Victoria Guy,
1. Battery Bias, triode wired:
Consider a Triode wired 6V6 with 250V B+, and a 5k Ohm output transformer that has 200 Ohms DCR.
Suppose the cathode current is 45mA, with the 12V battery bias.
The drop of B+ in the OPT is 200 Ohms x 0.045A = 9V. We now have 250V - 9V = 241V at the plate (241V plate to cathode).
2. Self Bias, triode wired:
Now, lets remove the battery bias, and replace the Cathode 10 Ohm current sense resistor with a 267 Ohm self bias resistor (45mA @ 12V).
Now we have 241V - 12 V = 229V plate to cathode.
Suppose the 6V6 Triode wired plate resistance, rp is 2,000 Ohms.
Self Bias reduced the B+ by 12V. 12V / 2,000 Ohms = 6 mA less plate current.
45mA - 6mA = 39mA
We would need to raise the B+ by 12V to get back to 45mA.
3. Battery Bias versus Self Bias, Pentode mode:
The plate to cathode voltage is 241V with battery bias.
The plate to cathode voltage is 229V with 12V self bias.
The Pentode plate resistance is 50,000 Ohms
12V / 50,000 Ohms = 0.24mA a very small change in plate current (45mA versus 44.76mA).
There is essentially no change in plate current from battery to self bias.
But pentode operation means 50,000 plate resistance is trying to drive 5,000 Ohm output transformer. The damping factor is very low. You would need negative feedback for a better damping factor, and for a better frequency response.
4. Battery Bias versus Self Bias, Ultra Linear Mode
The plate resistance in Ultra Linear is about 2.5 times the plate resistance in Triode wired.
2.5 x 2,000 = 5,000 Ohms.
12V / 5,000 Ohms = 2.4mA 45mA - 2.5mA = 42.5mA
There is hardly any change in plate current with battery bias versus self bias in Ultra Linear mode.
But Ultra Linear has better damping factor than (un-fedback) pentode mode.
5,000 Ohms rp and 5,000 Ohms plate load . . . damping factor is 1 (at least it is unity, not 0.1 of un-fedback pentode mode).
Triode mode, and Ultra Linear modes that do not have negative feedback are stable.
(Well, admittedly: triode mode with the screen tied to the plate is a Very Local negative feedback, and Ultra Linear with the screen tied to the UL tap is a Very Local negative feedback).
Pentode mode that needs to have negative feedback, may, or may Not, be stable with negative feedback.
And adjusting negative feedback can be tricky, especially if it is from the OPT secondary, and the OPT transformer selection and phase frequency varies widely.
Batteries:
When I used to use battery bias, I always used Duracell. At one of my jobs, I built a battery tester that used some active circuitry and an oscilloscope for a monitor. I tested open circuit voltage, and dynamic impedance (internal resistance).
I tested Duracell, Everyday, and Ray-o-Vac.
Duracell won, and that is what we used for medical equipment.
Another of my associates that worked for another company, came to the same conclusion.
"Your mileage may Vary".
Which kind of bias do I prefer to use:
I consider self bias to be the best combination of advantages versus disadvantages, whether for my own amps (I can test and watch over things), and for those others who do not have the ability or even want to test current over time, etc.
Don't forget to use a bypass cap across the self bias resistor (depending on the circuit, sometimes does not need one, and those purists do not seem to like electrolytics).
1. Battery Bias, triode wired:
Consider a Triode wired 6V6 with 250V B+, and a 5k Ohm output transformer that has 200 Ohms DCR.
Suppose the cathode current is 45mA, with the 12V battery bias.
The drop of B+ in the OPT is 200 Ohms x 0.045A = 9V. We now have 250V - 9V = 241V at the plate (241V plate to cathode).
2. Self Bias, triode wired:
Now, lets remove the battery bias, and replace the Cathode 10 Ohm current sense resistor with a 267 Ohm self bias resistor (45mA @ 12V).
Now we have 241V - 12 V = 229V plate to cathode.
Suppose the 6V6 Triode wired plate resistance, rp is 2,000 Ohms.
Self Bias reduced the B+ by 12V. 12V / 2,000 Ohms = 6 mA less plate current.
45mA - 6mA = 39mA
We would need to raise the B+ by 12V to get back to 45mA.
3. Battery Bias versus Self Bias, Pentode mode:
The plate to cathode voltage is 241V with battery bias.
The plate to cathode voltage is 229V with 12V self bias.
The Pentode plate resistance is 50,000 Ohms
12V / 50,000 Ohms = 0.24mA a very small change in plate current (45mA versus 44.76mA).
