Dirt cheap low noise remote microposition controller.

[The_Doctor]

I was disassembling an old pinboard for a new circuit, and noted the existing layout to save it. I thought it might be useful to someone here, so here it is...

This is a dirt-cheap and effective low noise microposition controller based on an LM317T regulator configured for adjustable constant voltage output. It can be useful for remote control of fine focus and many other things, adjusting right down to near zero volts. Accuracy, drift and repeatability depend mostly on the mechanics used, as will linearity. While it probably won't do for finished designs, it should be ideal for simple projects and testing stuff.

(Schematic made in Courier New font, size 10, Windows ANSI coding.)

             |¯¯¯¯¯¯¯¯¯¯¯|          D1   D2
     ¡-------|  LM317 T  |----x---|>|--|>|---x------¡
     ¦       |___________|    ¦              ¦      ¦
     ¦             ¦       R1 ¦              ¦      ¦
     x      ¡------x--/\/\/\--!       ¡------x      x
            ¦ VR1  ¦                  ¦ R2   ¦
            Z     _¦_C1               Z     _¦_C2
     x      Z<-¡                      Z             x
     ¦      Z  ¦  ¯¦¯                 Z     ¯¦¯     ¦
     ¦      ¦--!   ¦                  ¦      ¦      ¦
     !------x------x------------------x------x------!


              VR1     10K Lin
              R1      2K2
              R2      10K
              D1,D2   1N4001
              C1      4µ7 Tantalum
              C2      1000µF Electrolytic

Input +12 VDC.
Output, 5 watt mylar speaker or similar (better) actuator.
If using a mylar speaker, bond the magnet's back to a metal chassis for cooling, this is critical for neodymium magnets. Also heatsink the LM317T with 10°C/W or less.


VR1 sets position, with C1 to reduce noise from VR1.
R2 and C2 prevent shocks to the mechanism if fast changes occur on VR1.
D1 and D2 keep the output pin voltage at least two diode drops above ground, which allows the output drive to be varied close to zero.

If you like it, tell others; if you don't, tell me. :twisted:


Edit:
Is constant VOLTAGE, not current, as I first wrote, I got confused with a variant I was going to try, and didn't. A variable current source that can go to zero, be linear, and not fry the potentiometer IS possible, using a similar parts list to this circuit, but the config is not the same. It uses a floating voltage reference to bias the ajust pin, and maybe the two diode drops will serve, but you'd probably need better accuracy.

The circuit above is ideal if you've got a constant load though, where voltage control is enough.

[buffo]

Thanks for the tip! Also, thanks for posting it here at PL before posting it in alt.lsers. (Makes me feel special!)

Just kidding - with all the brewhaha over the PL post on alt.lasers, I figured a little humor was in order.

Off topic, but why didn't you keep the "lostgallifreyan" nick when you registered here?

Adam

[The_Doctor]

Good question. Not sure.. I think it's mainly cos I used The_Doctor on the Music-Reviewed forum and others before it. Lostgallifreyan isn't easily shortened either, as Skywise noted on alt.lasers...

[marconi]

Hey Crow... Nice circuit

That is very similar to a motor driver I built some years back !!!
I think it was for the lissajou generators we built for a while.
3 of these circuits and three small motors with mirrors attached
perpendicular to the shaft.

The good 'ol 317.. it's a great device, Almost can do anything
Of course It wont drive diodes very well on its own,,..
You'd still need a dropping resistor..

As for constant current source..even around half an amp
I still had to play with several low value resistors ...juggling
to get what I needed... but it worked.
Using the diodes as the current sense resistor will give funny results
as they change resistance with load...
Even a temperature change will change the output..

Good for general use..Altho I prefer Op-Amps.
Better control and all..
But still.. that circuit is easy to construct and do many things

Anyone think about using these for controlling seleniods for shutters?

OBTW, I know where Gallifrey is..hehe

--chuck

[The_Doctor]

I love op-amps. I'll be putting up a circuit using a couple of them soon, for the Scientech 360001 to Voltmeter with differentiator response acceleration, if I can make it work well.

Diodes change with load, true, but that circuit is for constant load. While it might be hard to predict with easy rigour where a value will be, it should be close enough, and it should stay put.

I wouldn't use it to drive diodes. But I might use this:

             |¯¯¯¯¯¯¯¯¯¯¯|         R1
     ¡-------|  LM317 T  |----/\/\/\----x----x------¡
     ¦       |___________|              ¦    ¦      ¦
     ¦             ¦                    ¦    ¦      ¦
     x             !------------¡       ¦    ¦      x
                                ¦ VR1   ¦   _¦_C1
                         ¡----/\/\/\----x
     x               R2  ¦     D1   D2  ¦   ¯¦¯     x
     ¦      ¡---/\/\/\---x---|<|--|<|---!    ¦      ¦
     ¦      ¦                                ¦      ¦
     !------x--------------------------------x------!


              VR1     10K Lin
              R1      1R 2W
              R2      10K
              D1,D2   1N4148
              C1      1000µF Electrolytic

That's the current controlled equivalent. I haven't built it to try it, so the resistances VR1 and R2 are only starting estimates. It should be good for 0 to 1.25 amps. If it works. I found the original version too. It's in the LaserFAQ, posted by Winfield Hill. He specifies a 1.23V bandgap reference and a low current source in place of D1,D2 and R2, but I think the simplest form will often work, at least as well as the voltage form I posted earlier.

Actually, I prefer to tweak LM317 diode drives as you do, by juggling fixed resistances in constant current config, adding or subtracting half-ohm resistors.

About the LM317, I think it's maybe the greatest single IC ever made. I've seen amazing circuits built around it. I even saw a high(ish) power radio transmitter once. Not working, but a ham radio guy built it, and I saw the design he posted online. I can't remember the URL though, I saved a few things once..

One feature of all the circuits I'll post, anywhere, is extreme low cost and part count. It's like the equivalent of machine code, or Haikus, or maybe a martial art move. I like the effort and the final simplicity that comes when it's perfect. It's also the right thing to come from someone with lots of time but little money...

[marconi]

Hey Crow ,..I hadnt tried that one yet

I'll give that one a go... next time I'm breadboarding a 317
it'l be interesting to see what happends...
I'm always interested in ways circuits handle temperature changes as well !!!

Yeah I like simple circuits as well, less parts to go bad..
Sometimes easier to troubleshoot too !!

Hey, If anyone here sees these circuits and tries to use them for driving diodes
Please be very careful..
You will need to experiment with component values quite a bit..
Could be fun tho !!!

[The_Doctor]

Oh, it's definitely NOT diode-ready...

There are several protections missing, apart from it being risky, variable. Should be fixed current with temperature compensation if needed, and have some capacitors across the LD and a fast diode in inverse parallel with it too. And maybe a series 2 ohm resistor after the current sense resistance too, before anything reaches the LD and other protection stuff.

The main point of these circuits is they can be adapted easily to micropositioning tasks that normally cost a lot more to solve.

[marconi]

yeap...True..

This is for other stuff other than for diodes.. I got in a hurry in that post.
I had to leave.


I really think Im gonna try it on some seleniods !!!
Thanks for posting it Crow