Log/antilog balance pot

[TheLoneRoger]

I've built a CCD phaser from a design I found in an old copy of ETI, and the mix pot calls for a 10k log/antilog pot, which I'm guessing is basically a pan pot, but I can't seem to source one anywhere.

I tried using an MN blend pot, but as pointed out here, https://guitarnuts2.proboards.com/thread/6761/blend-pan-pots, it's not the same thing, and it doesn't really do the job.

Any ideas where I can get such a thing?

[GGBB]

I can't be sure without seeing a schematic, but I suspect the pot would actually be log-log with the operation being log-antilog (CW-CCW). A log pot implemented/operated in reverse becomes antilog - e.g. left-handed guitar audio taper pots are really just reverse-audio taper pots.

EDIT: Please ignore my momentary lapse of reason.

[R.G.]

There are two possibilities. One is to look at the circuit and see if it's practical to put in an opamp based panning circuit, along the lines of "Panning for Fun" at geofex.com. Yeah, yeah, it uses opamps and resistors and caps, but it produce a very good approximation of a constant-power pan pot.

Another is this: if the circuit uses the log/antilog pot as two effectively counterrotating volume controls, you can probably use a dual 50K pot and add tapering resistors, as recited in "The Secret Life of Pots" again at geofex.com.

I think that finding and paying for a log/antilog pot may be an expensive and time consuming process.

[Mark Hammer]

I can't be sure without seeing a schematic, but I suspect the pot would actually be log-log with the operation being log-antilog (CW-CCW). A log pot implemented/operated in reverse becomes antilog - e.g. left-handed guitar audio taper pots are really just reverse-audio taper pots.

What he is describing is a W taper.  These are frequently used for things like tone controls and graphic equalizers, where you want opposite tapers for boost and cut.  The nice thing about that taper and their typical application is that they generally come with a centre detente.  Sadly, the smallest value that Tayda carries in them is 20k.

[willienillie]

I think he's talking about a dual-gang pot, like the balance control in many old stereos.

[Rob Strand]

Magazine page 58, 
https://worldradiohistory.com/UK/Electronics-Today-UK/70s/ETI-1978-05-78.pdf

Schematic on Magazine page 60,  pot at top right.

You could probably get away with just using a dual linear pot.
Next step you could add some fixed resistors to bend the taper like RG's article.

Another way would be to just use a single gang linear pot bridged between the two output points then take the output off the wiper.

As for the other single antilog pots.  You can use a linear pot to some degree.  The LFO speed might get bunched up at one end.   You can't fix that with resistors.   One way around that is to use a log pot backwards.   The downside of that is the LFO speed is high in the anticlockwise position.

[TheLoneRoger]

Thanks for your replies.

Yes, Rob, that's the one!

I did manage to source the single antilog pot for the LFO, so that's not a problem.

I did try using a dual linear pot with fixed resistors to fake the curves, but it didn't seem to work too well.
However, that was early on when the project probably had a few other issues, so I'll give it another try now that the rest of it is performing better.

It's quite possible I got the values wrong too, or even the tapers reversed - I had to reverse the gangs to get the MN pot working - luckily I wasn't using PCB mounted pots like the article suggests.

I hadn't heard of a W taper, I'll try searching for that too.

[Rob Strand]

I did try using a dual linear pot with fixed resistors to fake the curves, but it didn't seem to work too well.
However, that was early on when the project probably had a few other issues, so I'll give it another try now that the rest of it is performing better.

It's quite possible I got the values wrong too, or even the tapers reversed - I had to reverse the gangs to get the MN pot working - luckily I wasn't using PCB mounted pots like the article suggests.

I hadn't heard of a W taper, I'll try searching for that too.

When I look at the output pots (RV4a and RV4b) in the original article it doesn't look right.

With the pot set to 50%, Both RV4a and RV4b are going to divide the signal down by a large factor.  Seems like too much gain loss to me.    The levels from each path are equal, which looks right,  just there is a large gain drop.    (FWIW IC1b has a gain of about x 1.6 (3.9dB) to compensate for the 4dB loss in the BBD so the delay output should be about equal level wit the clean signal.)

If RV4a was log and RV4b was anti-log then the output level at 50% would not drop so much.  I have a feeling that might be the correct connection.    It makes a lot more sense as the only reason you would use log/antilog would be to do just that but the connection shown does the opposite.

Maybe with that fix the two linear pots with the taper adjust resistors would be OK.

The 2x5k1 at the output are going to cause more attenuation and to some degree they will mess-up the levels from the pot anyway.   Making those resistors higher, say 22k might help.

ETI often published errata once a year with a project Index.  Don't know what month that appears in for UK ETI, maybe Dec of the same year or Jan or April  of the following year.

The W-taper pot won't help.

[R.G.]

I think he's talking about a dual-gang pot, like the balance control in many old stereos.

Yes, he was. I was suggesting alternatives.

[Rob Strand]

I checked out the performance of the original pot wiring and the suspect fixed (ie. swapping the RV4 gangs on the ETI schematic).

It's pretty clear the original doesn't work well.    In the middle position there is a large volume drop compared to the ends.

The fix produces about a 4dB boost in signal at the middle setting.   This is OK.  If you think about a normal Flanger or Chorus it is quite common to just add the Dry and Delay signals.   If those two signals were the same signal that combination would give a 6dB (x2) gain.   We don't hear the 6dB boost on a Flanger or Chorus because the delay causes peaks and troughs when the signal is combined and on average the level is to 0dB.   A -1dB gain is often perceived as equal level, so that would equate to a target gain in your case of 5dB.    So the 4dB isn't far off.

If you used linear pots with taper tweak resistor you might be able tweak the level when the pot is at 50% position a little better.

FYI, there is only little benefit increasing the output resistors from 5k6 to 22k.    It certainly doesn't fix the problem.


The dB's in the table show the dB level relative the signal level at either extreme of the pot..

This is with the output mixing resistors at 5k.
                             
              Dry Signal Taper %     
              (%R at 50% rotation)               
                20           80
pot%    dB Sum  dB Sum
0           0.00         0.00
5          -0.96         0.86
10        -1.78        1.48
15        -2.51        1.95
20        -3.18        2.32
25        -3.83        2.65
30        -4.47        2.94
35        -5.16        3.21
40        -5.92        3.49
45        -6.82        3.77
50        -7.96        4.08
55        -6.82        3.77
60        -5.92        3.49
65        -5.16        3.21
70        -4.47        2.94
75        -3.83        2.65
80        -3.18        2.32
85        -2.51        1.95
90        -1.78        1.48
95        -0.96        0.86
100        0.00        0.00
               
min=        -7.959        0.000
max=        0.000        4.082
avg=        -3.68        2.35

[Rob Strand]

I tried a linear 50k pot with 3k3 to 8k2 taper-tweak resistor between the wiper and the opamp side of the pot, on both pots.

A lower value taper tweak resistor boost the signal level of the 50% position relative to the ends.

You could play with the value bu I suspect if you chose 4.7k it would be fine.

The 5k6 output resistors have even less effect on the behaviour because the taper-tweak resistors make the pot output impedance look almost constant.

[willienillie]

I think he's talking about a dual-gang pot, like the balance control in many old stereos.

Yes, he was. I was suggesting alternatives.

I was actually referring to M.C.'s post above mine about a W taper pot, but looking back it seems I took him out of context.