How do you decide pot taper?

pucho812

Well-known member
Joined
Oct 4, 2004
Messages
12,194
Location
third stone from the sun
Was given a demo, pre- production pair of mic pres to try out. They sound really nice but we noticed in taking measurements and when in use from about 7 o’clock far left on the Gain  pot till about 3 o’clock, moving the pot to the right. There is 0 change in level. Past 3 o’clock  Till about 5 o’clock  where travel ends, we have all the increase in gain action.
I am wondering if it’s due to pot taper or pot value or both.
This sounds like a problem we have heard before, any insight?
 

Dualflip

Well-known member
Joined
Nov 15, 2008
Messages
1,190
I guess it depends where you place the pot, for example imagine you have a non inverting amplifier, if you place the pot in place of a feedback resistor then gain is directly proportional to the resistance of the pot, a linear pot will give a linear gain change, however if you place the pot in another place, for example in series with the resistance going to ground, gain is proportional to the inverse of the resistance, and if you plot in a graph, you'll notice that gain will not change linearly, so a linear pot will not be the best choice. So in the end it depends on how the pot is being used, for example in an instrumentation amplifier, gain is inversely proportional to the gain resistor, such is the case of the THAT1510/1512 or the INA217, and thats why it is recommended to use a reverse log pot in that case.
 

pucho812

Well-known member
Joined
Oct 4, 2004
Messages
12,194
Location
third stone from the sun
well, I don't have the schematic so I can't say where it is in the chain.

I can tell you that the preamps have a mic input transformer, a mic output transformer and runs off a single opamp, an LM709.  I can't imagine anything  being out of the ordinary and considering it was based  off an old quad eight/eletrodyne model I  would suspect they stayed close to the data sheets.  the pot itself is a 25K pot with unknown taper at the moment. 
 

JohnRoberts

Staff member
Moderator
Joined
Nov 30, 2006
Messages
19,942
Location
Hickory, MS
The most popular topology for low noise mic preamps requires an extreme reverse log taper. At Peavey we tooled up a custom pot (required 3 or 4 screened resistance overlays).

This is ergonomics or human factors engineering... If you noticed it customers will notice too.

JR
 

john12ax7

Well-known member
Joined
Oct 14, 2010
Messages
1,796
Location
California, US
Sounds like a pot taper issue,  and perhaps value as well.  I agree with JR regarding the need for extreme tapers.  The typical 15% reverse log is generally not sufficient to give smooth response in a lot of 60dB mic pre applications.
 

pucho812

Well-known member
Joined
Oct 4, 2004
Messages
12,194
Location
third stone from the sun
Gold said:
A new design with an LM709? Bold. Metal can?

yes metal can. Not a new design so much as a  new product.  not mine I am just playing with them and for 80% of the throw there is no volume change, and in the last 20% it increases as one would think should happen for the full travel.  Perhaps the  original unit they modeled it after does this?
 

Youngwhisk

Well-known member
Joined
Oct 13, 2012
Messages
67
Location
Canada
At approx 45:35 in this video he explains how he solved a pot taper problem (similar to your problem). He explains the differences in the tapers.

https://www.youtube.com/watch?v=KPsRTScY4TI

I don't know much about tapers but this seems reasonable to me.

A very interesting piece of equipment.
 

abbey road d enfer

Moderator
Joined
Jan 22, 2008
Messages
11,953
Location
Marcelland
pucho812 said:
I can tell you that the preamps have a mic input transformer, a mic output transformer and runs off a single opamp, an LM709. 
With such a topology, it doesn't take much to make it right.
I would think the gain range to be much less than 60dB, rather about 40.
A standard Log pot in rheostat mode in the NFB loop gives adequate results. Connecting it as a potentiometer improves the gain law significantly.
On the first mic pre I designed professionally, I even managed to get acceptable law from a linear pot.
I cannot imagine why they did not make it right...
 

nedonnelly

Member
Joined
Jun 27, 2017
Messages
15
Youngwhisk said:
At approx 45:35 in this video he explains how he solved a pot taper problem (similar to your problem). He explains the differences in the tapers.

https://www.youtube.com/watch?v=KPsRTScY4TI

I don't know much about tapers but this seems reasonable to me.

A very interesting piece of equipment.

I have worked with Reverse Log, Linear, and Audio pots without realizing what they are actually doing. Thanks for sharing this. It may not be new info for those that are experienced (as many on here are), but for good for those of us that are new to learn these things.  :D
 

gridcurrent

Well-known member
Joined
Aug 28, 2009
Messages
590
pucho812 said:
I can tell you that the preamps have a mic input transformer, a mic output transformer and runs off a single opamp, an LM709.  I can't imagine anything  being out of the ordinary and considering it was based  off an old quad eight/eletrodyne model I  would suspect they stayed close to the data sheets.  the pot itself is a 25K pot with unknown taper at the moment.
ElectroDyne manufactured a mic preamp card based on the 709.
It used a 4  position slide switch on the front panel to vary gain in 10 dB increments from 0 to 30 dB (amplifier gain only excluding input transformer).
In my opinion, potentiometer taper is the least of any concerns with a variable gain 709 based preamp.
Compensation capacitors require significantly different values depending on gain.

