Ladder-type Pot, Coupling Cap, and Input Resistors, for a Gainclone

[mudihan]

I am thinking about getting a 25k ladder-type 24-step volume control. I went to this website to calculate the resistor values:
http://homepages.tcp.co.uk/~nroberts/atten.html
I put the 25k and 24 steps there, and got the results for Rx and Ry. Looking at it, I realize that Ry (signal to the ground) is in the range of 0ohm-5142ohm.

Now, my coupling caps (to get rid of the DC) is 10uf films and my gainclone has resistors Rin (signal to the ground at the input) valued at 25K (I may increase it to suit the volume control) . The cap was in the CD player. However, since Ry in the ladder volume control are of very low values, for the sake of good bass, should I move the cap inbetween the volume control and the amp so that the HPF will be based on the 10uf cap and the 25k input resistors Rin?

[mudihan]

Just to clarify:

In the CD player, there is nothing after the coupling caps in the signal path.

The new ladder-type pot will be used as a passive volume control.

By the way, what's the input impedance of LM1875 and LM3875? Just in case that I want to omit the Rin (signal to ground).

[AndrewT]

Originally posted by mudihan
Just to clarify:

In the CD player, there is nothing after the coupling caps in the signal path.

The new ladder-type pot will be used as a passive volume control.

By the way, what's the input impedance of LM1875 and LM3875? Just in case that I want to omit the Rin (signal to ground).

the last question gets a question first.
What topology are the chipamps using? inverting/non-inverting

A schematic of your assembly would help.

[AndrewT]

Originally posted by mudihan
I am thinking about getting a 25k ladder-type ..... volume control................ the 25k input resistors Rin?

A 25k attenuator followed by a 25k input impedance is not well suited for a passive pot (nor stepped attenuator).

I would suggest a 10kpot feeding a 50k Zin when used as passive pot.
or 5kpot feeding a 25k Zin. or 20kpot feeding 100k Zin.

The 5kpot will drive reasonable lengths of interconnect.
As the pot value increases, the permissible capacitance (from cables and RF filter) reduces drastically. A 25kpot may require direct connection to the Power Amp input filters.

BTW,
5142r ~=5k6//62k=5136r

[Peter Daniel]

If the gainclone "has resistor Rin (signal to the ground at the input) valued at 25K" that would indicate non inverting topology.

In such configuration you cannot omit Rin, nor increase it much over the value of feedback resistor, as DC offsets at the output will become very problematic.

For best results, input coupling cap should be between the pot and Rin.

[mudihan]

Originally posted by AndrewT
the last question gets a question first.
What topology are the chipamps using? inverting/non-inverting

A schematic of your assembly would help.

Thanks for the reply, Andrew. It's Non inverting. The Schematic is not much different from the standard dual supply one on the National datasheet.

Originally posted by AndrewT
A 25k attenuator followed by a 25k input impedance is not well suited for a passive pot (nor stepped attenuator).

I would suggest a 10kpot feeding a 50k Zin when used as passive pot.
or 5kpot feeding a 25k Zin. or 20kpot feeding 100k Zin.

The 5kpot will drive reasonable lengths of interconnect.
As the pot value increases, the permissible capacitance (from cables and RF filter) reduces drastically. A 25kpot may require direct connection to the Power Amp input filters.

BTW,
5142r ~=5k6//62k=5136r

Yes, a 25K pot feeding a 25K Rin is not technically ideal in most situations, but I will mount the pot inside the amp. Will that solve the problems? I remember people saying that the value of the input impedance (Rin) should be 10x the value of pot, but I believe that they said it for the same reasons you listed. In other world, if pot is within inches of the amp input, can the value of the pot be as high as Rin without having any kind of negative sonic effect?

Originally posted by Peter Daniel
If the gainclone "has resistor Rin (signal to the ground at the input) valued at 25K" that would indicate non inverting topology.

In such configuration you cannot omit Rin, nor increase it much over the value of feedback resistor, as DC offsets at the output will become very problematic.

For best results, input coupling cap should be between the pot and Rin.

Thank you for the reply, Daniel. Yes, I omitted the Rin once and burned my speakers. How high do you think the value of Rin can go without too much effect on DC offsets at the output?

A few more questions:
If I move the coupling cap inbetween the pot and Rin, would the DC coming out the CD player do anything sonically when passing through the resistors in the pot?

If the output stages in my CD players consist of OPA627s or AD843s (My favorite ones, well decoupled, but nothing more), would a 5K ladder-type pot be too hard a load for them? If so, what would be the minimum load for them to produce optimal quality? Thanks!

[AndrewT]

Originally posted by mudihan

If the output stages in my CD players consist of OPA627s or AD843s (My favorite ones, well decoupled, but nothing more), would a 5K ladder-type pot be too hard a load for them? If so, what would be the minimum load for them to produce optimal quality? Thanks!

the datasheets should give a good indication of minimum load impedance and also a guide to best load impedance.
It may drive 600r with some loss of performance.
It may drive 10k without any loss of performance. It may show that there is very little deterioration when reducing the load impedance down to 2k.
Read and interpret. The come back with specific questions and links to save us doing all the work.

[mudihan]

Originally posted by AndrewT
the datasheets should give a good indication of minimum load impedance and also a guide to best load impedance.
It may drive 600r with some loss of performance.
It may drive 10k without any loss of performance. It may show that there is very little deterioration when reducing the load impedance down to 2k.
Read and interpret. The come back with specific questions and links to save us doing all the work.

I actually examined the OPA627 datasheet before posting, and I just reexamined it. While many figures were given in regard of RLoad>/=1k, there is no graph (such as a voltage swing vs load resistance one) that plots the load resistance against something else. The application note does not mention load resistance either. Should I assume then that the RLoad should simply be greater than 1k? Thanks!