Lightspeed Attenuator DIY

[georgehifi]

 

I started these Lightspeed Attenuators back in the 1970's, and have used and made them for myself and friends ever since, first ones using Nikon camera light meter photo cells and neon lamps inside tubes.

I'm posting this for DIY'er to do for themselves, please don't make and try sell them for monitory gain, I'm over the clones trying to make a buck out of someone else's IP.     

 

I've tested and listen to just about every concievable passive potentiometer (volume control), Bournes, Alps Blue Velvet, Alps Black Beauty, Penny&Giles, Dact Switched resistors, and transformer (TVC) based volume controls.

Nothing has equalled the transparency and dynamics of sound that the Lightspeed Attenuator,  a series/ shunt LDR arrangement in the attachment, it has no contacts in the signal path, I believe that is the secret.

Also it is the purest way of making it, with nothing but matched quad ldrs in the signal path. No "forced matching feed back circuits" as some clones try to do to make things easier/cheaper to build, as they have all have an effect for the worse on the sound, just the KISS principal, keeps it honest to the source.  

 
The only stipulation that the Lightspeed Attenuator series/shunt LDR  passive needs is, 
1: Source output impedance (cd player/dac ect.) <1kohm (1000ohms)   (which most are)
2: Amplifier input impedance  > 33kohm. (which most are)

 

 

Now let's get to it and build this sucker.

To explain a little the 5 photo attachments.

 

1st: Is the basic led/ldr and how it works in the Lightspeed Attenuator, it mimics a 10kohm series shunt logarithmic volume potentiometer.

 

2nd: Is the Lightspeed Attenuator stereo version, complete circuit diagram, note how the signal goes through nothing but one soldered resistor, no sound robbing light contact of a metal wiper on a resistive track of a volume control (potentiometers) or source switchning  in the signal path. 

 

3rd: Is the same but a dual mono, good for room or system imbalances, (for some reason owners of both say the dual mono has the edge sound wise) I believe it's the fact you can precisely centre the image regardless of recording, environment room or speaker imbalances.

 

4th: Is the buffer Nelson Pass kindy designed for me a few years back for the Lightspeed Attenuator (later to be known as the B1 Buffer) for it to be able to drive low input impedance amps of less than <33kohm, not as good as "no buffer" for >33kohm amps, but almost there.

 

5th: Is a cheap $5 led tester diy'er can use, get on Ebay, (KEMO M087 or M087N) Do not use the 35mA or 50mA range it will damage the leds.

 With this you can quad match your own set of NSL32SR2S(sorted) Silonex led/ldr's you'll need get about 15 pieces to get 2 sets of quad matched led/ldr's for two x stereo Lightspeed Attenuators. So go halves with a buddy and you can have one each.

https://www.digikey.com.au/product-detail/en/luna-optoelectronics/NSL-32SR2S/NSL-32SR2S-ND/5039807

 
Match the LDR resistances using a DMM, powering the LED from 1mA to 20mA, give them time as they need a few minutes to warm up.

 

PS: You can skimp on the cheaper NSL32SR2 but you have to buy almost twice the amount, as their low resistance is not as low and constant, they cannot get to quietest minimum volume level as the NSL32SR2S (sorted) can. 

https://www.digikey.com.au/product-detail/en/luna-optoelectronics/NSL-32SR2/NSL-32SR2-ND/5039808

 

Cheers George

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Edited October 23, 2017 by georgehifi

[Happy]

Looks a little challenging for me but thanks for posting.

 

Also compared to digital volume control? They’re not passive?

[Addicted to music]

I want one that does XLR.

[soundbyte]

I might be missing seeing it, but the circuit of the Lightspeed MKI (?) seems to be missing an earth point on the power supply?

MKII circuit seems to be OK.

[Addicted to music]

12 minutes ago, soundbyte said:

I might be missing seeing it, but the circuit of the Lightspeed MKI (?) seems to be missing an earth point on the power supply?

MKII circuit seems to be OK.

Wouldnit matter?  The psu is to drive the LEDs.  Has no connection to the signal you are attenuating 

 

1 hour ago, Happy said:

Looks a little challenging for me but thanks for posting.

 

Also compared to digital volume control? They’re not passive?

It's not digital, it's all passive and analog sort of speak.  The Intensity of the light from the LED controls the resistive element hence varies your attenuation.

[Happy]

5 minutes ago, Addicted to music said:

Wouldnit matter?  The psu is to drive the LEDs.  Has no connection to the signal you are attenuating 

 

It's not digital, it's all passive and analog sort of speak.  The Intensity of the light from the LED controls the resistive element hence varies your attenuation.

