Optical Passive Attenuator

[Mike H]

Wot it does ~

pink = "light power" of lamp (an extra node enabled on original lamp model)

green = Voltage at output U1 (otherwise about 90% on R1)



 

[Graeme]

Yes, i have been using an LDR attenuator.

Its damm good.

Amazing how much better than the alps blue it is.

Unfortunately, it hasnt been too reliable. First time it just went a bit odd but dropping voltage to the LED's solved it.

This time either my amp has fried it, or its fried itself, not really sure.

I may buy/build another LDR or i may build a stepped attenuator.

Im reluctant to leave LDR's though as it was such a big leap in sound quality.

Lower noise floor, less distortion (didnt even know i had any untill it was gone), better seperation, better imaging, more space around everything etc etc.

[Mike H]

Yeah the Alps blue is not that terrific actually, structurally well made though must be said. But hence the "shunt pot conversion" way of using it like I do at the mo.


Think it was Ed or somebody posted some while ago "manages to sound both hard and dull at the same time, which is quite a feat" if I remember correctly!


 

[Mike H]

Voila ~

Gwarn be the first to make one

(I had to take out the lamp model cos it was too complicated and slowing things down too much)

[Mike H]

If 1V input, output stepped 0 - 100% of VR1 in 10% steps

At 0 output should be -40 dB or could be more

[Graeme]

Yeah the Alps blue is not that terrific actually, structurally well made though must be said. But hence the "shunt pot conversion" way of using it like I do at the mo.


Think it was Ed or somebody posted some while ago "manages to sound both hard and dull at the same time, which is quite a feat" if I remember correctly!

Yes, i should have mentioned, the LDR replaced a shunted alps pot.

I think i may go back to it though, use some stupidly expensive BMF resistors and call it a day.
Things tend to go wrong when i try and get too complicated

[Andrew]

Voila ~

Gwarn be the first to make one

(I had to take out the lamp model cos it was too complicated and slowing things down too much)

Sorry, Mike, I'm missing something, I admit to a cursory glance only, so tell me to bog off and read it if you like, but what's all the extra circuity do?

Andrew

[Mike H]

LOL I wouldn't do such a thing


In a nutshell, it forces the LDR resistance to follow a linear relationship with the VR1 control pot position. By using a control amplifier with feedback to compare two Voltages, the reference (pot) and the result (Voltage at bottom of the variable resistance LDR2).

Hope that makes sense.

Note you'd have to use a log pot if you want it logarithmic, in the above plots it's as if it were a linear pot, as you'll notice there's an identical gap between each output Voltage trace and the next one up.



 

[Mike H]

PS ~ if VR1 = 100k, and R4 = 1k, and wiper is at the bottom:

Then, if R5 = 10k, LDR2 must = 1 Meg to get the same ratio.

Ergo, LDR1 is 1 Meg as well. As R8 is also 10k. Giving attenuation = -40 dB.




 

[Andrew]

LOL I wouldn't do such a thing


In a nutshell, it forces the LDR resistance to follow a linear relationship with the VR1 control pot position. By using a control amplifier with feedback to compare two Voltages, the reference (pot) and the result (Voltage at bottom of the variable resistance LDR2).

Hope that makes sense.

Note you'd have to use a log pot if you want it logarithmic, in the above plots it's as if it were a linear pot, as you'll notice there's an identical gap between each output Voltage trace and the next one up.

Ok, thanks, let me put it another way, it might aid my thinking. What problem is the extra circuitry trying to address? Say above and beyond two LDRs (one for each chan) and a bog standard light bulb?

Andrew

[Mike H]

Oh right, well it makes the light output exactly the right amount to get the resistance it wants. The light output could vary a lot versus filament Voltage and so on, it's not linear, like 5V is not going to get you exactly 0.5 lux, kind of thing



Reason for edit: word missing!

&160;

[Andrew]

Ok that's cool.

So if I had one of your circuits for each channel, presumably running off the same pot, your circuit would force the two LDR's channels to track the single pot linearly? Thus eliminating the problem of matching the two LDRs to get channel balance... yes?

Andrew

[Mike H]

That is the basic idea. Still all theory though! It assumes the LDR's (there will be four for stereo) are at least reasonably matched in behaviour. Although thinking about it, should only be necessary to match two together to make each pair.

Yep should be able to run off one pot track.

Problee lots of things haven't been figured out in detail yet, like where is this 1V DC reference coming from. Don't suppose the precise value is critical so could be one of them 1.24V precision reference chips



 

[PeteC]

FWIW ,

from the original question ...

yes I built one early this year - on veroboard. I copied the Lightspeed circuit.
It makes a very nice volume control indeed.

I have my scrappy layout sketch somewhere which I could scan this weekend if needed ? Otherwise could take digital shot of the guts etc for layout.

cheers
Pete

[Mike H]

For the record, it's not that I'm poo-pooing the lightspeed system or anything like that, just me using a bit of lateral thinking (one of my favourite pastimes ) to see if it can be done a different way.

Using bulbs seems to work for cinema projectors so must be something in it