converting input pot resistance to gain (dB) [LA-2A]

Hi all,

with no peak reduction applied the LA-2A can serve as an amplifier with up to +45dB gain.
the amount of gain applied is determined by the input pot, which is a 100k log 'fader' type.
I am wondering if there is any sort of formula that allows me to convert input pot resistance to volume gain?

The idea behind this is to replace the input pot with a rotary switch and fashion a neve-1073-style gain control, with some line and some mic settings. I am thinking of having a fairly fine resolution in the line section (-5 to +10dB in 1dB steps) and then a coarser resolution for the higher gain settings (+15dB to +45dB in 5dB steps). But how do I calculate the required resistors?

or maybe Harpo has already done a spreadsheet for this sort of thing? 

Cheers
Briomusic

This one has proved its usefulness over and over again: http://homepages.tcp.co.uk/~nroberts/atten.html

I've done exactly this for my LA2A. Accuracy is great since the line amp of LA2A is feedback based and doesn't need matched tubes. I don't recommend this for the peak reduction. I did it initially but quickly found out the T4B's are almost impossible to match and you need the variability of pot.

Talking about a stereo unit of course.

Set the amount of steps and then select the exact attenuation in dB steps. I used 1-3dB steps to spread out the stepped attenuator over 60dB range. Not much use for attenuation below that. Can't remember the exact distribution of the steps anymore, but that's not critical. Maybe get a feel of the most important area of attenuation with a pot first, and use finer grained dB steps there.

briomusic said:

or maybe Harpo has already done a spreadsheet for this sort of thing? 

Click to expand...

I'll ask him when he shows up next time .. ;D

as Kingston said

wow, thanks to the both of you!

I am going to ponder the "peak reduction" problem, maybe implement an additional 'fine'-pot on one of the channels to achieve perfect stereo matching. I would just like to have total recallability and a lot of times the compressor will be used in dual mode, not even linked.

briomusic said:

I am going to ponder the "peak reduction" problem, maybe implement an additional 'fine'-pot on one of the channels to achieve perfect stereo matching.

Click to expand...

The problem is the highly non-linear opto-resistive element. One channel will compress low frequencies more than the other etc. so the peak reduction setting will be very program dependent.

In practise this is not a problem but it means that a fine/trim will not help much.

so if I was going to replace the peak reduction pot anyway - just for a try, how would I calculate the values for the resistors on THAT one?
I am sure I can reuse Harpo's spreadsheet, the top dB value in that case would be 0 (no attenuation) but what do I put in for the other 23 values?

briomusic said:

so if I was going to replace the peak reduction pot anyway - just for a try, how would I calculate the values for the resistors on THAT one?

Click to expand...

You would have to know the frequency and level depending threshold value when the EL-foil begins to glow and thereby decreasing the value of the LDR in the T4B cell.

I am sure I can reuse Harpo's spreadsheet, the top dB value in that case would be 0 (no attenuation) but what do I put in for the other 23 values?

Click to expand...

I am sure I won't use Harpo's spreadsheet for this purpose, but YMMV.
'The top dB value in that case would be 0 (no attenuation)' looks misleading. No attenuation in front of the sidechain amp probably will light up the EL-foil, so the LDR value goes down with the result of decreased program signal. This will be interacting.

briomusic said:

so if I was going to replace the peak reduction pot anyway - just for a try, how would I calculate the values for the resistors on THAT one?
I am sure I can reuse Harpo's spreadsheet, the top dB value in that case would be 0 (no attenuation) but what do I put in for the other 23 values?

Click to expand...

You seem to neglect the virtues of experimentation. Spreadsheets are for bean counters...