What constitutes a “properly designed” bias circuit? Battery of correct voltage? What else?
A typical battery bias circuit that works well:
Pick a battery voltage that gives the correct voltage for grid bias.
1.5, 2, 3, 4.5, 6, 7.5, 9, . . . 90V, etc.
Connect the input signal to a grid leak resistor to ground.
Connect the junction of the input signal and grid leak resistor to the battery + lead.
Connect the battery - lead to one end of a grid stopper resistor.
connect the other end of the grid stopper resistor to the grid.
The only time that battery current is drawn, is when the signal is positive enough to draw grid current (and that current direction does not discharge the battery).
But do not let the signal be large enough to draw grid current (turn the volume down, single ended and push pull A1, and push pull AB1 circuits generally sound distorted when grid current is drawn.
Do the above, and the battery should last as long as it would if it stayed on the shelf.
Simple, elegant, works.
Just my opinion and my experience.
Pick a battery voltage that gives the correct voltage for grid bias.
1.5, 2, 3, 4.5, 6, 7.5, 9, . . . 90V, etc.
Connect the input signal to a grid leak resistor to ground.
Connect the junction of the input signal and grid leak resistor to the battery + lead.
Connect the battery - lead to one end of a grid stopper resistor.
connect the other end of the grid stopper resistor to the grid.
The only time that battery current is drawn, is when the signal is positive enough to draw grid current (and that current direction does not discharge the battery).
But do not let the signal be large enough to draw grid current (turn the volume down, single ended and push pull A1, and push pull AB1 circuits generally sound distorted when grid current is drawn.
Do the above, and the battery should last as long as it would if it stayed on the shelf.
Simple, elegant, works.
Just my opinion and my experience.
@6A3sUMMER, thanks for your reply. I have noticed in past threads you have advised this scheme of battery bias where the signal has to propagate through the battery. In this thread for example:https://www.diyaudio.com/community/threads/battery-bias-question.350715/#post-6113091
I wondered if your scheme with a battery in series with the signal isn’t more deleterious to signal fidelity than when the battery negative terminal is connected to the grid stopper and +side to ground.
Does anyone have experience with both approaches; took measurements? Perhaps this isn’t the place for that discussion. Apologies to OP.
I wondered if your scheme with a battery in series with the signal isn’t more deleterious to signal fidelity than when the battery negative terminal is connected to the grid stopper and +side to ground.
Does anyone have experience with both approaches; took measurements? Perhaps this isn’t the place for that discussion. Apologies to OP.
I am also curious about the best way to implement battery bias in order to optimize SQ, and if people have any preference for particular battery types. The 300B is an excellent performer. It seems that the majority of issues with 300B circuits are found with the driver portion of the circuit. Any methods that optimize the driver would be most welcome.
Optimize the driver in a 300B amplifier?
For a single ended 2 stage 300B amplifier, with no negative feedback:
1. Is the optimum driver one that has no 2nd harmonic distortion, so that there is no cancellation of the 2nd harmonic distortion of the 300B?
2. Is the optimum driver one that has an amount of 2nd harmonic distortion that is equal to the 300B 2nd harmonic distortion, so that it completely cancels the 2nd harmonic distortion of the 300B?
3. Is the optimum driver one that has a medium amount of 2nd harmonic distortion, so that it partially cancels the 2nd harmonic distortion of the 300B?
The question is, what do you want from the amplifier: relatively high 2nd harmonic distortion (# 1), no 2nd harmonic distortion (# 2), or a medium amount of harmonic distortion (# 3)?
Note:
If you partially cancel the 300B 2nd harmonic distortion (# 3 above), or completely cancel the 300B 2nd harmonic distortion (# 2 above), then the 3rd harmonic distortion is dominant.
If you do not cancel the 300B 2nd harmonic distortion (# 1 above), then the 2nd harmonic distortion is dominant.
