Input volume control.......why not linear?

[ToneMerc]

Let me first preface this in the general context of a Dumble circuit topology.... Now considering that the VOL control is generally kind of set and forget, minus maybe +/- a knob position, does it really truly matter if the core taper of that 1M pot is linear or log?

If you toss out all the auditory theory of how humans actually perceive volume changes, the only the tangible issue left is the position of the knob right?

Thus, does it actually really matter considering the way that the knob is normally used?

Any thoughts................

TM

[Structo]

I always wondered why a linear pot seems to jump in volume all of a sudden where a log or audio pot seems more gradual in volume increase.

Seems backwards to me, because if a pot is linear shouldn't it be a gradual increase?

Semantics aside, for set and forget I'm sure a linear pot works just as well.

[ayan]

I always wondered why a linear pot seems to jump in volume all of a sudden where a log or audio pot seems more gradual in volume increase.

Seems backwards to me, because if a pot is linear shouldn't it be a gradual increase?

Semantics aside, for set and forget I'm sure a linear pot works just as well.

Think of it this way: the log of a number (greater than 1) is always smaller than the number itself. If a curve follows a number (could include a multiplying factor) we call it a linear relationship; if it follows the log of a number we call it a logarithmic relationship. The log function will always lag, and so the log pot will always "lag" with respect to the linear pot as well. Hpe this makes sense.

Gil

[Bob S]

I get a lot of comments / unease from guitar players using my TW type clones running the Volume at 10 or 12 o clock...
I've used pots with a series resistor to to allow them to use the Volume maxed. Whatever floats your boat I suppose.
Perceived volume isn't directly proportional to power/wattage.
Log pots are better at tracking perceived volume.
Unless you are hearing impaired, like me.
Loud is loud - be happy I say.

[Structo]

Thanks Gil and Bob that does help me to understand it better.

[brentm]

[img:751:571]http://www.geofex.com/article_folders/p ... ttaper.gif[/img]

Source: http://www.geofex.com/article_folders/p ... tscret.htm

In my dealings with subbing a linear pot where an analog pot taper should go, I find it hard to control the usable range of volume. So on a pedal I'm mentioning, the slightest tweak puts me either too loud or too soft... it's hard to find that "just right" setting.

[Structo]

I've seen that curve plot many times.

Now with the explanation by Gil and Bob it all becomes clear.

The Log or Audio pot has a much slower taper until about 75% rotation.

And with modern Audio taper pots that are 10-15%, it would even be longer.

With the Linear pot, it is a 50% taper meaning at half rotation it is 50% of the total value. (in theory)

[d95err]

Let's do some math! (Well, I used an online calculator to spare us most of the math, this is mainly the results...)

Example for a linear pot: (We assume that it's 100k and has 300 degrees of travel, i.e. 1k per 3 degrees).

If the pot is near the bottom at 99k serial resistance and 1k parallell, we have about 40 dB attenuation. Turn it up 3 degrees (98k serial, 2k parallell) and we get about 34 dB attenuation.

Now turn the pot to near max. With 2k serial and 98k parallell resistance, we get about 0.18 dB attenuation. Turning three degrees to 1k serial, 98k parallell, we get about 0.09 dB attenuation.

Summary:
Near bottom: 6dB change for 3 degree movement
Near max: 0.09dB change for 3 degree movement

That's a huge difference! A (truly) logarithmic pot would produce the same dB change for the same number of degrees of movement across the whole range of the pot.

[martin manning]

Hejsan Erik,

This is a very good way to think about pot tapers. I made some numbers for myself to plot attenuation for true log, linear, and the bi-linear taper that is common in real "audio taper" pots. In the plot the rapidity with which the linear pot comes up relative to the log is clearly shown. The bi-linear comes up faster than a true log curve, and also has a jump where the slope change occurs.

[roberto]

I though about it some years ago, and got to the decision of using log pots for gain too. The main point is not that you'll have the pot in one position or in another, the main reason for me is the control capability of a certain range of sounds: with a lin pot you'll have 30° (0-1 on the knob) of control instead of 150° (0-5 on the knob) where the amp usually has more nuances: up to 10% of the signal.