What is this circuit section doing??

[slacker]

Can you elaborate? What's it linearising and how?

[Gus]

I posted what it is doing in the other thread

It is a servo control to control the LED light output

[Digital Larry]

The op-amp and trimmer circuit form a light-bias circuit for the right hand LDR.  Pretty sure that the LED current does not change except in response to the trimmer.  Even though the circuit looks somewhat like a comparator, think of it as a really high gain amplifier (which it is).

As the trimmer increases the (+) input voltage, the output voltage goes up and more current goes through the transistor and LED.  This reduces the resistance of the LDR on the left and the (-) input goes up.  How far does the (-) input go up?  It goes up until it matches the (+) input, if it can.  If it goes a little higher than the (+) input, then the transistor base voltage starts to drop, reducing the collector current, etc. so it balances out.   This is the servo circuit in action.  It adjusts the current to whatever it needs to be, within the limits, to get the (-) voltage to match the (+) voltage.  It automatically creates a non linear current/illumination curve in order to get a linear resistance curve.  The same illumination is used for the audio control LDR.  Fortunately you adjust it by ear so that you can accommodate any differences in their sensitivity or illumination.

There are limits to (1) how much current you can pull through the LED and (2) how low the LDR resistance will go.  The op-amp's output voltage is (when the circuit is working properly) not slammed high or low as it would be for a comparator.  Once you hit one of those limits, the output voltage can't move far enough to reduce the error voltage to zero and so the op-amp output WILL be slammed into the rails.

I think figure 15 in the link I posted awhile back is instructive.  In that example there are two separate LDR/LED assemblies, but the LEDs are connected in series and the nonlinearity of the LED current to LDR resistance relationship are assumed to be identical for both devices (which might not be true, but it's as close as we're going to get).  In this example there is one LED shining on two LDRs and most likely the intensity is not the same, but that is just a matter of a gain correction factor (er... or it might explain why the left hand LDR is looking at the LED from the side).

[PRR]

> the LED current does not change except in response to the trimmer.

Exactly so for a hacker's guide.

It keeps the LED's *brightness* CONSTANT.

Secondarily it corrects LED and LDR temperature drift, age drift, and perhaps battery voltage drop.

Imagine you want to keep a 100fc lamp at 85fc, despite wall-voltage surge/sag or lamp glass darkening with age. You put a light meter on it. You compare the meter to the desired 85fc. If you read 80fc, you dump more power to the light; if 90fc you put less power in. This is a dummy job. If the light-meter output and the desired light level are both voltage signals, any (DC) amplifier can do it. Depending on the lamp, it may be necessary or convenient to buffer the lamp, hence Q3.

Sadly it does not linearize the LDR's curvature. (However a fair fit can be had by "bending" the shutter, either with a curve-taper slot or in the linkage to your shoe.)

What the trimmer probably does here is correct for the flanger's JFETs turn-on/off voltage or voltage-range. If you have alignment instructions(?) this will probably make sense: you trim for best response when the shutter is not blocking.

[Lurco]

somehow I had fig.3 in mind, which by accident just had crossed me earlier yesterday: http://www.keith-snook.info/Wireless-World-Stuff/Wireless-World-1976/Wideband%20compander%20design.pdf . Call it optical feedback or servo, ymmv.

[slacker]

Slaps head, I was being a complete buffoon, got too carried away with looking at what happened at the extremes of the LDR's resistance and forgot about all the points in between.
Thanks to Larry's explanation it now makes sense to me. Can I have some points for getting the right answer even though my working was wrong?

[Digital Larry]

Can I have some points for getting the right answer even though my working was wrong?

This circuit, a non inverting op-amp stage, and an inverting op-amp stage all can be understood on a few fundamental principals.
1) There is a feedback path from the output to the (-) input.  The difference between the (+) and (-) inputs is considered the "error" voltage.
2) The circuit, due to its extremely high gain, moves the output voltage to "whatever is necessary" to keep the voltage at the (+) and (-) inputs the same, reducing the "error" to zero.
3) The op-amp's input impedance is infinite (no current flows into either input) - this is an oversimplification.
4) The op-amp's output impedance is zero (another oversimplification).

