I want to be able to use tubes that are reasonably off balance in sections or reasonably unmatched tube to tube. But I'm hesitant to add a balance control where I lose gain or cause roll off problems or cross talk or impedace issues, etc. So what I was thinking was to design the amp with 2 VU meters at the output, and a switch on the back that will send a mono 2v test tone Y'd into both inputs. The little oscillator would be built in. Then have a series trimmer pot on each channel input (somewhere).
So to "calibrate" the amp for the mismatched tubes one would turn on the oscillator and attenuate the trimmer on the more powerful channel to match the weaker channel with the VU meters. This way end-to-end the amp would balance out where it counts, the output. When I change a tube I re-calibrate it. In headphone listening I've become very sensitive/annoyed to the channels not balanced, it even happened on a solid state head amp I bought years ago, that device got quickly relegated to the junk box, it had no balance control.
Is this a balancing technique seen before in the history of amplifiers? Could channels be trim balanced with series attenuation like this, or must I investigate other means?
So to "calibrate" the amp for the mismatched tubes one would turn on the oscillator and attenuate the trimmer on the more powerful channel to match the weaker channel with the VU meters. This way end-to-end the amp would balance out where it counts, the output. When I change a tube I re-calibrate it. In headphone listening I've become very sensitive/annoyed to the channels not balanced, it even happened on a solid state head amp I bought years ago, that device got quickly relegated to the junk box, it had no balance control.
Is this a balancing technique seen before in the history of amplifiers? Could channels be trim balanced with series attenuation like this, or must I investigate other means?
Your maching could work for one frequency, possibly not for the whole audio frequency range. Matching could depend on the amount of drive (2 V is just one value, you would have to balance all values, from 0 to 2 V). I don’t think it will work when measuring accurately...
Regards, Gerrit
Regards, Gerrit
Thanks,
To cover a wider frequency range the oscillator can have a series of frequency doublers so that a stack of octaves is actually sent to the input. Maybe compromise of 1.25 volts on the amplitude. Mostly I'm thinking series trim pots wouldnt be creating a filter, just attenuating.
To cover a wider frequency range the oscillator can have a series of frequency doublers so that a stack of octaves is actually sent to the input. Maybe compromise of 1.25 volts on the amplitude. Mostly I'm thinking series trim pots wouldnt be creating a filter, just attenuating.
If you use an amp with global feedback, just hand match the two resistors for each channel that set gain, then the outputs will be exceptionally well balanced (and far better balanced than you could ever get them by hand tuning with a VU meter).
You can balance with signals and meters. But I submit that if you can't balance "by ear", then it does not matter.
When I was young I was very good at this. Late middle age I got uncertain and then obviously wrong, as my hearing fell-off asymmetrically (>10dB difference).
When I was young I was very good at this. Late middle age I got uncertain and then obviously wrong, as my hearing fell-off asymmetrically (>10dB difference).
I see.
Can one say the use of CCS on tubes these days is essentially eliminating match variations as well? Making the circuit kind of immune to unmatched tubes?
My obsession with a balance control only exists on headphone amps. With speakers the whole room is blending everything in a sea of reflections and waves, so there balance that is "close enough" is "good enough" for me. At that point my rug, chair, curtains open vs closed, where I'm at on the couch, etc. matter more than a balance control. But with headphones, off-balance channels is the first thing I hear, there is no room component.
No, using a CCS will not make up for poorly balanced tubes.
If you don't want to use NFB, then a balance control is a good idea. It would be tough to imagine you don't have a few spare dB to give away to be able to use a balance control. I am not aware that crosstalk is an issue with balance controls. Impedance issues are part of the design process and not all that different than selecting an appropriate volume control in the first place.
If you don't want to use NFB, then a balance control is a good idea. It would be tough to imagine you don't have a few spare dB to give away to be able to use a balance control. I am not aware that crosstalk is an issue with balance controls. Impedance issues are part of the design process and not all that different than selecting an appropriate volume control in the first place.
Generalizations:
Left and Right channel balance:
My room acoustics, and speakers, and ears, are not perfectly "balanced".
And many recordings are not perfectly balanced.
Some recordings put everything at or near the middle, use pan-pots, and other manipulations, and then calls it 'stereo'
Look on an XY plot on your scope and see that.
But if there is true separation, and there is a Piano and Bass on the left side, and a drum and a saxophone on the right side,
and nothing in the middle, then what is the meaning of channel balance, and what is so important about balance.
