I've got a push pull 6BQ5 amp that is cathode biased.
The measurements are as follows with a 130.1 ohm cathode resistor.
Plate voltage 340.3Vdc
Screen voltage 341.7Vdc
Cathode voltage 10.27Vdc
Cathode current 78.94mA
I'm thinking of converting it to fixed bias as that eliminates the need for the cathode bypass cap and gives slightly more power output.
All I should need to do is set the grid bias voltage to where the cathode current is around 78.94mA, right?
I have a 500 ohm pot listed on the schematic as that's what I had on hand. I suspect that I will need a second filter stage for the bias voltage.
I'm also thinking of adding a 100mA meter to measure the cathode current so that I do not need an external meter to measure it. I may also install a SPST switch across the meter so that I can bypass the meter if needed.
I am, however going to have to get creative on installing the meter as it will not fit on the chassis due to the chassis being rather compact.
I currently have a 15Vac supply for the - bias power supply. Is it possible to use the 5 volt ands 6.3 volt windings in series for 11.3Vac and get enough bias voltage to bias the tubes at 70mA? If so that would eliminate one transformer.
What I will do is use a variable voltage power supply to set the bias voltage and see what voltage is needed for 70mA cathode current.
Also will the B+ voltages be too high for the 6BQ5 and 6SQ7 or are they ok?
Here's the measurements for the 6SQ7
Plate voltage 299.7V
Cathode voltage 43.2V
Cathode resistor 22.28k
Cathode current 1.94mA
Are the voltages and currents ok?
I could use a 6SR7 for lower gain if so desired.
In the communications receiver that is driving this amp I used a voltage divider at the 1 meg volume control consisting of a series 1 meg resistor and a 1 meg resistor in parallel with the volume control so lower gain from the amp would be beneficial.
It isn't shown on the schematic, but there's a 1 ohm resistor between the CT of the HV winding and ground which I used to measure the total B+ current. I might reduce the wattage of that resistor so that it acts as a fuse should B+ current jump up too high due to a fault or red plating tube.
I do have the Russian equivalent of the 6BQ5 installed.
Also the output transformer was what I had on hand (came from a GEC 912 PLUS amp I built maybe 15 years ago) and I know I could get more output power with a higher primary impedance.
One further question. Keep in mind this is not a HI-FI amp.
There any faults with the design or improvements that can be done to the amp?
The measurements are as follows with a 130.1 ohm cathode resistor.
Plate voltage 340.3Vdc
Screen voltage 341.7Vdc
Cathode voltage 10.27Vdc
Cathode current 78.94mA
I'm thinking of converting it to fixed bias as that eliminates the need for the cathode bypass cap and gives slightly more power output.
All I should need to do is set the grid bias voltage to where the cathode current is around 78.94mA, right?
I have a 500 ohm pot listed on the schematic as that's what I had on hand. I suspect that I will need a second filter stage for the bias voltage.
I'm also thinking of adding a 100mA meter to measure the cathode current so that I do not need an external meter to measure it. I may also install a SPST switch across the meter so that I can bypass the meter if needed.
I am, however going to have to get creative on installing the meter as it will not fit on the chassis due to the chassis being rather compact.
I currently have a 15Vac supply for the - bias power supply. Is it possible to use the 5 volt ands 6.3 volt windings in series for 11.3Vac and get enough bias voltage to bias the tubes at 70mA? If so that would eliminate one transformer.
What I will do is use a variable voltage power supply to set the bias voltage and see what voltage is needed for 70mA cathode current.
Also will the B+ voltages be too high for the 6BQ5 and 6SQ7 or are they ok?
Here's the measurements for the 6SQ7
Plate voltage 299.7V
Cathode voltage 43.2V
Cathode resistor 22.28k
Cathode current 1.94mA
Are the voltages and currents ok?
I could use a 6SR7 for lower gain if so desired.
In the communications receiver that is driving this amp I used a voltage divider at the 1 meg volume control consisting of a series 1 meg resistor and a 1 meg resistor in parallel with the volume control so lower gain from the amp would be beneficial.
It isn't shown on the schematic, but there's a 1 ohm resistor between the CT of the HV winding and ground which I used to measure the total B+ current. I might reduce the wattage of that resistor so that it acts as a fuse should B+ current jump up too high due to a fault or red plating tube.
I do have the Russian equivalent of the 6BQ5 installed.
Also the output transformer was what I had on hand (came from a GEC 912 PLUS amp I built maybe 15 years ago) and I know I could get more output power with a higher primary impedance.
One further question. Keep in mind this is not a HI-FI amp.