For a single ended 2 stage 300B amplifier, with no negative feedback:
1. Is the optimum driver one that has no 2nd harmonic distortion, so that there is no cancellation of the 2nd harmonic distortion of the 300B?
2. Is the optimum driver one that has an amount of 2nd harmonic distortion that is equal to the 300B 2nd harmonic distortion, so that it completely cancels the 2nd harmonic distortion of the 300B?
3. Is the optimum driver one that has a medium amount of 2nd harmonic distortion, so that it partially cancels the 2nd harmonic distortion of the 300B?
The question is, what do you want from the amplifier: relatively high 2nd harmonic distortion (# 1), no 2nd harmonic distortion (# 2), or a medium amount of harmonic distortion (# 3)?
Note:
If you partially cancel the 300B 2nd harmonic distortion (# 3 above), or completely cancel the 300B 2nd harmonic distortion (# 2 above), then the 3rd harmonic distortion is dominant.
If you do not cancel the 300B 2nd harmonic distortion (# 1 above), then the 2nd harmonic distortion is dominant.
Last edited:
Most batteries are low impedance.
1. The signal source drives the low impedance of the battery, and the battery drives the grid stopper, which drives the high impedance of the grid.
That circuit causes the signal source to 'see' a high impedance.
That is what I described and recommended.
2. Post # 565 above: "when the battery negative terminal is connected to the grid stopper and +side to ground."
That circuit causes the signal source to 'see' a low impedance to ground.
Not recommended.
But that may not be what you meant to describe.
1. The signal source drives the low impedance of the battery, and the battery drives the grid stopper, which drives the high impedance of the grid.
That circuit causes the signal source to 'see' a high impedance.
That is what I described and recommended.
2. Post # 565 above: "when the battery negative terminal is connected to the grid stopper and +side to ground."
That circuit causes the signal source to 'see' a low impedance to ground.
Not recommended.
But that may not be what you meant to describe.
No measurements, but I've used both methods of battery grid bias that you describe.
I prefer the same method that 2A3sUMMER suggests but only because I like the simplicity of it. I didn't hear any difference between the two and I also didn't hear any difference between an alkaline and a lithium type. That said, I didn't spend a lot of time trying to listen closely for any audible differences and there may be measurable differences.
Since the battery life is essentially the same as its shelf life, unless grid current is being drawn, the advantage that a lithium offers is simply that it will last longer - perhaps as long as 10 years. The advantage of an alkaline is that it's cheaper.
As I understand it, another advantage of using the battery in series with the signal is that it eliminates the need for a DC blocking cap on the input. I was told that the other battery grid bias method requires a blocking cap. I'm sure someone will correct me if I'm wrong.
The first time I tried battery bias I used a battery to bias the cathode. In that scheme you need to use a rechargeable battery because there is current going through it during use and the + connects to the cathode and the - goes to ground. I abandoned that method because the amp I was using it on was only used sporadically so the battery didn't stay charged and it was a pain to have to remove it and put it on a charger.
Otherwise, it worked fine when I used it on an input tube (a 27) but I also tried to bias an output tube (forget which type) and the battery couldn't handle the amount of current being drawn. Luckily I discovered the damaged battery before it exploded.
euro21,
The major problem with using an FFT to measure very low frequency noise is:
There is lots of noise at frequencies below 100Hz, 10Hz, 1 Hz, or 0.1Hz.
Those noises are larger than KTB (-174 dBm / Hz).
The noise of a resistor is about -174 dBm / Hz at 20 or 25 degrees C.
The resistor noise.
(Not the noise across a resistor that has DC current passing through it)
The major problem with using an FFT to measure very low frequency noise is:
There is lots of noise at frequencies below 100Hz, 10Hz, 1 Hz, or 0.1Hz.
Those noises are larger than KTB (-174 dBm / Hz).
The noise of a resistor is about -174 dBm / Hz at 20 or 25 degrees C.
The resistor noise.
(Not the noise across a resistor that has DC current passing through it)
matejsirk,
Nice looking schematic!