Tube Balance:
For a push pull output stage, the first thing to balance is the DC current.
Push pull output transformers do not work as well with un-balanced DC.
Then, there is the issue of AC gain balance in the same output stage.
Using individual CCS in the output tubes, does not fix AC balance, just DC balance, which makes for a lot of circuit that can not solve both problems at once.
And if the outputs are single ended, CCS does not fix AC balance (back to channel balance as in the top paragraph).
Negative feedback can help reduce the distortion due to un-balanced AC gain.
But if you have to use too much negative feedback to fix it, the amplifier might be unstable, especially when driving a difficult loudspeaker load.
Make things work good Before you apply negative feedback.
In a cathode coupled phase splitter that uses a True Current Sink in the cathode circuit,
and that has perfectly matched plate resistors, and perfectly matched grid resistors in the next stage, the balance is intrinsic (Automatic) . . .
. . . Unless you have a gassy tube; or unless you drive the phase splitter so hard that it draws grid current; or unless you completely cut off one of the tubes.
All of those exceptions are just things you need to be sure to keep from happening.
In todays world, automated test equipment can do good matching. Why not purchase matched tubes?
Even though it is a little more expensive, you save money and time by not having to use extra parts, extra connections,
and extra negative feedback.
Life is to short to spend time and effort to save a little money. Time is money.
If you are using Very Rare tubes, then that is one extra problem you have to overcome.
Left and Right channel balance:
My room acoustics, and speakers, and ears, are not perfectly "balanced".
And many recordings are not perfectly balanced.
Some recordings put everything at or near the middle, use pan-pots, and other manipulations, and then calls it 'stereo'
Look on an XY plot on your scope and see that.
But if there is true separation, and there is a Piano and Bass on the left side, and a drum and a saxophone on the right side,
and nothing in the middle, then what is the meaning of channel balance, and what is so important about balance.
Tube Balance:
For a push pull output stage, the first thing to balance is the DC current.
Push pull output transformers do not work as well with un-balanced DC.
Then, there is the issue of AC gain balance in the same output stage.
Using individual CCS in the output tubes, does not fix AC balance, just DC balance, which makes for a lot of circuit that can not solve both problems at once.
And if the outputs are single ended, CCS does not fix AC balance (back to channel balance as in the top paragraph).
Negative feedback can help reduce the distortion due to un-balanced AC gain.
But if you have to use too much negative feedback to fix it, the amplifier might be unstable, especially when driving a difficult loudspeaker load.
Make things work good Before you apply negative feedback.
In a cathode coupled phase splitter that uses a True Current Sink in the cathode circuit,
and that has perfectly matched plate resistors, and perfectly matched grid resistors in the next stage, the balance is intrinsic (Automatic) . . .
. . . Unless you have a gassy tube; or unless you drive the phase splitter so hard that it draws grid current; or unless you completely cut off one of the tubes.
All of those exceptions are just things you need to be sure to keep from happening.
In todays world, automated test equipment can do good matching. Why not purchase matched tubes?
Even though it is a little more expensive, you save money and time by not having to use extra parts, extra connections,
and extra negative feedback.
Life is to short to spend time and effort to save a little money. Time is money.
If you are using Very Rare tubes, then that is one extra problem you have to overcome.
Last edited:
A very smart technique used by one amplifier manufacturer:
An onboard low level oscillator was applied to both channels, but out of phase.
Ono of the speakers was connected to the Left channel Hot speaker tap, and the other speaker's second connection was connected to the Right channel Hot speaker tap.
You merely turned the balance control until there was no sound from the speaker.
This Null method is the most sensitive, Nulls are Very sharp, do not turn the balance control quickly, or you will pass over the null and not even know it.
Then the normal operation is restored to the amplifier and speaker connections too . . . with perfectly balanced amplifier channel gains.
An onboard low level oscillator was applied to both channels, but out of phase.
Ono of the speakers was connected to the Left channel Hot speaker tap, and the other speaker's second connection was connected to the Right channel Hot speaker tap.
You merely turned the balance control until there was no sound from the speaker.
This Null method is the most sensitive, Nulls are Very sharp, do not turn the balance control quickly, or you will pass over the null and not even know it.
Then the normal operation is restored to the amplifier and speaker connections too . . . with perfectly balanced amplifier channel gains.
Last edited: