LM3886 Input Resistor Position

[danielwritesbac]

Howabout a dual gang stereo volume pot used on a monobloc amplifier, whereby the first voltage divider is used normally and the second voltage divider is reemployed as a variable resistor series to the amp, set up so the load on the amp doesn't change when you turn the knob? Sort of a t-net volume control idea there.

 

[tsiros]

so we need low and high frequency attenuation at the input and the feedback network while simultaneously avoiding π phase shifts in the feedback loop across the entire spectrum or at least where the gain is >= 1?

 

[ChristianThomas]

Howabout a dual gang stereo volume pot used on a monobloc amplifier, whereby the first voltage divider is used normally and the second voltage divider is reemployed as a variable resistor series to the amp, set up so the load on the amp doesn't change when you turn the knob? Sort of a t-net volume control idea there.

I don't see why that shouldn't work.

Because of my background, and because there is a limited number of things one can do with a chip amp, I'm quite particular about not changing loads and defining exactly where the roll-offs should be. (Or at least having them stay in the same place.) I don't think I've liked many of the solutions we've seen in datasheets and app notes from National - which is odd since they presumably know more about the chips than we ever shall. They look to me as though they are no more than a couple of days work and often they seem to be merely lifted from the previous one. And one or two have solutions that I really don't like, such as setting the mid voltage on the single supply example.

The way I got around this in the first amp of these I did was to have an active volume control - ie. the pot in the fb loop of an inverter. (Incidentally you put an R in parallel with it, which gets over the problem of it going open circuit and of noise as you twiddle it.) This also has the theoretical advantage of linearising the pot, though I'm not quite sure whether that is entirely the case.

I'm not entirely sure how I'd do it without an extra op amp but probably I'd strive to use the v. high input impedance of IN+ and have the pot then set the full input R, irrespective of where the centre was. You could then put the Cap in front of that for the high pass. An RC in front of that would then do the trick but might be a bit dependent on the output Z of the source. Doing it afterwards runs into the same problem again, in principle, so you'd have still be quite careful about your choice of ratios between what is R and C for the same product. Nothing that LTspice and a bit of measurement can't sort out, though.