optimum pot value/ ubication?

[federico moreno]

Concerning amps and preamps, I have seen a lot of discussion about pot and attenuators types, values, ubication -sometimes at the input, sometimes at the output of a preamp, for example-, etc.

Now I´m using some sort of integrated amp. The power amp section has an input resistance of 46k. Input cap value is 1,5 uF, Epcos polyprop.

The preamp section -a simple OPA627 noninverting affair with gain 3,- has a 100k input resistor. Input cap is Iskra 1,5uF polyprop. Output series resistor is some 40r.
Dc at preamp output is nonexistent -or at least at the threshold of my testing facilities-; the blocking cap at the input of the power amp section is more of a safety issue than a neccesity.

Previously, I was running the thing with a 50k pot at the input of the preamp section. The sound was good and detailed.
I decided to experiment a little putting the pot at the output of the pre -or at the input of the power amp-.

I´m using an 10k Alps pot, assuming that it isn`t a hard load for opa627, and assuming that it`s output impedance is smaller than
the rest of the pots that I have in my desk.

The change in the sound character was dramatic: On the good side, ¨blacker¨ and more silent backgrounds; and very good dynamics.
On the evil side, I have to turn the pot a good deal -12 o´ clock- to obtain the same volume as with the 50k pot at the input of the pre.
Maybe this issue is related to different law in the logarithmic progression of the pot, so it doesn´t matter really.
But bass is a little strange: more dry, recessed, fast and with an ¨staccatto¨ quality that is very different from the original preamp section.
The difference in sound character is very noticeable; much, much more than a swap of opamps, capacitors or any other audiophile consideration.
And makes me think that topology issues are much more important -and noticeables- than different type of components.

But the real (and difficult) question is the following:
Of the two different types of sound, -pot at the input and pot at the output of the preamp section-,
Which is, in theory, closer to the sound of the source?
I know that the question is difficult to tell, for the source is part of the equation, because when loaded with a pot, it reacts in more or less unpredictable ways.
But, assuming that the source is a good CDP, with decent output stages...
loading it with a pot... is better than load the following preamp?
Maybe the optimal solution for preamp functionallity would be, in this case, opamp/potentiometer/opamp, and then the input section of the power amp. I suppose that sandwiching the pot between two chips acts as an ¨isolator¨, making the effect of the pot less noticeable...
But I wish to use only one opamp per channel-I have no more than two opa627- ...
So, in this scheme of things, what topology -and values of pots- are suggested?
Thank you very much for your input.

[AndrewT]

Hi,

pot at the input and pot at the output of the preamp section

of this pair of options, I would prefer to see the pot after the pre-amp to preserve the maximum voltage through the pre-amp.

putting the pot at the output of the pre -or at the input of the power amp-.

again the better option is the second since it keeps the cable capacitance at minimum after the pot's output (source) impedance.

But you have to decide which of all three options sounds best.

The input impedance of the 10K pot is able to be driven by the opamp or the pre-amp and so is of no great importance.
The source impedance of the 10k pot is a maximum of 2k5 when at -6db , so it is usually important to keep the following capacitance (to ground) to below 1n6F (-3db @ 40kHz). If you can achieve this and keep the next stage input impedance above 25K then either end of the cable is acceptable. But some opamps & pre-amps can struggle to drive 1n6F without showing overshoot, which is the first symptom of reducing phase margin.
This simply means that capacitance should be kept low.

[federico moreno]

Thank you, AndrewT.
The preamp is encased in the same cabinet as the power amp -is some sort of minimalist integrated amp-; the distance between the preamp output pins, pot and power amp input pins is minimal -maybe 15 centimetres-; I haven`t measured the link capacitance, but I believe is very little, some pF´s, no more...
So I think that your nice guidelines are more or less covered.

I have made this question because of the huge difference in sound quality between the two options. It was so noticeable that it made me think seriously about the importance of some basic issues that I tend to take for granted -pot ubication is one of them-; so I always end up spending more time swapping opas or changing capacitors -with audible, but very subtle results-, without priorly experimenting with the basic topology of the thing...
If you ask me what kind of sound I like the most, well... I don´t know yet -I made this modification just yesterday, bored because of a cold and rainy day-... I wasn´t expecting so much difference to be perceived, but yes!
Normally, changing opamps, or putting those bloody things in class a, -or adding or taking a cap or two, for that matter- make for barely audible changes; but we`re talking real difference here -during a brief moment I checked everything to reassure myself that nothing was out of place or miswired, so huge the effect in sound charecter was-...
To put it in other words: this is the kind of difference that even mere mortals can hear. And the kind of difference that some audiophiles -me included- tend to dismiss, so busy as we`re discussing differences between opa627 vintage 1999 and opa627 model 2004...

With the pot at the input the sound is fine, with a little more background noise, but with some hifi-ish sheen and some ¨sameness¨... as if something would always be viewed tru`a tinted glass, to put it in words.

With pot at output, and minimal cabling between pot and power amp section, sounds -even tiny ones-appears to surge from a background of nothing: I think that this is what reviewers define as ¨black background¨ or whatever...
Dinamycs are exceedingly noticeable, and percussive instruments take another dimension.
At first instance the effect is a little surreal; but the opposite approach is surreal too, but in dissimilar ways.

I don´t know. I have to do some additional listening tonite, to see if I`m not fooling myself...

Thanks again.

[lndm]

Hi federico moreno,

I often ask my self the same questions as you are. What you have observed seems to make sense with what I find.

When you put the pot in a small signal area like before the op-amp, you are more likely to have more noise. When you put your pot after the op-amp, in theory your signal to noise ratio should be much better.

This might explain your 'blacker' silences.

As far as the sound changing, I would wonder whether the pot is changing the sound itself. For example, if it is a carbon pot, increasing the signal to it might cause it to produce some noticeable distortion. This would depend on different things but if you are noticing it, maybe you could try a different pot.

I think you might want to keep it after the op-amp if possible.

[federico moreno]

Lndm, thanks for your advise. I will try another, better pot as soon as possible -good pots are something difficult to find here in Buenos Aires; good pot is another issue, hehe-.
This city is getting more and more strange and psychedelic...

Maybe some plastic film Alps, or a good resistor ladder -maybe diy- will do the job. It`s in my plans.
But I think that the signal is not so hot to be a problem there -the pre is currently running at gain 3-; I have a new 5k carbon pot of uncertain origin -I bought it at a surplus store- that I´ll try too... Maybe the pot that I´m using currently has some flaw in the carbon film -It`s a NOS Alps of decent quality, but nothing special-.

Interesting thing would be to check out some other pot configurations, such as shunt resistor...
I always asked myself why so many times, in some particular cases when less s/n ratio degrading solutions are feasible, a pot at the input is viewed as adequate. I think that maybe because is difficult to find low value pots -the vast majority is some 15k upwards-.

Cheers