It looks like a relatively linear amplifier that does not have any negative feedback.
Have you measured and listened to that amplifier?
Power, distortion, damping factor, gain, etc.
It would love to know the results.
What are the makes and models of the input transformer, plate choke, and output transformer?
Nice looking schematic!
It looks like a relatively linear amplifier that does not have any negative feedback.
Have you measured and listened to that amplifier?
Power, distortion, damping factor, gain, etc.
It would love to know the results.
What are the makes and models of the input transformer, plate choke, and output transformer?
Hi,
I'll be looking forward to answers to those questions. c3g triode has gain of 48 according to Ale's uTracer curves, so this schemo is giving A=196 with a 1:4 input transformer. Could the poster of this design please elaborate instead of just popping out a schemo that in his mind is ultimate? Did you build it?
Best, Robert
P.S. Looks similar to schemo posted in message # 488, Pg 25, but with input step-up transformer.
Hi thaiDIY,
Cool ! ! ! Do you know what he moved on to? I haven't tried to track down his present systeme at the Lenco Heaven site. Did you have a cup of tea when you were in that room?
Robert
Hi,
RC coupled version of the same for a bit more output, and then to 76 >10Y IT 845.
IIRC there was a cup of tea, yes
Hi Andy, et al,
Just catching up on this...sorry, sometimes I am slow.
With (Z1/Z2) = 5k/5 = (n1/n2)^2 = 1000, if Z2 = 8, then Z1 = 8K! So much for your 5K transformer, and perhaps you have gone further than Sakuma sanin exploring alternatives...what is/was your loudspeaker Z when you did that?
Best, Robert
loadlines, impedances, all based on sine waves to design the output transformers, but in the end it is the music that we listen to.....
OPT impedances is all about turns ratio, what the secondary load impedance reflects back to the plates...
the thing is, with a pure resistor at the secondary the turns ratio holds....but real world speakers are never a pure resistor load, it varies with frequencies from 20hz to 20khz, so what now?
OPT impedances is all about turns ratio, what the secondary load impedance reflects back to the plates...
the thing is, with a pure resistor at the secondary the turns ratio holds....but real world speakers are never a pure resistor load, it varies with frequencies from 20hz to 20khz, so what now?
Last edited:
What now?
Hasn't anybody learned anything from all these threads in Tubes/Valves?
The Wright brothers probably did not have all the calculations worked out; but their first real flight went a distance equal to the wingspan of a 747.
From small beginnings . . .
Even with a non-inductive load resistor on the secondary, the primary impedance from 20Hz to 20kHz depends on more than the Square of the Turns Ratio.
There are other factors, such as:
Primary DCR, secondary DCR, leakage inductance, distributed capacitance, and core losses.
Load resistors and sine waves are merely tools. Lowly resistors and sine waves.
Try responding to all your problems if all you have in your toolbox is a Hammer.
Oh, does your leg hurt, just hammer your thumb, then you might stop noticing the pain in your leg.
By the way, I listened to a couple of hours of music played on Sakuma-Sans tube amplifiers.
Some design, some build, some listen. And, I am sure some do not do any of those three.
Hasn't anybody learned anything from all these threads in Tubes/Valves?
The Wright brothers probably did not have all the calculations worked out; but their first real flight went a distance equal to the wingspan of a 747.
From small beginnings . . .
Even with a non-inductive load resistor on the secondary, the primary impedance from 20Hz to 20kHz depends on more than the Square of the Turns Ratio.
There are other factors, such as:
Primary DCR, secondary DCR, leakage inductance, distributed capacitance, and core losses.
Load resistors and sine waves are merely tools. Lowly resistors and sine waves.
Try responding to all your problems if all you have in your toolbox is a Hammer.
Oh, does your leg hurt, just hammer your thumb, then you might stop noticing the pain in your leg.
By the way, I listened to a couple of hours of music played on Sakuma-Sans tube amplifiers.
Some design, some build, some listen. And, I am sure some do not do any of those three.
Last edited: