Dual-Drive designs

[jmcd]

The Marshall 1987x and a host of other amps (including the Trainer YBA1, Viva Canada!) have dual inputs, each with its own volume. The parallel inputs come together by way of equal-value resistors (470k in the Marshalls; 100k in the YBA1), often with one of the resistors bypassed with a cap. Many people mod the YBA by swapping out the 100k resistors with 470k resistors in order to 'plexi-fy' the Bass Master. I have several questions about these resistors:

1. Why are they there? If they act as grid stoppers, why not have just one?
2. In what way does their value matter?
3. If bypassed with a cap, as on the brighter input of the 1987x, does this form a bright boost or cut? Initially, I assumed that it was a boost, but now I'm thinking that it interacts with the tube's miller capacitance to form a cut. 


Thanks in advance for the edification...

[K O'Connor]

Hi Guys

As TUT refers to these resistors, they are for mixing and for isolation. Yes, contradictory functions

With a single signal path, the output of the level control could be fed more or less directly into the next gain stage. If we add a second input path and Level-2 control, the controls would interact with each other. Assuming we simply tie the wipers together, adjusting either pot effects both signals. To reduce this interaction, isolation resistors are added in series with each wiper.There is still interaction but it is greatly reduced.

With one level control set to zero, its resistor is grounded. The other pot can be set however we need and the signal sees a voltage divider comprised of the two resistors, which if they are equal values reduce the signal by half. If both pots are up, then both signals can pass to the next stage, both being attenuated by the divider-action of the two resistors.

If we add a cap across one resistor, that signal will have a treble emphasis. When the related pot is at zero, there will be a treble
de-emphasis for the other signal.

Why 470k versus 100k? The higher values load the pots less and allow the use of lower-value caps for frequency shaping.

In one of the versions of the Fender Bassman, the stage after the passive mixing had feedback applied, which made the mixing active as the overall circuit becomes a virtual-earth mixer. This minimizes the interaction between the pots and the frequency shaping works only on the signal path it is built into.

In neither form do the mixing/isolation resistors act as grid-stops in the true sense. There is usually quite a lot of wire between these Rs and the tube grid, so things can happen along the way. However, these Rs are large enough to work against the next stage's internal capacitance to impose a high-frequency roll-off.

Have fun

[jmcd]

Awesome. I can see it now! Thanks, KOC!