Yeah i know the op started with some comparisons so was curious

i mean those digitally controlled analog vc btw, they’re not passive I guess 

[soundbyte]

2 minutes ago, Addicted to music said:

Wouldnit matter?  The psu is to drive the LEDs.  Has no connection to the signal you are attenuating 

It might, as you need a current through the LED's.

Cannot see a return path for the +5V, so missing a connection between -5V and earth, check the MKII circuit for the -5V.

I do understand how this attenuator works.

[Addicted to music]

8 minutes ago, soundbyte said:

It might, as you need a current through the LED's.

Cannot see a return path for the +5V, so missing a connection between -5V and earth, check the MKII circuit for the -5V.

I do understand how this attenuator works.

See the bottom left: chassis ground and gnd for RCA.

for that  to work you need to tie the gnd to the -Ve  to the -Ve of the LEDs.

[soundbyte]

You are missing the point, you assume everyone will know to make that connection, I do but others probably will not.

So missing a connection, as the circuit as shown will not work until that missing connection is made.

 

[Addicted to music]

2 minutes ago, soundbyte said:

You are missing the point, you assume everyone will know to make that connection, I do but others probably will not.

So missing a connection, as the circuit as shown will not work until that missing connection is made.

 

You are correct, not everyone is like us.  To me it just wasn't a big deal!  

[Addicted to music]

29 minutes ago, Happy said:

Yeah i know the op started with some comparisons so was curious

i mean those digitally controlled analog vc btw, they’re not passive I guess 

The digitally controlled volume control such as a PGA 2310, the signal goes through an opamp, it's the feedback resistor that it's digitally controlling to control attenuation,  so sort of speak no it's not passive.  

[georgehifi]

2 hours ago, Addicted to music said:

I want one that does XLR

I have the circuit for one, but I'm not going to post it, as it will not stay in calibration over an extended period. No matter what's done, unless a feedback type of "forced matching" is used, then the whole thing becomes active and loses it's magic.

 

1 hour ago, soundbyte said:

seems to be missing an earth point on the power supply?

Sorry circuit amended, it was an old one I had lying around.

 

The power supply floats and is completely devoid of the sound circuit to keep any noise out of it, as it only supplies the led voltage and has nothing to do with the signal. 

So an smp wall wart even though I hate them, can work with min amount of switch mode noise intruding into the signal path ground plane, if the power supply negative was grounded this would then have greater noise, from an smp wall wart.

 

Cheers George 

Edited October 22, 2017 by georgehifi

[georgehifi]

Here are also the data sheets for the NSL32RS2S the preferred one to get better feel and performance.

 

Also the cheapskate NSL32SR2 which will work but need more of them to get quad sets from, and also they don't go as low in volume as the "S" version can.

 

Cheers George 

 

 

NSL32SR2.pdf

NSL-32SR2S Data Sheet.pdf

Edited October 22, 2017 by georgehifi

[georgehifi]

On 22/10/2017 at 9:01 AM, Addicted to music said:

I want one that does XLR.

Messaged you the build circuit for an XLR (balanced) Lightspeed, but it will be hard work and you'll have difficulty keeping it in calibration over time. I've tried too many times, all different ways of doing it and it's just not possible without effecting the quality of sound with active matching circuits. Just use single ended unless you run  abnormal lengths of interconnects.

 

Cheers George  

[neixian09]

Guys, 

 

Came across this thread recently.  Anyone knows if there is no power supply to the attenuator, such as 5V  power supply fails, will the volume be half the maximum? 

 

It appears to me so, as two LDR will both be at maximum value.

 

Same as that you can't get zero volume as shunt LDR will never be zero.

 

Best approach to address this, is to trigger a relay somehow, to cut of input. 

 

 

Edited August 4, 2018 by Jing

[georgehifi]

When power is interrupted the volume goes up slightly for a split second then fall away to it's minimum state.

 

Cheers George 

[neixian09]

Thanks George, but I couldn't understand that. It is like two 10k in series when no current to power the led , and the output is in the middle. Wouldn't that be half maximum volume?

Cheers

When power is interrupted the volume goes up slightly for a split second then fall away to it's minimum state.

 

Cheers George 

 

[neixian09]

After reading the datasheet, I guess you are right. 

When power is off, both LDR resistance will be very high, 25M ohms at least, therefore there won't be any output.

 

Asking this because I am thinking use power cut to mute it completely, as zero volume is not possible in your design.

 

Edited August 4, 2018 by Jing

[georgehifi]

6 hours ago, Jing said:

Asking this because I am thinking use power cut to mute it completely, as zero volume is not possible in your design.

Yes this will work, but remember it goes up a little before going down when power is removed.

 

Cheers George