PEC pot imbalance

[mortenpedersen]

Recently installed a PEC pot in my stock crack 1.1 and there seem to be a veeery slight imbalance - barely audible.

There is a slight difference on upper and lower middle lugs when measuring the input volts from a 60hz test tone. Also measuring the resistance on the two lugs reveal the same discrepancy. Both readings are in favor of what I also believe I'm hearing.

In fact pulling a brand new pot right out the bag and directly measuring resistance between input and output, reveals a difference on the two decks.

Guess my question is if it's normal for this type of volume control? And would padding the pot as suggested in the Crack FAQ #3 fix this, or do I have to opt for a stepped attenuator type for perfect balance?

Thanks, Morten.

[kgoss]

Measure at 50% volume. If the imbalance is gone padding to get you in the middle of the pot will work.

[mortenpedersen]

Well, I measure a difference at any given volume. Also, I'd like to find a better solution than trying to hit a useable range by padding, if I understand you correctly.

Morten.

[Doc B.]

Yes, you have to opt for a stepped attenuator with close tolerance resistors for "perfect" balance. But padding the existing pot is not really that much less optimal if it gets the pot into a range where balance is good. For me what determines optimal has to with with what sounds best as much as how well the pot tracks. The Pec is a good sounding pot, and probably worth adding the pad to. If you opt for a stepped attenuator you could end up with one that sounds better, or one that sounds worse.

[mortenpedersen]

Thanks, looking at the readings I might be better off leaving it unpadded. The difference is smaller around 25% volume, which I actually find a little strange - here are some resistance readings on the middle pot lugs:

VOL        L          R        DIFF
25%      1.5       1.5       0%
50%     12.9     12.6     0.4%
75%     54.3     46.3     4.3%
100%   96.5     96.5      0%

It kinda jumps in diff in between those levels, so could be a toss-up

[Doc B.]

That is one part of the puzzle. To get the most useful result, put a mono, steady tone through the amp and measure the signal level of each channel's output at different pot level settings. There are other components in the circuit that can also influence the balance, and what you actually hear may not correlate precisely to those measurements. For example if one half of a tube has a bit less gain than the other, or resistor values elsewhere in the circuit are slightly different between channels. All these small differences can skew the results a little.

[troplin]

Thanks, looking at the readings I might be better off leaving it unpadded. The difference is smaller around 25% volume, which I actually find a little strange - here are some resistance readings on the middle pot lugs:

VOL        L          R        DIFF
25%      1.5       1.5       0%
50%     12.9     12.6     0.4%
75%     54.3     46.3     4.3%
100%   96.5     96.5      0%

It kinda jumps in diff in between those levels, so could be a toss-up

How did you calculate those difference percentage values?

[Paul Birkeland]

VOL        L          R        DIFF
25%      1.5       1.5       0%
50%     12.9     12.6     0.4%
75%     54.3     46.3     4.3%
100%   96.5     96.5      0%

A percentage doesn't equate to what you're hearing.  At the worst position, you have 34.6dB of attenuation on one side and and 33.3dB on the other.  That's about 1dB, it's audible, but barely, and frankly at 75% volume, you likely have so much signal coming out of your amplifier that it's unlistenable.

If you are able to hear a channel imbalance at 50% rotation, you have a problem that is not with the pot.

-PB

[mortenpedersen]

How did you calculate those difference percentage values?

Just calculated it based on the two readings - L / 100 * (L-R).

If you are able to hear a channel imbalance at 50% rotation, you have a problem that is not with the pot.

And it may all be in my head I thought I could hear a difference and when the measurements indicated the same I got curious. I don't have the experience to put the impact into perspective, so thanks for that. Still from measuring the pot alone, it seems like my best option is to leave it unpadded.

And Doc, thanks for the tips on how to investigate further, I'll look into that.

Thanks, Morten.

[troplin]

Just calculated it based on the two readings - L / 100 * (L-R).

The correct formula is:
100*(L-R) / L
which will give you higher difference values, i.e. 2.3% diff at 50% vol and 14.7% diff at 75% vol.

[mortenpedersen]

Thanks troplin, should've looked that up

[diynewbie]

As Paul pointed out, you really want to compare the attenuation of the two channels in dB.  I assume you measured resistance at the center lug.  The formula for attenuation of a channel would be 20*log(R/R100%) - e.g. 20*log(12.9/96.5).

If the resistance at 100% is the same, the difference in attenuation between the two channels can be calculated directly using the formula 20*log(RL/RR) – e.g. 20*log(12.9/12.6).  The louder channel will have the higher resistance.

Paul, please correct me if I got it wrong.

Doc has pointed out that most people will have little trouble hearing a 1 dB difference in volume.  at 0.5 dB it gets harder.  Here is a website where you can here different dB shifts in volume for a test tone:  http://www.audiocheck.net/blindtests_index.php

The above will work for most pots, I've read that the TKD 2511 is an exception.  A more accurate equation would be 20*(log (Rout to grnd/(Rin to Rout + Rout togrnd)).  But I am not sure if this works for the TKD pot.  The sure fire way is to play a 150hz or thereabouts test tone and measure input and output AC voltage; using the equation 20*log(Vout/Vin) to calculate attenuation for each channel.

[mortenpedersen]

Great great info diynewbie, I'll look into that.

Thanks for all the help guys, Morten.