A comparator lacks negative feedback, so (1) and (2) do not hold in that case.  (3) and (4) are still valid.

[armdnrdy]

We've got a winner!!

Thanks Larry for the explanation!

I was completely wrong in my assumption that this circuit had something to do with a cheese grating attachment! 

I'm working on a prototype of the LDR/LED tubes with dimensions provided by Dino. The idea is to fabricate them from easily obtainable plumbing/sprinkler parts.

So Larry, when are you or Mick going to post one of those cool sounding FV-1 projects?

[Digital Larry]

So Larry, when are you or Mick going to post one of those cool sounding FV-1 projects?

Hi Larry,

Glad to have helped.  Part of my enjoyment hanging out here is dusting off my ancient analog skilz.

If you MUST know, I have a couple top-secret possibly revolutionary designs in progress.  These leverage the FV-1 in a way that I have not seen done and also tie into some things which are very commonly discussed here.  Interested yet?     By the way I sure hope it works!

I was at first motivated by the contest but I started a new job a few months back and so I don't think I'll be able to make the deadline and still give it enough time.  Hopefully by the end of the year.

[armdnrdy]

If you MUST know, I have a couple top-secret possibly revolutionary designs in progress.  These leverage the FV-1 in a way that I have not seen done and also tie into some things which are very commonly discussed here.  Interested yet?    

Does this design have anything to do with easily obtainable plumbing/sprinkler parts? 

You threw out just enough to make me curious! Let me see....Octave divider perhaps?

[digi2t]

If you MUST know, I have a couple top-secret possibly revolutionary designs in progress.  These leverage the FV-1 in a way that I have not seen done and also tie into some things which are very commonly discussed here.  Interested yet?    

Does this design have anything to do with easily obtainable plumbing/sprinkler parts? 

You threw out just enough to make me curious! Let me see....Octave divider perhaps?

The octave divider might be something to consider, but I`m definately in for the cheese grater attachment. 

Thanks for the explination. Especially in terms that I can understand.

[Digital Larry]

Does this design have anything to do with easily obtainable plumbing/sprinkler parts?  

The octave divider might be something to consider, but I`m definately in for the cheese grater attachment.  
Thanks for the explination. Especially in terms that I can understand.

(furiously scribbling notes on old In-n-Out Burger wrappers)...

Hmmm, an Octave Divider is one thing this WON'T do, although it will support pitch shifting.

Now I'm wondering why you want "easily obtainable" plumbing/sprinkler parts, when I just so happen to have some NOS RainBird timers over here... pre-RoHS for sure... well, I don't want to give TOO much away... cheese grater... flat or one of those hand crank dealybobs?

[tubegeek]

Let me get this straight. What you are saying is that you are 100% sure that you're 50% confident? 

It might be the other way around.

But how can you be sure?

"60% of the time, it works every time."

[digi2t]

cheese grater... flat or one of those hand crank dealybobs?

Flat... definitely flat. Nothing beats that old school elbow grease. Makes the cheese taste better.

[armdnrdy]

Okay,

Here's the "tube" I fabricated that houses LDR1 and D18.

[digi2t]

LAIR!!!! You actually salvaged that from a scrapped Mutron Flanger.










Beautiful work Larry. What materials did you use?

[Digital Larry]

Let me get this straight. What you are saying is that you are 100% sure that you're 50% confident? 

It might be the other way around.

But how can you be sure?
"60% of the time, it works every time."

I checked with my cat, but he doesn't appear to be feeling well.

[armdnrdy]

LAIR!!!! You actually salvaged that from a scrapped Mutron Flanger.










Beautiful work Larry. What materials did you use?

I wasn't joking about the sprinkler parts! 

A 1/2" X 12" cut-off riser from Home Depot, a 3/4"plug, party favor (plastic) magnifying glass for the lens from Party City, red, diffused LED, and a 10K-1M 20ms-30ms LDR that I had some stock of.

I'll take some construction photos for the next tube.

[digi2t]

Brilliant!!

[R.G.]

But how can you be sure?
"60% of the time, it works every time."

There is a serious saying in programming that "almost always is almost always as good as always".