There is essentially no change in plate current from battery to self bias.
But pentode operation means 50,000 plate resistance is trying to drive 5,000 Ohm output transformer. The damping factor is very low. You would need negative feedback for a better damping factor, and for a better frequency response.
4. Battery Bias versus Self Bias, Ultra Linear Mode
The plate resistance in Ultra Linear is about 2.5 times the plate resistance in Triode wired.
2.5 x 2,000 = 5,000 Ohms.
12V / 5,000 Ohms = 2.4mA 45mA - 2.5mA = 42.5mA
There is hardly any change in plate current with battery bias versus self bias in Ultra Linear mode.
But Ultra Linear has better damping factor than (un-fedback) pentode mode.
5,000 Ohms rp and 5,000 Ohms plate load . . . damping factor is 1 (at least it is unity, not 0.1 of un-fedback pentode mode).
Triode mode, and Ultra Linear modes that do not have negative feedback are stable.
(Well, admittedly: triode mode with the screen tied to the plate is a Very Local negative feedback, and Ultra Linear with the screen tied to the UL tap is a Very Local negative feedback).
Pentode mode that needs to have negative feedback, may, or may Not, be stable with negative feedback.
And adjusting negative feedback can be tricky, especially if it is from the OPT secondary, and the OPT transformer selection and phase frequency varies widely.
Batteries:
When I used to use battery bias, I always used Duracell. At one of my jobs, I built a battery tester that used some active circuitry and an oscilloscope for a monitor. I tested open circuit voltage, and dynamic impedance (internal resistance).
I tested Duracell, Everyday, and Ray-o-Vac.
Duracell won, and that is what we used for medical equipment.
Another of my associates that worked for another company, came to the same conclusion.
"Your mileage may Vary".
Which kind of bias do I prefer to use:
I consider self bias to be the best combination of advantages versus disadvantages, whether for my own amps (I can test and watch over things), and for those others who do not have the ability or even want to test current over time, etc.
Don't forget to use a bypass cap across the self bias resistor (depending on the circuit, sometimes does not need one, and those purists do not seem to like electrolytics).
Last edited:
Everybody,
Do always use B+ Bleeder Resistors, yes, yes, yes . . . Always, Forever, Eternally, Amen.
Safety First!
If it is your own amplifier, then for whatever reason you pass on, someone else will love to have your amplifier.
When they remove the cover, and look inside to see your skillful handiwork, don't let them be shocked by the still charged electrolytic.
(They just listened to your amp, were curious what was inside, unplugged it, and removed the cover).
Do always use B+ Bleeder Resistors, yes, yes, yes . . . Always, Forever, Eternally, Amen.
Safety First!
If it is your own amplifier, then for whatever reason you pass on, someone else will love to have your amplifier.
When they remove the cover, and look inside to see your skillful handiwork, don't let them be shocked by the still charged electrolytic.
(They just listened to your amp, were curious what was inside, unplugged it, and removed the cover).
Last edited:
6A3-
Thanks for that long explanation - I appreciate it.
I don't want to pull Eli's thread off-topic, but...
I thought plate resistance (internal plate resistance) was a feature of the tube construction. I don't understand how the connection to the tube can make that vary by a factor of 10 (50k in the tube specs vs 5k).
..which shows how much I still have to learn!
Thanks for that long explanation - I appreciate it.
I don't want to pull Eli's thread off-topic, but...
I thought plate resistance (internal plate resistance) was a feature of the tube construction. I don't understand how the connection to the tube can make that vary by a factor of 10 (50k in the tube specs vs 5k).
..which shows how much I still have to learn!
VictoriaGuy,
Back to Eli's post #4, the Decware link schematic [hard to see link there: "here"] is in Triode Mode (screen tied through a stopper resistor to the plate). You could use a 6AQ5, 6V6, 6BQ5, or EL84 instead of the SV83 with either some resistor changes, or B+ voltage adjustment, or both.
Yes, I love this thread too.
Negative Feedback effect on plate impedance:
Negative Feedback reduces the gain of the amplifier.
The intrinsic 6V6 plate resistance in pentode mode is 50k at 250V and 45mA.
If you apply negative feedback, the amplitude of the feedback signal is proportional to the amplitude of the plate voltage swing.
Put a heavy load on the amplifier output, then the plate voltage swing is low amplitude. So, the negative feedback signal is low (less feedback signal). With less signal feedback, the gain goes up . . . correcting for the low amplifier plate voltage swing.
Put a light load on the amplifier output, then the plate voltage swing is high amplitude. So, the negative feedback signal is high (more feedback signal). With more signal feedback, the gain goes down . . . correcting for the high amplifier plate voltage swing.
