Lightspeed Attenuator a new passive preamp

Thanks for sharing the design of this George.

I've gotten the itch to solder something and I hear that the pre on my new(ish) C370 is slightly weaker than the power section, so I'm going to build myself one of these.

I'm a little concerned about the impedances being out of spec. I may have to look at building a buffer but I wondered if anyone could advise on whether it's worth a try without or is guaranteed to need it.

Input will be a Cambridge 740C - <50ohm
Power will be NAD C370 - 20kOhm

I'm thinking perhaps having a low impedance on the input might compensate a bit for the low on the out?
 
Last edited:
Just thought you DIY'ers and owners of the Lightspeed Attenuator would like to know, that it has made it again into 2013 Octobers Stereophile Recommended Components, with the highest $$$ value rating once more.

Cheers George

I had actually seen this on the week end. Congratulations again George and thank you for allowing the DIY crowd to gain access to this. We are, thanks to you enjoying sound that many reviewers compare against multi k$ boutique offerings.

I went looking for the 2010 review in Stereophile and could not find it. Do you have a link to that article?
walter
 
I went looking for the 2010 review in Stereophile and could not find it. Do you have a link to that article?
walter


I have a draft of it Stereophile sent me "for my eyes only", but as it has never been posted on the Stereophile website for all to read free of charge, I'll need to ask them permission to post it up here.

Maybe they just get a lot of back issue sales of that copy still, so they wait till sales drop off before they put it on the web free to read.
Makes sense to me, it's called maximising back issue sales.

Cheers George
 
My dual pot wich is rated 100k is having 87kohm on one channel and 95 kohm on the other one. Does it matter wich section goes for shunt LDR's and serial ones ?

Would be a bad idea to use only one section of the pot connecting +5v on middle pin and extremes to shunt/series ldr's ?

I measure min 1.4V and max 1.89V across diode as i'm rotating pot from one extreme to another. Is that range ok ?
 
Last edited:
No Atupi, it won't matter with a dual mono setup if one pot is 87k and the other 95k.

Looking at the back of the pot with a dual mono pots. I take the +5v to the center, pin, the series to the right pin and the shunt to the left pin.

You mean after the 100ohm resistor? That's about right from memory.

And yes the Lightspeed Attenuator mimics a 10kohm log passive pot with it's i/o resistances.


Cheers George
 
If you measure the ouput impedance of a 10k log potentiomer at different positions you will see, it's close to the Lightspeed, that is if you have used quad matched NSL32SR2S ldr's.
Even if as you state a 5k measurement that then is better to drive into lower input impedance amps say at say 30kohms. But then your source is looking at a lower load impedance and needs to be low output impedance.
I can't see any problem for you so long as your source is not high output impedance as some silly tube output dacs and cdp's are

Cheers George
 
May be old news or buried somewhere in this thread but has any effort been made on optimization of supply voltage to the LDRs? I see (very) early posts by George as well as a schematic indicating 12V being reduced to a regulated 5V.

What about lower voltage, e.g., 2.5-3V? Or higher than 5V?

Thanks.
 
May be old news or buried somewhere in this thread but has any effort been made on optimization of supply voltage to the LDRs? I see (very) early posts by George as well as a schematic indicating 12V being reduced to a regulated 5V.

What about lower voltage, e.g., 2.5-3V? Or higher than 5V?

Thanks.

Hi Big Guy, as far as optimizing the power supply we have with a group of "golden ears" done a/b testing. Powering just by battery vs 12vdc wall wart, and out of 20 or so "golden ears" only a couple "thought" they heard an improvement with pure battery power, but they couldn't say what it was.

The noise on the output with input shorted of the Lightspeed Attenuator using the 12vdc linear wall wart then through the secondary 5vdc regulated supply as I have outlined, was a miniscule 20uV, which was the base line of the Tek cro as well. So it was not noise the couple of "golden ears" thought was the difference they heard. (it may have been audio voodoo)

Sure if you start to use smp wall warts then they are more noisy than linear wall warts, but after being smoothed and regulated for a second time by the internal 5vdc regulated suppy, even good quality smp wall warts are exceptable.

And it won't make a difference if you reconfigure for lower or higher inital voltage as it all comes down to how many mA the led part of the NSL32SR2S package can take. I picked 5vdc as the regulators are common and easy to get.

Cheers George
 
Last edited:
Thanks for the info on supplying power. I figured with all these posts, someone must have investigated.

I understand that the design is at 5V but, not knowing much about LDRs, what if voltage supplied, e.g., batteries connected directly to LDRs, was 2.5+/-V or 7+/-V rather than the 5V? Would they not operate, etc.?