Correct.
You would need to insert a gain limiting resistor between the differential amplifier stage's output and the volume pot., connecting the two in series. The value of this resistor is to be selected to set the maximum gain of the stage, where the inserted resistor functions as Rin, and the pot. at it's maximum volume setting functions as Rf. So, suppose you wanted to limit the maximum gain to 10x. Since Rf/Rin determines the gain, 10K pot./ 1K Rin = 10x. Therefore, the value of the inserted resistor would be 1K.
Last edited:
Correct, but note that Bruno did NOT do that because it would spoil the perfect tracking of the two channels. This may or may not be an issue in your case with your particular pot - the ideas was to use a regular pot, but there are others (more expensive) that have better tracking so you may need to get one of those in this case.
This is all documented and discussed in the article, it really pays to read it!
jan
You can get "better" tracking by adding extra fixed values resistors around or to a variable vol pot.
If one track has a value of 10030r and the other track has 9980r then the added fixed resistor would not be 1k0, you would choose 1003r for the higher value track and 998r for the lower value track. At maximum gain both channels will have an exact 10:1 ratio and as the vol pot is turned down any small errors in vol pot tracking are ameliorated by the two fixed value resistors. i.e. the tracking at the higher gains is improved by added the fixed value gain limiting resistors.
You can apply this "trick" of improved tracking at the lowest gain (max attenuation) by adding a fixed value resistor in front of the vol pot, instead of getting the two channels reaching-90dB and -85dB at the point just before the tracks short to the end stops one can get much closer tracking by aiming for -80dB for BOTH tracks just as the tracks short to the stops and when the wiper moves off the end stops the error between the channels is reduced by the in front series resistors.
If one track has a value of 10030r and the other track has 9980r then the added fixed resistor would not be 1k0, you would choose 1003r for the higher value track and 998r for the lower value track. At maximum gain both channels will have an exact 10:1 ratio and as the vol pot is turned down any small errors in vol pot tracking are ameliorated by the two fixed value resistors. i.e. the tracking at the higher gains is improved by added the fixed value gain limiting resistors.
You can apply this "trick" of improved tracking at the lowest gain (max attenuation) by adding a fixed value resistor in front of the vol pot, instead of getting the two channels reaching-90dB and -85dB at the point just before the tracks short to the end stops one can get much closer tracking by aiming for -80dB for BOTH tracks just as the tracks short to the stops and when the wiper moves off the end stops the error between the channels is reduced by the in front series resistors.
Last edited:
Andrew, I wonder how that would impact the tracking itself between the end-stops. I haven'really looked at it, but someone should do a list of attenuation versus angle with your numbers, in Excel or something and graph the list. Hopefully it will be linear, but I am doubtful.
jan
jan
Many thanks, I have trouble reading with google translator, I tried to simplify the project even if not very precise between the two channels, only important good music without noise.
I also ask if I can add a tape output of PC doubling in parallel after the two LM4562 op amp TL 072 with pot independent?
Can i use Alp pot? Is a good option?
Otherwise I will use circuit relay Hendrics.
Best regards.
Alessandro
I also ask if I can add a tape output of PC doubling in parallel after the two LM4562 op amp TL 072 with pot independent?
Can i use Alp pot? Is a good option?
Otherwise I will use circuit relay Hendrics.
Best regards.
Alessandro
I read and realised I did not understand B.Putzeys' comment about linearity of the vol pot. I have re-read that comment and the sentences around it, in trying to fathom what he meant and what could be implied from the claimed linearity.
I still don't understand what special property he is claiming for linearity by using a sole vol pot.
Just to clarify. You are happy with his assertion that a lin pot generally is more matched between channels than a log pot? You are also happy with the claim that using the pot in the feedback circuit reduced a potential source of distortion? It's just the statement about law changing you don't like?
FWIW I think his comment was about using a law faking resistor for low volume settings rather than an additional series R to limit max gain. I cannot currently explain why this is an issue, as I've not put thought into the mechanism. Interesting to consider.
Yes, J.Didden restates B.Putzeys' claim is for improved linearity and implies he agrees with that claimed attribute. Then a post later say he has not looked at it.
Yes, and many other have said the same. a dual track linear vol pot has better channel to channel matching than a dual track log law vol pot.
I don't remember reading that, could you explain?
I don't understand this question.
B.Putzeys never used any series connected fixed value resistor in front or after the vol pot. The rather part in the latter half of your sentence does not add up.
I said I didn't understand his claimed "linearity" attribute so I can't surmise what he is actually claiming is improved. Is he referring to distortion? or something else?
Andrew, you read wrong. I have not looked at your suggestion that including resistors to bound the max-min level does not destroy linearity. Not saying that your suggestion is wrong; just have no prove one way or another.
Intuitively, having two pots of unequal value in series with resistors also of unequal value, and then still expecting good tracking with rotation seems to me, let's say, a leap of faith. Look at the pot system as a bridge.
Have you actually read the article and the relevant section? Seems clear to me.
Jan
I have reads all of the excellent article twice and some of three or four times.
But I just don't understand what linearity improvement is claimed.
But I just don't understand what linearity improvement is claimed.
OK.
This meant you have not looked at my suggested modification.
I had misread it to mean you had not tried to analyse B.Putzeys' claimed linearity.
Ok first contentious part of the article below
Second part
Could you pick out all the parts you are unhappy with.
Does it?
I don't agree that adding a gain limiting resistor after the vol pot will immediately put the "linearity of the track resistance back into the equation".
And yet BP says
as if to mean that any external resistor/s must not be added
Or could he be referring back to the diff stage being impervious to linearity errors and that by adding a gain limiting resistor to each channel will reduce the circuit's ability to amplify/attenuate with good/low distortion?
Last edited:
I think he means a parallel law faking resistor. But would need to think that through. It's not obvious to me, but Bruno IS a lot better than me at electron herding.
I wonder if a spreadsheet could show the effect of adding resistors to a pair of slightly mismatched vol pot tracks.
Could that show or compare the no added resistors to the added resistors to show what if any change in channel to channel matching is achieved, or lost?
It would not show any added distortion as a linearity effect.
Could that show or compare the no added resistors to the added resistors to show what if any change in channel to channel matching is achieved, or lost?
It would not show any added distortion as a linearity effect.
Last edited:
That's log faking again.
His circuit in the feedback loop does not use log faking resistors to gain a dB volume scaling effect from a linear law vol pot.
The NFB vol pot has a small gain in non linearity with rotation, but it is far from what one would require of a constant dB attenuation for equal angles of mechanical rotation of a rotary vol pot.
One would get closer to constant dB with rotation by adopting a log law vol pot, but as with a conventional attenuator that brings back in the poorer channel to channel matching.
I see I have used the "channel to channel matching" phrase a few times now.
That makes me wonder if B.P. uses the "linearity" word instead of that?
In amplification lack of "linearity" means added distortion. Maybe that is why I cannot see what he is getting at?
Last edited:
well bruno states 'The only drawback is that the thing gets a bit fiddly at low volume settings' This is normally fixed with a law faking pot. Big jump on my part but I think he is talking about that.
So my view is that a series resistor would add to RiH and not be a problem. But I could be wrong.
So my view is that a series resistor would add to RiH and not be a problem. But I could be wrong.