Are these linear or audio taper pots?

I'd forgotten there was another thread discussing what this whole circuit was intended for so yes, linear pots

MagnetoSound said:

There were two questions being asked.
Some of us were responding to the 'rule of thumb' question in Mbira's third post.
No harm done.

Click to expand...

Yep. I was responding to the "how do you tell which pot to use" question so no harm done here either Mbira 

Brian Roth said:

I have seen circuits in some desks (certain Neve and ADM desks).... Typically, the 'slug' R was 1/3 or 1/4 the value of the R of the pot.

Click to expand...

It won't make a huge difference.  In classic Neve desks, I've seen a 5K linear with a 5K1 pull-down (47K in parallel on the amp side) but that's for the 'set it and forget it' bus amps. 

"Small Signal Design" by Self goes into this and determines that, with a 10K linear pot, a 4K7 value pull-down fits the logarithmic curve best.  That ratio seems to work well in practice too IMHO.
BTW, standard (read cheap) log. pots don't fit the curve especially well either and cheap linear pots with a pull down will have a closer tolerance.

Again, it won't make a huge difference but, just for completeness, the attenuation at 50% with a 10K linear pot  is:
4K7 = 9.73dB,  3k3 = 10.91dB, 2K2 = 12.61dB.  No other load on wiper of course. 

Jean Clochet said:

"Small Signal Design" by Self goes into this and determines that, with a 10K linear pot, a 4K7 value pull-down fits the logarithmic curve best.  That ratio seems to work well in practice too IMHO.
BTW, standard (read cheap) log. pots don't fit the curve especially well either and cheap linear pots with a pull down will have a closer tolerance.

Again, it won't make a huge difference but, just for completeness, the attenuation at 50% with a 10K linear pot  is:
4K7 = 9.73dB,  3k3 = 10.91dB, 2K2 = 12.61dB.  No other load on wiper of course.

Click to expand...

---veeer alert---

If it was this simple, and worked so well, there would never be any reason to buy another log taper pot. Think of the savings in inventory utilization. Big companies manage such things. 

The way pots are typically manufactured even cheap ones will track taper better than absolute resistance. while it will depend on the designed-for taper. If the tool is not accurate log, the result won't magically be log, while as I mentioned there are multiple flavors of audio taper based on loss at 50% (-10/15/20 dB) . Since the loaded taper depends on the actual resistance and the beginning taper this is not very reliable in production.  For DIY where you are typically using pots from a single production lot, go for it, but in my experience this is not normal practice for large scale production.

For small one-off DIY projects taper is not a huge deal... when selling a console with tens of identical channels side by side, customers actually expect the channels to be identical, and complain if they aren't.

JR

JohnRoberts said:

---veeer alert---

Click to expand...

Not sure if you're referring to your, about to follow post or to me. 
If it's me then I apologise.  Just felt like being thorough (maybe more than necessary?) in trying to answer Mbira's question regarding which pot is used on a schem. when it isn't indicated... 

JohnRoberts said:

when selling a console with tens of identical channels side by side, customers actually expect the channels to be identical, and complain if they aren't.

Click to expand...

This is all I meant by tolerance, nothing more.  As well as cheap dual pots for stereo which don't track too well.  I happen to find it's true that, given inexpensive parts, it's a lot easier to get two channels (or left and right) to track identically (or near enough as damn it) with a pulled-down dual linear than a dual log.  Companies such as P&G go to great lengths to fabricate the multiple resistive tracks used for the taper to ensure good tolerance from unit to unit.  If it was so easy to get it right without the effort, they'd do it.

Not sure I agree with you regarding the assumption that there would never be a need to buy log pots if it worked so well though. 
There are other factors to consider such as minimum load on the source and source resistance/impedance to the destination versus noise that would make using a log. pot more desirable.  Or maybe the need for a log. taper that doesn't change its load on the source with position  (when terminating a transformer for instance) that, I think we can assume, the manufacturers of cheap log pots have nothing to worry about for the time being 
Anyway, if we have a difference of opinion, no biggie.  Differences make the world interesting 
Peace.

Jean Clochet said:

JohnRoberts said:

when selling a console with tens of identical channels side by side, customers actually expect the channels to be identical, and complain if they aren't.

Click to expand...

This is all I meant by tolerance, nothing more.  As well as cheap dual pots for stereo which don't track too well.  I happen to find it's true that, given inexpensive parts, it's a lot easier to get two channels (or left and right) to track identically (or near enough as damn it) with a pulled-down dual linear than a dual log.  Companies such as P&G go to great lengths to fabricate the multiple resistive tracks used for the taper to ensure good tolerance from unit to unit.  If it was so easy to get it right without the effort, they'd do it.

Not sure I agree with you regarding the assumption that there would never be a need to buy log pots if it worked so well though. 
There are other factors to consider such as minimum load on the source and source resistance/impedance to the destination versus noise that would make using a log. pot more desirable.  Or maybe the need for a log. taper that doesn't change its load on the source with position  (when terminating a transformer for instance) that, I think we can assume, the manufacturers of cheap log pots have nothing to worry about for the time being 
Anyway, if we have a difference of opinion, no biggie.  Differences make the world interesting 
Peace.

Click to expand...

I am trying to share my actual experience from having specified and used many pots from multiple vendors over the years. At Peavey when I was over all mixer engineering I was nominally responsible for using millions of pots a year.

The primary issue with tracking between dual pot sections is mainly mechanical, as the indexing of wipers and resistive elements will not be very precise.  Indeed a better approach is to screen a dual track on a single pot element which will perfectly index the two tracks together but can cause other issues. I recall one special dual element used in a sweep EQ, where one of our two vendors cut corners and tried to use a standard wiper inside that special part. This led to problems from wiper bounce due to too few wiper fingers in contact with the narrow tracks. 

Better pot vendors will spec tracking between sections but for audio taper I don't recall ever seeing great +/- XdB between sections at the bottom of the pot range. I actually paid a technician back in the early '80s (at a small company) to hand tweak a 4 section audio taper pot that was the best I could buy from Alps at the time. This was the frequency control for a 4 pole crossover and I was not satisfied with filter tracking using stock parts. So we were measuring and adding shunt resistors to each one of these premium audio taper pots to get acceptable tracking between sections. I do not accept that I could have just used cheap linear tapers, and let if fly. Well I could have,,, but I didn't.

The bottom line is bulk potentiometer resistance due to the process of screened resist elements that are then heat cured (baked) to reach nominal resistance ranges is not a very accurate or reliable process...(typical 20%).  I had one pot vendor decline our business at Peavey because we were in his word's "too picky"  8). Another long story I won't bore you with again.

FWIW I am mostly talking from a production perspective, for low volume DIY, do whatever works for you.

For the OP relax, whatever taper you get will work just fine for that application.

JR

Cheers John.
I appreciate you sharing from experience. 

JohnRoberts said:

This was the frequency control for a 4 pole crossover and I was not satisfied with filter tracking... I do not accept that I could have just used cheap linear tapers, and let if fly. Well I could have,,, but I didn't.

Click to expand...

In that situation, I probably wouldn't have either 

I agree with pretty much everything in your post.  The only point I'd add is that, even though it's DIY here, I don't think we should either 1/ settle for channel to channel missmatches on our little 8 channel out-rigger mixing boxes or 2/ face paying through the nose for a P & G, TKD or XYZ...
Sometimes a faked linear works well.  If it's still not close enough for Rock 'n' Roll, a trimmer in place of the pull-down resistor will balance out the pot R tolerance so you can hit -10dB or whatever point you choose
Cheers.