Lightspeed Attenuator a new passive preamp

[Ben Hanke]

Thanks for your replies, everyone. I'm still reading through it all, but I think it's worth saying that I feel as though it should be possible to rule out any outlandish operating environment presented by my equipment. My loudspeakers are probably the easiest to eliminate because of their unremarkable efficiency rating of 88dB. My amp and CD player, while heavily tweaked for sound quality, should not differ much in terms of their respective gain and output voltage from their original mass-market specifications.

[Ben Hanke]

The resistance of my LDRs measures as follows:

Shunt / Series
Min: 30R / 10k
Max: 10k / 160R

I think by Andrew's math that sets max attenuation at -50dB.

The voltage over the 100R limiting resistors at min volume is 1.6v and 2.0v.

[udailey]

Your amp:
" Line-level input sensitivity is 150mV which may lead to rather low settings of the volume control for some source units"
From Review Marantz PM66 KI amplifier (January 1998) - DutchAudioClassics.nl
I think if you put in the extra series LDRs you will be good.
Uriah

[AndrewT]

Originally Posted by Ben Hanke

The voltage over the 100R limiting resistors at min volume is 1.6v and 2.0v.

This confirms the min volume LED currents at 16mA and 20mA.

Something wrong here.
5V from the regulator.
1.9V across the LEDs and 1.6V across the 100r.
Where has the other 1.5V gone awol?

Can anyone confirm if it is safe to reduce the 100r slightly, to increase the LED current and thus decrease the LDR resistance below the existing measured 30r?

-50dB and 150mV sensitivity is the compatibility problem.

adding a pair of series LDRs will increase your max attenuation to ~-56dB
Or
adding a pair of shunt LDRs will increase your maximum attenuation to ~-50dB
Or
adding both a pair of series LDRs and a pair of shunt LDRs will increase the maximum attenuation to ~-62dB.

It is far easier to increase the impedance of your 100k control pot to increase the LDR resistances to far above the 10k you have measured. Try for LDR~50k or even ~200k.

You can do the equivalent to the high value control pot by instead reducing the regulated voltage from 5V to ~3.5V. This roughly reduces LED current by ~ 50%. Then you will definitely need to reduce the 100r resistors.

[udailey]

I find that below 40R the LDRs start to lose matching so I dont go lower. I cant say for certain what current because I use the resistance of the LDR as my guide. I dont like to see numbers like 30-34. I dont mind 35 and up to much but like to stick at 40 and up.
You could do similar to AndrewT's suggestion and stick a few trimmers after the current limiting resistors for the series LDRs and raise max resistance of LDRs by raising resistance of these trimmers but this will limit your max volume. You could make another LDR voltage divider that never moves, static rather than dynamic LDR resistances. These will all work.
Taking the LDRs up to 50-100k is possible but guaranteed to not match any longer. If you dont mind using a balance pot this is fine.
I think the easiest to try right now is reduce voltage and see if it solves your problem. You should still have same or real close to same min resistances this way.
Uriah

[Ben Hanke]

Great, thanks guys - I have some things to try, and I've got a much better understanding of the problem. Nice find w/ the review on the KI input sensitivity, btw.
A lower max volume is not a problem as I never go past 12 o'clock anyway.
I'll let you know how I get on.

[udailey]

NP.
Sweet CDP by the way. Tried finding specs on it but the glowing reviews took up all the google hits

[georgehifi]

Originally Posted by Ben Hanke

The resistance of my LDRs measures as follows:

Shunt / Series
Min: 30R / 10k
Max: 10k / 160R

I think by Andrew's math that sets max attenuation at -50dB.

The voltage over the 100R limiting resistors at min volume is 1.6v and 2.0v.

Look at the Silonex data spec sheets. This is why I use quad matched sets of the NSL32SR2S, to get lower min volume to get better logarithmic feel to the volume and better channel balance.

On resistance of a NSL32SR2S is typically 40ohms @20mA
On resistance of a NSL32SR2 is max 40ohms @ 20mA, typically much higher >100ohms
On resistance of a NSL32SR3 is max 60ohms @ 20mA, typically much higher >100ohms

Cheers George

[AndrewT]

Originally Posted by georgehifi

..............I use quad matched sets of the NSL32SR2S, to get lower min volume

George,
Can you show your data to justify your claim that "quad" matching achieves "lower" minimum volume.

Basically, I put it less diplomatically, I don't believe this claim.

[georgehifi]

Andrew:

Like I have said all along more times than I care too now (to put it less diplomatically), you will get better lower levels if you use the NSL32SR2S, and when quad matched you will also get better logarithmic feel to the volume instead of nothing at the start till you get to 12 o'clock or too much at the start and can't get passed 8 o'clock. And also more stable impedances which I believe gives a better more consistant sound at different levels also, over the NSL32SR2 or the NSL32SR3. As you have typically 40ohms min value with the NSL32SR2S where the others have well over 100ohms and if anyone pushes these harder than 20mA you will over time end up with a dead led/ldr and that's a given.

Cheers George