[sdiy] LED/LDR modules

[Grant Richter]

> most of the people who like ldr's I've spoken to like the slowness in them,
> it's part of the sound I spose.. generally they all will be pretty slow,
> nothing like close to OTA or other semiconductor device, but then they
> prolly have different speeds of LDR so maybe you could get a faster one..but
> it will prolly still be notably slow..

The Hamamatsu high speed ones are practically indistinguishable from an OTA.

> maybe the rate of change is proportional to the amt of instantaneous light
> change.. ie if you blast it with a huge amt of light it will get to a point
> much more quickly than with a lesser amt of light [so with the high light
> exposure, it obviously would be a point midway along somewhere, not the
> stable value for this light lvl].. if this is the case, you could cause a
> large light pulse to be pulsed when a change occours, speeding it up.. could
> work ok for interesting manipulation?

Exactly what Buchla does. Bypass the LED current limiting resistor with a 1
uF cap. Gives a current spike on the raising edge and speeds up response by
about 5 ms.

The thing that makes a Buchla 200 unique, is that there are no OTAs used at
all. The entire thing is built from opto-couplers. That is why the audio and
CV use  different connectors. The audio and control voltage systems are
optically isolated from each other, and use separate power supplies. No CV
feedthru.

Opto couplers are also inherently lower noise than a OTA because they are
passive noise sources (like a resistor), rather than active noise sources
(like a transistor). All electrical current is noisy because the current
flow is not uniform, but flows in "clumps". A resistor converts this current
noise to a voltage noise by ohms law. That's why low noise circuits use low
value resistors, less resistance = less voltage noise produced for the same
current noise.

A transistor controls the noisy current, but adds additional clumpyness to
the already clumpy current.