In effect, negative feedback reduces the plate impedance versus non-feedback mode.
Connecting the screen to the plate is negative feedback (1:1 triode mode). The plate moves "1", the screen moves "1". This is as low as the plate impedance can get, without some additional negative feedback path around the circuit. In triode mode, the effective gain of the output tube is the lowest it can be.
Connecting the screen to the Ultra Linear tap is negative feedback (i.e. 1 : 0.4). The plate moves "1", the screen moves 0.4 (40%). This reduces the effective plate impedance, but not as much as in triode mode. In ultra linear mode, the effective gain of the output tube is larger than in triode mode.
Pentode mode is back to having 50k plate resistance, and can only be reduced by some feedback around the output stage.
Back to Eli's post #4, the Decware link schematic [hard to see link there: "here"] is in Triode Mode (screen tied through a stopper resistor to the plate). You could use a 6AQ5, 6V6, 6BQ5, or EL84 instead of the SV83 with either some resistor changes, or B+ voltage adjustment, or both.
Yes, I love this thread too.
Negative Feedback effect on plate impedance:
Negative Feedback reduces the gain of the amplifier.
The intrinsic 6V6 plate resistance in pentode mode is 50k at 250V and 45mA.
If you apply negative feedback, the amplitude of the feedback signal is proportional to the amplitude of the plate voltage swing.
Put a heavy load on the amplifier output, then the plate voltage swing is low amplitude. So, the negative feedback signal is low (less feedback signal). With less signal feedback, the gain goes up . . . correcting for the low amplifier plate voltage swing.
Put a light load on the amplifier output, then the plate voltage swing is high amplitude. So, the negative feedback signal is high (more feedback signal). With more signal feedback, the gain goes down . . . correcting for the high amplifier plate voltage swing.
In effect, negative feedback reduces the plate impedance versus non-feedback mode.
Connecting the screen to the plate is negative feedback (1:1 triode mode). The plate moves "1", the screen moves "1". This is as low as the plate impedance can get, without some additional negative feedback path around the circuit. In triode mode, the effective gain of the output tube is the lowest it can be.
Connecting the screen to the Ultra Linear tap is negative feedback (i.e. 1 : 0.4). The plate moves "1", the screen moves 0.4 (40%). This reduces the effective plate impedance, but not as much as in triode mode. In ultra linear mode, the effective gain of the output tube is larger than in triode mode.
Pentode mode is back to having 50k plate resistance, and can only be reduced by some feedback around the output stage.
Last edited:
I do "hen scratch" schematics and, at this time, I don't have a functioning scanner. 
Somebody else will have to look at links, digest narratives, and post something that will be discussed, modified, and settled on.
I'll toss a possibility out to be "kicked around".
IXCP10M45S CCS load the 12AV7 voltage amplifier at an IB of 6 mA. An adequate net anode load will be present, even if a "fixed" bias "final" is employed. Notice the comparatively low 100 Kohm grid to ground limit for the 6V6, when "fixed" bias is selected. I'm not suggesting "fixed" bias, but am allowing for it.
A 500 Ω cathode bias resistor ensures adequate compliance for the 2 VRMS signal from a "standard" CDP. IMO, there's a good chance that bypassing the cathode bias resistor will be unnecessary.
Somebody else will have to look at links, digest narratives, and post something that will be discussed, modified, and settled on.I'll toss a possibility out to be "kicked around".
IXCP10M45S CCS load the 12AV7 voltage amplifier at an IB of 6 mA. An adequate net anode load will be present, even if a "fixed" bias "final" is employed. Notice the comparatively low 100 Kohm grid to ground limit for the 6V6, when "fixed" bias is selected. I'm not suggesting "fixed" bias, but am allowing for it.
A 500 Ω cathode bias resistor ensures adequate compliance for the 2 VRMS signal from a "standard" CDP. IMO, there's a good chance that bypassing the cathode bias resistor will be unnecessary.
The DECWARE SE84 uses μ 33 tubes: 6922 and 6H1Π (6n1p) to drive the 6Π15Π (6p15p), AKA SV83. There have been reports of insufficient gain in combination with a "standard" CDP and the 6Π15Π (like the EL84/6BQ5) is slightly easier to drive than the 6V6. Moving to a μ 40 tube, while retaining low RP and high gm, gets us over those obstacles. The 12AV7 also happens to be linear and inexpensive. For instance, Jim McShane charges $6.50 for a 12AV7.
BTW, anybody with 5965s on the "pile" should exploit that industrial 12AV7 "equivalent".
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.