Tube Bias Control by LM337 or similar IC reg Voltage?

[opacheco]

Hi,

I am thinking in use a LM337 or similar IC reg Voltage for a Push-Pull 6L6 Vacuum Tube Amplifier Bias Control in order to get a stabilized Bias Voltage to the Grid and so a good stable Bias Current; I have some dude about the noise od this kind of IC and I don't know exactly how this will be in the performer of the amplifier??.

Any comments will be appreciate !

Thanks
Opacheco.

 

[opacheco]

Any comments?

Opacheco.

 

[volker]

Any regulator will be magnitudes less noisy than a conventional bias supply. They are almost always only half wave rectification and manufacturers skimp on the filter caps. But ripple on the bias voltage isn't really a big concern in the first place, since it is common mode to the output stage and will be suppressed by the differential operation.

One reason I can see why a regulator is actually a disadvantage is mains voltage fluctuation (if your HT voltage isn't regulated too). If mains drifts, the unregulated bias will drift in the same direction and keep the bias somewhat stable. A regulated bias voltage though will stay constant so the bias point will shift with changing mains.

Personally I don't think there is much to be gained from such effort. But if you're really keen on a "regulated" bias voltage a cap multiplier should be good enough, and it will move along with mains fluctuations.

 

[opacheco]

Any regulator will be magnitudes less noisy than a conventional bias supply. They are almost always only half wave rectification and manufacturers skimp on the filter caps. But ripple on the bias voltage isn't really a big concern in the first place, since it is common mode to the output stage and will be suppressed by the differential operation.

One reason I can see why a regulator is actually a disadvantage is mains voltage fluctuation (if your HT voltage isn't regulated too). If mains drifts, the unregulated bias will drift in the same direction and keep the bias somewhat stable. A regulated bias voltage though will stay constant so the bias point will shift with changing mains.

Personally I don't think there is much to be gained from such effort. But if you're really keen on a "regulated" bias voltage a cap multiplier should be good enough, and it will move along with mains fluctuations.

Dear Volker!

Very nice and interesting analysis; I thinking to evaluate the +B drift with normal mains voltage  fluctuation in order to investigate the maximum and minimum.....but we know that some volts more or some volts minus will not do any effect the Q point of the power tubes (Idle Current, due to the +B variation of course !!).

but I can think that install a 6.3Vdc LM for the filaments too and test with the LM Voltage Bias regulator be a very stabilized amp in performance, what do you think?

Thanks a lot for comments
Opacheco

 

[opacheco]

Any comments?

Opacheco

 

[Tubetec]

I did a test one time with a pre of my own design using a single transformer coupled el84, I had a regulated 6.3v on the heater(317), fed in balanced fashion with the usual 2x100 ohm to ground . I had used conventional Rc for cathode bias . The cathode on the el84 was running fairly close to 6 volts with the ht I had on it , so I grounded one side of the heater ,tied the other end to cathode and hooked up my 6.3 volt regulated supply . It worked but compared to the cathode bias arrangement it had much more noise , I went back to the previous arrangement , maybe if I had a higher voltage heater winding available and used two regs in series the noise could have been reduced . I didnt do any major testing with the funky bias/heater arrangement in place ,but the impression I got with the low impedence between cathode and ground was that treble was smoother ,not nessesarily extended ,as bandwidth was well beyond 100khz even with the rc cathode arrangement. Anyway hope theres some food for thaught in there somewhere

 

[opacheco]

I did a test one time with a pre of my own design using a single transformer coupled el84, I had a regulated 6.3v on the heater(317), fed in balanced fashion with the usual 2x100 ohm to ground . I had used conventional Rc for cathode bias . The cathode on the el84 was running fairly close to 6 volts with the ht I had on it , so I grounded one side of the heater ,tied the other end to cathode and hooked up my 6.3 volt regulated supply . It worked but compared to the cathode bias arrangement it had much more noise , I went back to the previous arrangement , maybe if I had a higher voltage heater winding available and used two regs in series the noise could have been reduced . I didnt do any major testing with the funky bias/heater arrangement in place ,but the impression I got with the low impedence between cathode and ground was that treble was smoother ,not nessesarily extended ,as bandwidth was well beyond 100khz even with the rc cathode arrangement. Anyway hope theres some food for thaught in there somewhere

Tubetec,
Interesting experience!...I will proceed to do a similar 317 installation in the filament in order to have a stabilized 6.3Vdc and I will install another 317 or similar in order to have a stabilized DC BIAS Voltage in the Grids of tubes.....This Bias and Filament stabilization can return a very stable performer state like Volker told us in previous comments.

Thanks
Opacheco

 

[Tubetec]

With that arrangement , any reduction  of the 6.3volts reduced the emission from the cathode while simultaneously increasing the current through the tube , once set up right though seemed rock solid . Main reason I did it was ,no need for an input cap on the grid like you'd need with grid bias, and also two less components ,getting rid the R and C on the cathode.
Ive also done some more recent experiments with using a battery to provide cathode voltage ,the battery actually charges in series with the tube. A small rechargeable cell presents a fraction of an ohm at the cathode to ground so maximises gain and removes the LF timeconstant  like youd get with standard Rc on the cathode

I just love simplicity in electronics ,with the heater cathode coupling method ,you had a grid resistor ,a tube and an output transformer in the signal path ,I dont think you could make an audio  amplifier with any less components.

 

[opacheco]

With that arrangement , any reduction  of the 6.3volts reduced the emission from the cathode while simultaneously increasing the current through the tube , once set up right though seemed rock solid . Main reason I did it was ,no need for an input cap on the grid like you'd need with grid bias, and also two less components ,getting rid the R and C on the cathode.
Ive also done some more recent experiments with using a battery to provide cathode voltage ,the battery actually charges in series with the tube. A small rechargeable cell presents a fraction of an ohm at the cathode to ground so maximises gain and removes the LF timeconstant  like youd get with standard Rc on the cathode

I just love simplicity in electronics ,with the heater cathode coupling method ,you had a grid resistor ,a tube and an output transformer in the signal path ,I dont think you could make an audio  amplifier with any less components.

I don't understsnd so well your idea but look like stabilization of BIAS voltage and Filament by two LM 337 will be a good deal!...Let me to know if I am lost in that.

Thanks
Opacheco