I think Grey means that the resistor reduces the
performance, assuming that the diff pair is current
sourced by a resistor. In this case it's not such a
great phase splitter, although it does work.
It hinges on the current sourcing; if the current source
is constant, then the resistor simply reduces the gain,
and the amplitudes of both halves are equal.
performance, assuming that the diff pair is current
sourced by a resistor. In this case it's not such a
great phase splitter, although it does work.
It hinges on the current sourcing; if the current source
is constant, then the resistor simply reduces the gain,
and the amplitudes of both halves are equal.
Bricolo
I modified my BOSOZ, using constant current source tails recommended by Nelson Pass. It is here in a detailed circuit.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=3949&perpage=15&pagenumber=30
It does great work as a pre plus unbalanced-balanced converter for my system.
If you adopt one CCS tail, you might need two gain control resistors of the same values between the sources: one to the left side and another to the right side of the centered CCS.
Anyhow, thanks for your kindness.
JH
I modified my BOSOZ, using constant current source tails recommended by Nelson Pass. It is here in a detailed circuit.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=3949&perpage=15&pagenumber=30
It does great work as a pre plus unbalanced-balanced converter for my system.
If you adopt one CCS tail, you might need two gain control resistors of the same values between the sources: one to the left side and another to the right side of the centered CCS.
Anyhow, thanks for your kindness.
JH
I've got no more questions, the figure explains all of them 
the resisitor will be used to set the DC offset, am I right?
I just have another question, something I thought about this morning.
We are all trying do zero the DC offset. But we mesure the offset at idle; with no input signal.
It is possible, than the DC offset can have been set to 0 ad idle, and that with an incominc signal, we will still have an offset
in other words: can the dissimetry be function to the load?
the resisitor will be used to set the DC offset, am I right?
I just have another question, something I thought about this morning.
We are all trying do zero the DC offset. But we mesure the offset at idle; with no input signal.
It is possible, than the DC offset can have been set to 0 ad idle, and that with an incominc signal, we will still have an offset
in other words: can the dissimetry be function to the load?
The offset noise isn't very noticable,
My amp starts out cold with 4 volts offset and ends up hot with 20 to 40 millvolts. Along the way I can hear the AX smoothing out and becoming creamier and lush. I don't understand the need to worry about a small imbalance. Just match the MOSFETs to a close tolerance, build it correctly and enjoy. Now if you're one of those types that can't sit still and smell the roses, so to speak, then keep on plugging away. A large jar of aspirin helps.
John
My amp starts out cold with 4 volts offset and ends up hot with 20 to 40 millvolts. Along the way I can hear the AX smoothing out and becoming creamier and lush. I don't understand the need to worry about a small imbalance. Just match the MOSFETs to a close tolerance, build it correctly and enjoy. Now if you're one of those types that can't sit still and smell the roses, so to speak, then keep on plugging away. A large jar of aspirin helps.
John
Now that I have built a couple production runs I can tell
you that 4 volts offset cold is out there. Most of the time
we see a small fraction of a volt that settles in to < 50 mV
over a range of temperatures. You might want to check your
diff pair matching; perhaps they are not from the same lot
code. Are you running DC input? This will exacerbate
(new vocabulary word to learn) the offset, but then even a
volt under these circumstances is a bit much.
you that 4 volts offset cold is out there. Most of the time
we see a small fraction of a volt that settles in to < 50 mV
over a range of temperatures. You might want to check your
diff pair matching; perhaps they are not from the same lot
code. Are you running DC input? This will exacerbate
(new vocabulary word to learn) the offset, but then even a
volt under these circumstances is a bit much.