There any faults with the design or improvements that can be done to the amp?
If you go fixed bias for almost inaudibly more power, you should have separate bias pots for the output tubes and lower grid leak resistors.
The ds says 300k max for stable fixed operation. So 220k instead of 470k.
The pots don't have to be 500ohm, more like 10k.
But what is wrong with bypass caps in a non hifi amp ?
If you want to improve, use separate cathode resistors and bypass caps.
So 270ohm and 47uF...100uF for each tube.
No nred for a meter and bias checking every other day either, the separate cathode resistors will do the job automatically and safer.
If you think the meter - which is hard to fit anyway - is required for the show, then use a 15V range and on-off-on switch to display bias voltage across cathode resistors from either tube or none.
The ds says 300k max for stable fixed operation. So 220k instead of 470k.
The pots don't have to be 500ohm, more like 10k.
But what is wrong with bypass caps in a non hifi amp ?
If you want to improve, use separate cathode resistors and bypass caps.
So 270ohm and 47uF...100uF for each tube.
No nred for a meter and bias checking every other day either, the separate cathode resistors will do the job automatically and safer.
If you think the meter - which is hard to fit anyway - is required for the show, then use a 15V range and on-off-on switch to display bias voltage across cathode resistors from either tube or none.
Separate pots would be beneficial as I wouldn't need a matched pair of tubes.
The 500 ohm pot was what I had on hand. May see if I can find a higher value in my stash.
Definitely need lower grid leak resistors for sure.
The 500 ohm pot was what I had on hand. May see if I can find a higher value in my stash.
Definitely need lower grid leak resistors for sure.
IMHO these “benefits” are not sufficient reasons to go into all the trouble of converting to fixed bias. Especially on an amplifier with no voltage amplification (VA) stage before the Cathodyne phase inverter, which provides a gain of less than 1). Perhaps it would be worthwhile rethinking the amp design and consider rebuilding a design where you have an integrated VA stage, such as the Magnavox modifications on www.audiokarma.org
If you are set on fixed bias, one way of converting to fixed bias would be Dave Gillespie’s EFB modification. You can read all about it on Dave’s web page, where he describes the conversion of a cathode-biased Dynaco SCA35 to the EFB biased output and discuss the improve,ents in performance to expect. See: http://www.tronola.com/A_New_Look_At_An_Old_Friend.pdf
His store offers the PCBs for the mod. http://www.tronola.com/html/daves_store.html I suggest for consideration the Dynaco ST35 boards (PC-13 with SCA35 power supply board)
However, this PCB does not enable individual bias.
Last edited:
The voltage amplification stage is a 12SQ7 in the communications receiver.
I've never thought of using a voltage regilator in the cathode circuit.
I've never thought of using a voltage regilator in the cathode circuit.
Last edited:
Plate and screen voltage (measured to cathode) limit for 6BQ5 is 300V.
Fixed bias will increase actual voltages by another 10V.
Cathode bias is much safer regarding thermal runaway.
Fixed bias will increase actual voltages by another 10V.
Cathode bias is much safer regarding thermal runaway.
The advantage is that you don’t to deal with the effective higher B+.
The 6BQ5 is running hot in the original schematic, 330V*39.5mA=13W.
Better to keep cathode bias and separate cathode resistors and caps, 2x300ohms would be safer for the 6BQ5.
The first stage has no gain, you need like 10V input signal to reach full power, right?
Better to keep cathode bias and separate cathode resistors and caps, 2x300ohms would be safer for the 6BQ5.
The first stage has no gain, you need like 10V input signal to reach full power, right?
The gain is a 12SQ7 on the communications receiver chassis.
Gonna be dropping the B+ voltage some. I first tried eliminating the first cap making it a choke input power supply, but that dropped B+ too much. The next thing I'll try is to lower the value of the first filter cap to maybe under 10uF. That will lower the B+ voltage. I'll shoot for 280-290Vdc with the amp cathode biased.
Gonna be dropping the B+ voltage some. I first tried eliminating the first cap making it a choke input power supply, but that dropped B+ too much. The next thing I'll try is to lower the value of the first filter cap to maybe under 10uF. That will lower the B+ voltage. I'll shoot for 280-290Vdc with the amp cathode biased.
The B+ was about 300Vdc to the 6BQ5 plates when I had another load connected at the 1.2k resistor.
Another way I've seen - bias done is to install a resistor between the circuit ground and the center tap of the high voltage winding.
Another way I've seen - bias done is to install a resistor between the circuit ground and the center tap of the high voltage winding.
So after doing some experimenting this is what I've come up with.
The input jack is wired to short the input when nothing is plugged in.
The measurements are as follows.
6BQ5 Plate 318.8Vdc
6BQ5 Screen 321Vdc
6BQ5 cathode current 56.4mA
Bias voltage -11.29Vdc
Unfortunately I cannot drop the B+ voltage more with the parts I have on hand.
Given the bias is not adjustable by a pot, I will not need the meter. What I may do is install a couple banana jacks on the amp connected across the 1 ohm resistor so that I can use a DMM to measure the cathode current.
Since it's relatively easy to do I might try the ultra linear connection real quick just to see if that's any better.
I have a tuner connected to my McIntosh C-24 preamp and it drives the amp quite nicely. Even though the amp isn't intended for HI-FI use, it sounds pretty good.
According to the tube data I should be using an 8k primary impedance output transformer for the operating point I am close to.
This one will give me 4 8 and 16 ohm output impedances.
https://edcorusa.com/products/cxpp25-8k-23-25w-8k-ohms-push-pull-tube-output-transformer
Not sure if I will use it or just leave well enough alone.
The input jack is wired to short the input when nothing is plugged in.
The measurements are as follows.
6BQ5 Plate 318.8Vdc
6BQ5 Screen 321Vdc
6BQ5 cathode current 56.4mA
Bias voltage -11.29Vdc
Unfortunately I cannot drop the B+ voltage more with the parts I have on hand.
Given the bias is not adjustable by a pot, I will not need the meter. What I may do is install a couple banana jacks on the amp connected across the 1 ohm resistor so that I can use a DMM to measure the cathode current.
Since it's relatively easy to do I might try the ultra linear connection real quick just to see if that's any better.
I have a tuner connected to my McIntosh C-24 preamp and it drives the amp quite nicely. Even though the amp isn't intended for HI-FI use, it sounds pretty good.
According to the tube data I should be using an 8k primary impedance output transformer for the operating point I am close to.
This one will give me 4 8 and 16 ohm output impedances.
https://edcorusa.com/products/cxpp25-8k-23-25w-8k-ohms-push-pull-tube-output-transformer
Not sure if I will use it or just leave well enough alone.
If the quoted cathode current is for one tube, then you are well above the dissipation limit for the 6BQ5, 18W vs 14W (Pa+Pg2).
Yup, I looked at the curves after posting and realised it had to be for two tubes. Should have checked first before posting.
When I have time I'll connect up a 4 ohm non-inductive load to the 4 ohm output then feed the amp a sinewave and see how clean the sinewave is. I can also get the approximate output power by increasing the signal until right before the sinewave begins to distort as seen on a scope and then using ohms law.
I might see if there's room on the amp chassis for an IEC jack as that would be the best way to connect the amp to power.
Had a slight buzz in the speaker until I connected the 220 ohm resistors.
I might see if there's room on the amp chassis for an IEC jack as that would be the best way to connect the amp to power.
Had a slight buzz in the speaker until I connected the 220 ohm resistors.
You can also use a sound card and REW software to measure distortion levels and even use it as a signal generator.
Could we go back to your motivation for this tinkering? As well as the schematic in post #13? And lowering the B+? I think I have missed the point of all this effort.
Why exactly do you want to eliminate the cathode capacitor? In your schematic in #13 you have not gone to fixed bias but removed the bypass capacitor, if I understand the schematic correctly. The bypass capacitor serves a good purpose in a cathode biased PP pair. What value of cathode resistor do you now use? If you are concerned about uneven dissipation in unmatched tubes, have you considered going to individual cathode resistors (including a trimmer pot to vary the bias) and each with a bypass cap.
And why lower the B+? Is it to reduce dissipation in the power tubes? I think your B+ was fine, as long as you use the correct cathode resistors for proper bias.
What I did was create a negative bias supply in a way using the 196 ohm resistor given it goes between the center tap of the HV winding and ground.
The cathode resistor of the 6BQ5 is a 1 ohm resistor so I can measure the cathode current.
I can ditch the 3k resistor, however that raises the 6BQ5 plate voltage to about 350Vdc.
What I could do is install a 5Y3 or other similar 2 amp filament rectifier and eliminate the 3K resistor.
I'll likely then have about 300Vdc.
Not sure why I didn't think of that last night.
The cathode resistor of the 6BQ5 is a 1 ohm resistor so I can measure the cathode current.
I can ditch the 3k resistor, however that raises the 6BQ5 plate voltage to about 350Vdc.
What I could do is install a 5Y3 or other similar 2 amp filament rectifier and eliminate the 3K resistor.
I'll likely then have about 300Vdc.
Not sure why I didn't think of that last night.