Photoresistor for gain - noobish question

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

shaddai

Active member
Joined
Dec 10, 2005
Messages
35
Location
North Idaho
Been reading the Ultimate analog desk thread & about page 9 & 10 the discussion of digital pots come up. I've been tempted to monkey with a little linux->usb->i2c->digipot setup to light up an LED or something simple. Not for any specific purpose, but just to learn the path to making it work.

Anyway, one of those brain farts happened and it led me to wondering if there was any reason a photoresistor/LDR as R1 or R2 in a negative feedback opamp situation would be bad news? I read on here somewhere it's because of thd & noise, but couldn't that somewhat conquered with math & part selection? Enlighten me please  :p

todd
 
LDRs have been around for a long time and were widely used to vary level. Typically they were used as a shunt element (leg to ground) to pad down an input signal. This way the LDR is off at max gain (min attenuation).

They could be used as one leg of an opamp feedback path but it doesn't immediately appear advantageous to me.

I looked at them seriously back in the 70s to use as a practical gain element. I even tried using half of a center-tapped LDR to linearize the gain control law of the other half. I could make it work but it wasn't attractive for commercial use even compared to the VCAs of the day.

They remain attractive for some specialized applications like to mimic the classic gain element used in an old legacy limiter, but not because they're better than modern technology, perhaps because they're not.

JR
 
To give you specific answers I'd have to look at specific data sheets, that I haven't opened in decades.

My active imagination can not imagine anything advantageous about using them that way.

What is your specific need or application?

JR

 
My knowledge of these things is a little basic but disadvantages I can think of are:

Significant inconsistency between one LDR and the next at any given current through the LED

Memory effect where an LDR that has been 'dark' for a long time will give a different resistance than one that has been on recently

Cost? Maybe not such a problem for DIY.

Large-ish current draw for the LED?

A better use for an LDR might be as the shunt in a pad for volume control... there will be a limit to the amount of attenuation you can expect though I would guess. (Theres another one!)

But for experimenting, why not make a little analog test circuit and see what you get?
 
One more disadvantage for me is that LDRs respond to infrared  as well - means to heat, means that it is impossible to get a repeatable value without temperature compensation. I had a great sounding programmable tube guitar amp back in the 80s (50 watts in 2 HE). They used LDRs to control the beast and it was impossible to keep the sound from changing over time. The same program sounded totally different in the morning than in the afternoon after some hours of practicing.

On the other hand LDRs can be used successfully, otherwise noone would care for LA2As. I guess the effect is less dramatic there, because the temperature change isn't as drastic, the LDR has more cooling and it's 'just' gain that varies, not the whole EQ and distortion like in that guitar amp.

Michael
 
disadvantages:

Significant inconsistency between one LDR and the next at any given current through the LED

Memory effect where an LDR that has been 'dark' for a long time will give a different resistance than one that has been on recently

One more disadvantage for me is that LDRs respond to infrared  as well - means to heat, means that it is impossible to get a repeatable value without temperature compensation. 
These can be somewhat overcome by using dual LDR elements with DC feedback and two trimmers. The need for trimming makes it unsuitable for industrialization, but OK for DIY.
Unfortunately, in audio, Vactrols are used essentially as soft switches, so manufacturers do not care about matching.
There are two applications where vactrols may offer significant advantage:
Frequency control of state-variable filters: multiple RevLog pots are becoming unobtainium
Gain control of Cohen-type mic pres: they need a very specific RevLog pot (about 3% at mid-rotation) that are also impossible to get off the shelf.
 
Back
Top