Volume control pot selection

[Andrew Howard]

What is the best way of selecting a volume control pot for amps, short of
trial and error? Is there a formula that I can use? I'm guessing it might
have to do with ohms law, but I'm not sure.
Also, can anyone direct me to sources of information regarding to what
happens with power rating, impediance, etc when connecting speakers
together?

Thanks in advance
Andrew Howard

 

[Dana Raymond, a minor God]

Well, since this is a DESIGN group, I'd thought I share some design
solutions I've come up with over the last 25 years or so.

I once had a PIC with only two spare output pins but needed several more
additional digital outputs. These outputs weren't high speed at all (simple
control signals) so a contrived solution was acceptable. This is what I did:

I used an 8 bit serial-in shift register with parralel output load strobe
control. The 8 bits of data were clocked using the serial clock pin and the
data input pin. The 8 bits of data clocked in serially were clocked onto the
output pins via the load clock pin.

One PIC pin drove the shift register serial data input. The other served as
the data clock and the load clock. It was connected directly to the shift
register serial clock, and also to the load clock through a passive network
consisting of a resistor in parallel with a signal diode and then to a cap
to ground and the load input clock of the shift register.

The pin was normally held low, which kept the shift clock and load clock
(through the diode) both low. Data would be setup on the data pin and then
clocked with a positive pulse on the serial clock. This pulse would not be
seen by the load clock pin due to the RC time constant of the resistor and
cap. Once all 8 bits were clocked in, the serial clock would be held high
for a period longer than 3tc of the RC time constant (actually much longer
to account for component+life, etc.) and the 8 output pins would update at
the same time. The clock line would then be returned to its nominal state,
low.

What's nice about this method is that both input and output can be expanded
for as many devices you care to add, all using 2 pins. In the case of input
expansion, the data pin can be switched to input mode while the parralel to
serial shift register is being clocked.

Anyone else have an interesting solution to share?

Dana Raymond

 

[Ban]

What is the best way of selecting a volume control pot for amps,
short of trial and error? Is there a formula that I can use? I'm
guessing it might have to do with ohms law, but I'm not sure.
Also, can anyone direct me to sources of information regarding to what
happens with power rating, impediance, etc when connecting speakers
together?

Thanks in advance
Andrew Howard

Andy,
of course there are several possibilities connecting your pot, like inside
the feedback loop of an inverting opamp, but if you do the usual (right end
to input, wiper to output, left end to GND) you can take any value between
10k to 100k. 10k is preferred if you drive the pot with low impedance, like
CD-player etc.
If you buffer the pot with a high input impedance non-inverting opamp stage,
it will be beneficial to use a logarithmic taper.
If the buffer is low impedance, choose a linear tapered one and use around
15-20% of the pot value as input impedance for the next stage. This again
will create kind of logarithmic behaviour in most of its range.

Our ears react logarithmically so these ways a volume control will give a
similar increase in loudness over its whole range.
If you have a volume control already in the preamp, you better take a linear
tapered pot. They have better tracking of the stereo sections and can
dissipate more power(in this case not of much relevance).

ciao Ban

 

[Dana Raymond, a minor God]

Nothing to do about audio resistor calculations. I thought I was simply
posting to the NG.
I changed the subject line... isn't that enough?
Thanks, Dana Raymond

 

[YD]

Nothing to do about audio resistor calculations. I thought I was simply
posting to the NG.
I changed the subject line... isn't that enough?
Thanks, Dana Raymond

No, the references are still there, so it ends up in the original
thread. You'd do better posting a new message rather than a randomly
chosen follow-up.

- YD.