Lightspeed Attenuator a new passive preamp


Hi SCM, I just had a look at my records and as I thought, your unit was from a Stephan Tremblay from Canada. And it is at least third hand now as he bought it second hand and was heavily modded by the owner/s that were ham fisted with a soldering iron.
I did replace all the quad matched LED/LDR's which were potted and soldered into 2 x 8 pin DIL sockets which also had to be replaced and then it had to be recalibrated at a labour and parts cost of $120.

This was the email I got back from the Stephan after I repaired it.

"Hi George,
It arrived safely yesterday. I just want to thank you for helping me with the Lightspeed. It now sound so much better. The sound is fast & clear with tons of details. The soundstage is huge and the balance is well centered! The music is just flowing...What a difference!!!
I am very happy that I can enjoy it properly. This is a awesome product and I can listen to 12 or 1 o'clock with plenty of gain.
My setup: Macmini > cEntrance DAC (25 ohms) > Lightspeed > Yamamoto A-08s (200K input impedance) > Klipschorns
Best, Stephan (from Canada)"

Two weeks later, he was having problems again.

" Hi George, Greetings from Canada. Sorry to ask for your help again but since yesterday I am only hearing a little bit of music throught my LS. Almost no vocal and no power whatsoever! I have always triple checked every connection with the RCA in/out wiring to make sure I was not doing anything wrong."

I had the feeling that maybe there might be some big dc offset from the source or back from the poweramp that was cooking the led/lrd packs. As I have said above first change the 100kohm dual log pot I suspect this now.

Cheers George
 
Well George you are absolutely correct.
I just received it, now I do I go about getting the cover off do I work from the front, the back or do both end plates have to come off ? I want to post pictures of the insides with your approval of course.

Steve

You have P/M re dismantling procedure.
This is the pot you'll need to buy. Please do not turn any of the calibration pots inside this unit has been calibrated to the new quad set of led/ldr's, hopefully they are still OK.
http://au.mouser.com/ProductDetail/...GAEpiMZZMtC25l1F4XBU8%2bWhIiv2u5sp8VJhNJL8jA=

PS: Break of the little lock tab on the new volume control, as it will not sit flat if you don't.

Cheers George
 
Last edited:
Can you imagine how much better his sound would be if he used a Lightspeed Attenuator for each control!

Feeling cheeky, had my double shot of caffine this morning
Cheers George
 

Attachments

  • He needs Lightspeed Attenuators.jpg
    He needs Lightspeed Attenuators.jpg
    119 KB · Views: 486
Bits of solder spatter here and there...one component in the middle of the board looks like two out of four wire connections are blown/melted apart.

Not too encouraging at this point :(

I have asked scm not to post pictures of this unit as it was heavily f****d with, as is now not representative of how a Lightspeed Attenuator looks, and I would not like a picture of this circling around on Goggle images.

The spatter is remnants of all the mods this has gone through don't worry about that, the little blue trim pot is for calibration don't touch that, and the 4 loops of wire are for calibration also and dictate which channel needs calibration to the other, two are open circuit and two are closed don't touch these either.
The front panel 100kohm dual log pot is what you first need to change as this is the maybe the culprit and the easiest thing to do, I just hope the 4 matched led/ldr's are ok if not then your seller (steven) had a serious problem with his system that has done damage to them again.

Cheers George
 
Last edited:
Thanks George. So at each corner of the blue trim pot only 2 out of the 4 wires loop back into the board then ( being closed) ?
Just to clarify, 2 of those wires located on opposite corners of the trim pot are by the looks of them burned through (being open)..if that`s by design then ok I`ll move on to the dual log pot.

Any chance the log pot may be found at Radio Shack ?

Thanks
Steve
 
Thanks George. So at each corner of the blue trim pot only 2 out of the 4 wires loop back into the board then ( being closed) ?
Just to clarify, 2 of those wires located on opposite corners of the trim pot are by the looks of them burned through (being open)..if that`s by design then ok I`ll move on to the dual log pot.

Any chance the log pot may be found at Radio Shack ?

Thanks
Steve

That's correct by design, only two diagonal of the 4 links are connected for calibration purposes, to the untrained they may appear burned, but that is solder that's seen.
As for the pot the only good one I've seen from the US that has the correct pot shaft and diameter is the Mouser I've linked to in the above post. The voltage to the pot should be 5vdc measured from the negative side of the power supply.

