Hello there,

Many of us use voltage regulators but i wonder how much attention is payed to the way the pot rotation actually affects the voltage output of the circuit. Many circuits will have a very non linear turn angle to output voltage character. I found one circuit i intend to work with soon to have this same problem, and the problem is pronounced enough to have made me look for another solution. A solution i did several years back worked pretty well, but it was a bit too complicated for what i want to use this new circuit for, so i was hoping to find a simpler solution. It turned out that an audio taper pot could do the trick.

The problem was that all of the change in voltage output came from the first maybe 10 percent of the pot angle turning, so turning the pot shaft a little would change the voltage by a significant amount, but then as we get past maybe 10 percent of the total angle the voltage changes only very little. This could be very annoying. Luckily using an audio taper pot makes the relationship almost linear.

The chip to be used is the XL4016 buck regulator and the circuit feedback circuit with adjustment pot is shown in the attachment. The curve of angle vs voltage output is also shown, but should be taken to be over the full range of the pot travel not just 93 percent. That was just the number used originally to study this effect.

The output voltage has a light curve to it, but it's hardly anything. The original circuit has a very very non linear curve that swoops way down and then goes almost horizontal.

There are other advantages too. The only caution though is that not all tapered pots are the same. The one being used here is a standard audio taper characterized by having about a 10 percent change in resistance for a complete half turn of the pot shaft. So turning the pot from start to about 50 percent of it's total angle of travel results in only 10 percent resistance change, then the rest of the change comes with the next 50 percent of the travel angle. Using any other taper will not work as well.

Note log(x) in the drawing is actually ln(x).

 

Hello there,

Many of us use voltage regulators but i wonder how much attention is payed to the way the pot rotation actually affects the voltage output of the circuit. Many circuits will have a very non linear turn angle to output voltage character. I found one circuit i intend to work with soon to have this same problem, and the problem is pronounced enough to have made me look for another solution. A solution i did several years back worked pretty well, but it was a bit too complicated for what i want to use this new circuit for, so i was hoping to find a simpler solution. It turned out that an audio taper pot could do the trick.

The problem was that all of the change in voltage output came from the first maybe 10 percent of the pot angle turning, so turning the pot shaft a little would change the voltage by a significant amount, but then as we get past maybe 10 percent of the total angle the voltage changes only very little. This could be very annoying. Luckily using an audio taper pot makes the relationship almost linear.

The chip to be used is the XL4016 buck regulator and the circuit feedback circuit with adjustment pot is shown in the attachment. The curve of angle vs voltage output is also shown, but should be taken to be over the full range of the pot travel not just 93 percent. That was just the number used originally to study this effect.

The output voltage has a light curve to it, but it's hardly anything. The original circuit has a very very non linear curve that swoops way down and then goes almost horizontal.

There are other advantages too. The only caution though is that not all tapered pots are the same. The one being used here is a standard audio taper characterized by having about a 10 percent change in resistance for a complete half turn of the pot shaft. So turning the pot from start to about 50 percent of it's total angle of travel results in only 10 percent resistance change, then the rest of the change comes with the next 50 percent of the travel angle. Using any other taper will not work as well.

Note log(x) in the drawing is actually ln(x).

That's a neat idea. One thing to be careful of is that low quality audio taper pots (of the carbon composition type), lose their ideal curve pretty quickly with use.....the graphite gets "dragged" across the resistor with use....gradually forming a linear pot!
This seems to be more of a problem with newer pots....old ones seem to be better made.

 

That's a neat idea. One thing to be careful of is that low quality audio taper pots (of the carbon composition type), lose their ideal curve pretty quickly with use.....the graphite gets "dragged" across the resistor with use....gradually forming a linear pot!
This seems to be more of a problem with newer pots....old ones seem to be better made.

Hi,

That's interesting i'll have to watch for that. I never used audio pots except for repairs.

 

Not sure if this would work in your application, but here's a discussion of how to make a linear pot look more like a logarithmic pot.

 

Not sure if this would work in your application, but here's a discussion of how to make a linear pot look more like a logarithmic pot.

Hi,

That looks interesting, i'll have to take a close look a little later. I'll post results then also.

 

Not sure if this would work in your application, but here's a discussion of how to make a linear pot look more like a logarithmic pot.

Hello again,

I dont seem to be getting a log pot out of the connections in that link but perhaps you can clarify.

What i do get is an exponential pot which is the inverse of the log pot, but maybe you did something different.

In the link the graph shows the voltage acting logarithmic, but for a 'standard' audio pot we need the resistance itself to be logarithmic. For this audio taper pot the resistance varies approximately like this...
For a 10k pot for example, when we turn the pot from 0 to 50 percent we get a resistance change from 0 to 1k, and when we turn the pot from 50 percent to 100 percent we get a change in resistance from 1k to 10k, which means that last 50 percent of full rotation gave us 10 times as much resistance change as the first 50 percent change did. This equates to a linear scale where the distance between 0.01 to 0.1 is the same as the distance from 0.1 to 1.0.
So if the pot had a 300 degree rotation, turning it 150 degrees would give us R=0 to R=1k, then turning another 150 degrees (to the end of rotation) the pot would give us R=1k to R=10k.

Note i did not look into any other ways to connect this yet.
Also note that there are other tapers but the kind i need is as described above. This may or may not actually be a 'standard' audio taper though.

 

Hello again,

I got the board yesterday and tested it today and traced out part of the circuit in order to determine what mods i would have to make in order to use a pot as described previously.

What happened though is they have it connected the way i thought they would, with the pot as the 'top' resistor and a fixed resistor for the 'bottom' resistor in the feedback divider. Of course that gives a linear adjustment with a linear pot, but there is a bit more risk involved if the pot becomes defective so i did not want to do it that way. If the pot arm disconnects from the substrate material the arm goes 'open' and so there is no longer any feedback, so the voltage shoots up to maximum. So for example with 20v input and 5v output with the pot adjusted correctly, if the pot arm becomes disconnected from the substrate the voltage will jump up to nearly 20 volts. So a load that normally takes 5v will get banged with 20 volts. Of course that's very bad, that's why i wanted to put the pot in the bottom location. I also wanted to use a 10k pot and they are using a 50k pot.

So rethinking this now, maybe i'll just parallel the bottom resistor with another resistor and then set the pot to 50k and parallel that with a 10k pot, which will then give me the range of adjustment i need, and find some other way to limit the output to the max that the load can take without a problem. Maybe a zener in parallel with the pot.

Either that or i'll just have to put up with the non linearity as i also found that the pot i wanted to use (a slide pot) is not audio taper. Too bad.

Interestingly, there also seems to be some disagreement throughout the web on what an audio taper pot really is supposed to do. One source says the voltage output is log, the other says the resistance itself is logarithmic. I dont know if i have one around now to measure.

 

Hello,

Perhaps the following pages on pots might interest you:
http://www.geofex.com/Article_Folders/potsecrets/potscret.htm
http://www.sengpielaudio.com/calculator-voltagedivider.htm

Bertus

 

Hello,

Perhaps the following pages on pots might interest you:
http://www.geofex.com/Article_Folders/potsecrets/potscret.htm
http://www.sengpielaudio.com/calculator-voltagedivider.htm

Bertus

Hi,

Yes thanks, i'll have to read over them today. The first link even goes into the tapering, and the same circuit with the fixed resistor added to the pot.