I've serviced a couple of both, but for the first time lately converted the time constants from original to the factory modification of January 1949, which makes it a dual constant, more like what the BA-6 is like.
I've never seen an updated late stage schematic showing the time constant changes, perhaps that is a short run 86-D in that last year of production?
This is the second version of the single time constant, the first (in the 96-A) is a 10M release resistor.
https://groupdiy.com/index.php?topic=44740.msg560803#msg560803
These are the update notes:
https://groupdiy.com/index.php?topic=44740.msg971657#msg971657
The single constant with 3M3 release resistor is more like a Gates SA-39, somewhat fast and bouncy/pumpy, which can be a feature or a negative depending on what you're wanting.
The dual is remarkably slow and unobtrusive, makes the unit more of a wide-range AGC, capable of very deep compression that slow rides a program, they were definitely thinking about classical music here.
It stands out that I've never heard these differences discussed, probably because so few people have one of each to compare, or have done the mod, or even know there was one.
The way power supply paths and loads are implemented, and the way meter zero and input stage bias are derived from those resistor chains, makes it less obvious what can be changed.
I may be wrong, but the meter zero range appears to be a function of tube bias point affecting the current measurement resistor load and thus the meter zero point, with no isolated zero tuning control. They wanted it to land at a very exact state of being, which today may not be as easily achievable with available tubes.
If you look at the bottom release resistor in the path to the meter zero control as variable, and change it to a smaller/faster value, you'll find you've changed the bias point and meter zero (out of range) along with it. Resistance from the top of the dual timing stack to ground will speed up release without changing the meter zero.
This most recent example did not have much range at all on the meter zero pot, and all the resistances in the chain measured within tolerances. Perhaps a function of exact 6K7 current.
I've never seen an updated late stage schematic showing the time constant changes, perhaps that is a short run 86-D in that last year of production?
This is the second version of the single time constant, the first (in the 96-A) is a 10M release resistor.
https://groupdiy.com/index.php?topic=44740.msg560803#msg560803
These are the update notes:
https://groupdiy.com/index.php?topic=44740.msg971657#msg971657
The single constant with 3M3 release resistor is more like a Gates SA-39, somewhat fast and bouncy/pumpy, which can be a feature or a negative depending on what you're wanting.
The dual is remarkably slow and unobtrusive, makes the unit more of a wide-range AGC, capable of very deep compression that slow rides a program, they were definitely thinking about classical music here.
It stands out that I've never heard these differences discussed, probably because so few people have one of each to compare, or have done the mod, or even know there was one.
The way power supply paths and loads are implemented, and the way meter zero and input stage bias are derived from those resistor chains, makes it less obvious what can be changed.
I may be wrong, but the meter zero range appears to be a function of tube bias point affecting the current measurement resistor load and thus the meter zero point, with no isolated zero tuning control. They wanted it to land at a very exact state of being, which today may not be as easily achievable with available tubes.
If you look at the bottom release resistor in the path to the meter zero control as variable, and change it to a smaller/faster value, you'll find you've changed the bias point and meter zero (out of range) along with it. Resistance from the top of the dual timing stack to ground will speed up release without changing the meter zero.
This most recent example did not have much range at all on the meter zero pot, and all the resistances in the chain measured within tolerances. Perhaps a function of exact 6K7 current.