Cheers George
 

Attachments

  • Lightspeed Attenuator MkII Circuit.JPG
    Lightspeed Attenuator MkII Circuit.JPG
    72.3 KB · Views: 446
Last edited:
If the ldr dies,is it full volume or 0? I'm expecting to get it as 0 to avoid 'BOOM' on amplifier and speaker

Regards

If you make it as I have posted, if power is pulled the volume goes up just a touch then fades away.
If the shunt only ldr goes open circuit, (I have never seen this) yes you will get max volume, this will only do that if you over power them, mine run below max rating.

Cheers George
 
I'm trying to experiment with the Silonex's reference circuit from figure 10: Advanced Photonix Canada, Inc.: Technical Reference: Audio Level Control with Resistive Optocouplers.

I'm using an Arduino board with:
- PWM 8bits signal (5V) for the control
- 2.048V as an input
- the output is plug to an input of the Arduino so I can measure the ADC value in my program

Do I need to put a kind of load on the LDR side of the Silonex device to simulate an amplifier? ...somethime I can get low as 48mV on the output and most of the time around 110-130mV

Another thing is ~3.6V the optimum voltage for the OP1 (from the series/shunt), instead of directly using 5V ?

Otherwise I'm getting a slightly S shape for the output response curve, so mainly linear in the most part of the curve. This is with a matched pair of LDRs. At the moment I'm just playing with a single channel.

Just a side note I'm getting my first output from the 60th value of the PWM control input; is it because that's the difference between 5V and 3.6V?

@george: I hope you won't mind; it's still a series/shunt implementation... :p
 
Cedric, I've looked at that page before; my impression is that the circuits are nice in theory, but they aren't adequate for a stereo potentiometer -- to control four different LDR devices with precision and minimizing the effects of temperature on the device.

Here is an article on LEDs that I found useful -- it explains the relationship between voltage, current, temperature, and light output in an LED, and I feel that one really needs to understand this relationship in order to make progress toward creating an effective control circuit.

LEDs Magazine - Driving LED lamps - some simple design guidelines
 
To address the S shape I think you would have to build the circuit almost twice. Basically build a circuit for the series and another for the shunt and program the arduino to give X frequency to Series when Y is given to Shunt or something similar. Maybe a lookup table so that it measures the frequency given to one LDR and then looks up what the frequency should be for the other LDR. If you can program this would be a good way to do it. An even better way would be a circuit that builds this table automatically. Fun!
Wish I could help more but programming is not my strong suit.
 
wapo54001 said:
Here is an article on LEDs that I found useful -- it explains the relationship between voltage, current, temperature, and light output in an LED, and I feel that one really needs to understand this relationship in order to make progress toward creating an effective control circuit.

LEDs Magazine - Driving LED lamps - some simple design guidelines

Thanks for the article.

udailey said:
To address the S shape I think you would have to build the circuit almost twice. Basically build a circuit for the series and another for the shunt and program the arduino to give X frequency to Series when Y is given to Shunt or something similar.

Yes I've already tried with the following diagram (which we already talked via PM) but I couldn't get a descent response curve which I could reproduce/follow. The LDRs are not precise over time! If you apply the same voltage at two different periods then you are getting two different resistance values which are maybe close but not enough to have a match with a margin of errors. Basically the response curve drift over time to the left or the right...
Maybe I'm a bit too picky and I should "trust" your matched LDRs and use the same voltage for both channels; doesn't matter if the attenuation is different over time! :)

udailey said:
Maybe a lookup table so that it measures the frequency given to one LDR and then looks up what the frequency should be for the other LDR. If you can program this would be a good way to do it. An even better way would be a circuit that builds this table automatically. Fun!
Wish I could help more but programming is not my strong suit.
I'm a software developer so I'm OK with that, it's just with the hardware which I'm lacking a bit... :dunno:
 
One thing I was wondering, but have not gotten around to testing, is how the resistance varies with input signal amplitude. Has anyone explored this? The last time I measured, it was quite well behaved over the audio frequency range.

You find a tighter tolerance for the NSL32SR2S (selected), this is why it's a bit more expensive but worth it, it gives typical 40ohm min to 5mohm max.
Where the NSL32SR2 they won't give the typical low figure it can be anywhere with a max to 5mohm.
And the NSL32SR3 is even worse at typical 150ohm min to 25mohm max.

Cheers George