You will need two capacitors, one for the 100V floater, another one for the 200V floater.
A more simple approach is to float all heaters at 180-200 VDC and pray that no tubes fail, because this is at the limit of the 6SN7 specs.
Good luck
Tubes connected in cascode, SRPP or other vertical / series arrangements like the Aikido tend to amplify any small heater or cathode movements; the practical effect is increased microphonics as compared to regular cascade circuits.
The way out is to make the heater supply more positive than the cathode and choose well-made tubes with the internal elements well supported (vibration-free). That's why you have noticed that some tube brandss have less microphonics.
Looking at the schematic posted on #6, all the heaters are lifted to the same DC potential, about 100V.
Since the B+ is about 300V, the upper tube cathodes are at about 150 VDC. Heaters floating at 100 V DC will be positive in relation to the lower tubes cathodes (great) but negative in relation to the upper tubes cathode (bad, it increases microphonics and noise).
That's why I suggessted that you separate the heater PS (-) terminals:
one heater PS feeds the bottom tubes and floats about 100 V above ground;
the second heater power supply feeds the upper tubes and should float above the cathode; 200 V is fine.
So, your voltage divider could a very simple one: ground> 100 K *>100K **> 100K > B+
Connect * to the heater supply feeding the bottom tubes;
Connect ** to the heater supply feeding the upper tubes.
Increase the capacitance size, 0.1uF is good only for radio frequencies.
IME 10-100uF is fine. You will need two capacitors, one for the 100V floater, another one for the 200V floater.
A more simple approach is to float all heaters at 180-200 VDC and pray that no tubes fail, because this is at the limit of the 6SN7 specs.
Good luck
Hi Carlos,
so I am still a little confused by the designation of upper and lower. I am under the impression that oner tube shares both the upper and lower triodes for the input of a channel and one for the output of the channel. If each tube has one heater for a pair of upper and lower triodes how can it have two different biases?
I'm sure I am missing something basic but looking at the schematic for the Aikido it seems the cathode of one triode feeds the plate of the other triode of the same tube.
If you see where I am going wrong here please give me one more try.
thanks,
gary
so I am still a little confused by the designation of upper and lower. I am under the impression that oner tube shares both the upper and lower triodes for the input of a channel and one for the output of the channel. If each tube has one heater for a pair of upper and lower triodes how can it have two different biases?
I'm sure I am missing something basic but looking at the schematic for the Aikido it seems the cathode of one triode feeds the plate of the other triode of the same tube.
If you see where I am going wrong here please give me one more try.
thanks,
gary
Gary,
Carlos gives good advice regarding the preferred voltage relationships between heater and cathode, i.e., keep the heater more positive than the cathode whenever possible, but within the tube's ratings, of course.
Unfortunately, the Aikido design boards use the same twin triode for both the upper and lower triodes, so you cannot take advantage of the separately referenced heater supplies.
I've used a few types of upper/lower tube arrangements: SRPP, mu follower, beta follower, Alan Kimmel's mu stage, and good old cascode. Never had the sort of problems that you are describing with the NOS tubes you have.
That makes me wonder if something else isn't going on.
Stuart
Carlos gives good advice regarding the preferred voltage relationships between heater and cathode, i.e., keep the heater more positive than the cathode whenever possible, but within the tube's ratings, of course.
Unfortunately, the Aikido design boards use the same twin triode for both the upper and lower triodes, so you cannot take advantage of the separately referenced heater supplies.
I've used a few types of upper/lower tube arrangements: SRPP, mu follower, beta follower, Alan Kimmel's mu stage, and good old cascode. Never had the sort of problems that you are describing with the NOS tubes you have.
That makes me wonder if something else isn't going on.
Stuart
One other thought, Gary. Assuming you haven't tried the excellent Herbie's Damping Instruments, take a roll of good plumber's Teflon tape. Make several wraps about the belt of the tube. Make them tight. This helps with internal microphonics. If it helps you, then the Herbie's are the answer. They worked on my wonderful sounding but very microphonic VT-231 Sylvanias.
I've attached a couple of pictures, one showing the Herbie's Ultrasonic Damping Instruments on the VT-231's and another showing the Teflon tape I used to use.
http://herbiesaudiolab.home.att.net/ultra.htm
TRY IT! You'll be pleased.
Stuart
I've attached a couple of pictures, one showing the Herbie's Ultrasonic Damping Instruments on the VT-231's and another showing the Teflon tape I used to use.
http://herbiesaudiolab.home.att.net/ultra.htm
TRY IT! You'll be pleased.
Stuart
Attachments
thanks Stuart,
I used silicon tubing as damping material wrapped around a tube for a while but it made little difference. I'll try the teflon next time. It seems intuitive that mass is what is needed but also interrupting nodes of oscillation might be a factor too. Come to think of it, wrapping a tube in cumbersome silicon tubing might only make oscillation worse by adding mass.
Someone was telling me recently about a bed that was suspended off the floor by magnets. Lots of little magnets were placed in the mattress and then repelled by a much larger one on the floor and held roughly in place with laniards. This sounds like sort of a magic carpet. I wonder if the same technology could be employed to suspend chassis off shelves etc. I'm sure the magnetic fields wouldn't hurt a thing...
g
I used silicon tubing as damping material wrapped around a tube for a while but it made little difference. I'll try the teflon next time. It seems intuitive that mass is what is needed but also interrupting nodes of oscillation might be a factor too. Come to think of it, wrapping a tube in cumbersome silicon tubing might only make oscillation worse by adding mass.
Someone was telling me recently about a bed that was suspended off the floor by magnets. Lots of little magnets were placed in the mattress and then repelled by a much larger one on the floor and held roughly in place with laniards. This sounds like sort of a magic carpet. I wonder if the same technology could be employed to suspend chassis off shelves etc. I'm sure the magnetic fields wouldn't hurt a thing...
g
- Status
- Not open for further replies.