There were techs in LA decades ago that worshipped the 709 and its designer, Mr. Widlar.
 

Attachments

  • response.jpg
    response.jpg
    273.6 KB · Views: 9

Gold

Well-known member
Joined
Jun 23, 2004
Messages
2,619
Location
Brooklyn
The only experience I have with the LM709 is in later versions of the Neumann VG66 lathe amplifier rack. Later versions of the SI66 RIAA encode and the decode on the WV66 Monitor modules use one in place of the previous discrete versions. The phono preamp also switches over to an LM709. All with compensation and transistor current buffers at the output.

Neumann considered it an improvement over the discrete version. I and everyone else I’ve talked to who has heard both prefer the discrete version.
 

JohnRoberts

Staff member
Moderator
Joined
Nov 30, 2006
Messages
19,942
Location
Hickory, MS
Gold said:
The only experience I have with the LM709 is in later versions of the Neumann VG66 lathe amplifier rack. Later versions of the SI66 RIAA encode and the decode on the WV66 Monitor modules use one in place of the previous discrete versions. The phono preamp also switches over to an LM709. All with compensation and transistor current buffers at the output.

Neumann considered it an improvement over the discrete version. I and everyone else I’ve talked to who has heard both prefer the discrete version.
The 709 was a very early (1960s) op amp and as I recall it was pretty fragile.. Easy to blow up and cost something like $50 each early on.

JR 

 

Gold

Well-known member
Joined
Jun 23, 2004
Messages
2,619
Location
Brooklyn
The discrete versions of those modules have 1968-69 Gotham Audio stamps. The ones with LM709’s were stamped 1970-1972. After that I guess they started working on the next generation SAL74.
 

abbey road d enfer

Moderator
Joined
Jan 22, 2008
Messages
11,953
Location
Marcelland
When the 709 came out, many thought it was the next big thing, and they were damn right!
But the fact that some venerated at the time must not hide the fact that much better parts have been made available.
It's taken a good decade to turn IC opamps into a part that could be used in a pro audio chain. The 741/748 family was a major step since it allowed a compensation scheme that worked almost unconditionally, but the real start was the coming of the TL0xx and 5534/32.
Except for historic correctness, I wouldn't want to use a 709 today. Actually I've never used one. Went directly to 748.
BTW I remember the first time I saw a Midas desk. The guy there proudly showed me the 3 grades of 709's they used, all hand-sorted at the factory.
 

JohnRoberts

Staff member
Moderator
Joined
Nov 30, 2006
Messages
19,942
Location
Hickory, MS
abbey road d enfer said:
When the 709 came out, many thought it was the next big thing, and they were damn right!
But the fact that some venerated at the time must not hide the fact that much better parts have been made available.
It's taken a good decade to turn IC opamps into a part that could be used in a pro audio chain. The 741/748 family was a major step since it allowed a compensation scheme that worked almost unconditionally, but the real start was the coming of the TL0xx and 5534/32.
Except for historic correctness, I wouldn't want to use a 709 today. Actually I've never used one. Went directly to 748.
BTW I remember the first time I saw a Midas desk. The guy there proudly showed me the 3 grades of 709's they used, all hand-sorted at the factory.
+1

Don't forget the old LM301  (10V/uSec when using feedforward compensation).

The much maligned 741 was historic for being internally compensated for unity gain stability giving design engineers one less way to screw up. While 0.5V/uSec seems lethargic by modern expectations decent audio could still be crunched using lower nominal 0VU levels (like -10 dBV).

JR
 

Dualflip

Well-known member
Joined
Nov 15, 2008
Messages
1,190
JohnRoberts said:
+1

Don't forget the old LM301  (10V/uSec when using feedforward compensation).

The much maligned 741 was historic for being internally compensated for unity gain stability giving design engineers one less way to screw up. While 0.5V/uSec seems lethargic by modern expectations decent audio could still be crunched using lower nominal 0VU levels (like -10 dBV).

JR

The LM301 is an interesting opamp IMHO, several tricks could be done with the compensation circuitry, for instance, you could replace the input section of the opamp with a discrete transistor stage using the compensation pins like it was done on the Studer 169 I/O Module. Also, they make decent comparators when used uncompensated, you can get nice crisp edges, its my go to opamp for stuff like square wave oscillators.

Almost all of the National Applications Handbook is based on the LM301/101 opamp.
 

JohnRoberts

Staff member
Moderator
Joined
Nov 30, 2006
Messages
19,942
Location
Hickory, MS
Dualflip said:
The LM301 is an interesting opamp IMHO, several tricks could be done with the compensation circuitry, for instance, you could replace the input section of the opamp with a discrete transistor stage using the compensation pins like it was done on the Studer 169 I/O Module. Also, they make decent comparators when used uncompensated, you can get nice crisp edges, its my go to opamp for stuff like square wave oscillators.

Almost all of the National Applications Handbook is based on the LM301/101 opamp.
The feedforward cap I mentioned allows you to bypass the slow input stage for 10V uSec, but only inverting topology.

Since the 70s we haven't had to use compensation tricks or even compensation, because modest cost op amps were better than audio.

JR
 

Latest posts

Top