One pot replacing four

[mth5044]

In particular, one pot replacing the four that would be needed in four pt2399 circuits in parallel for the feedback controls. It's not being used as a variable resistor to ground, as far as I can tell. Not sure if that means anything anyway. Someone had brought up voltage controlled resistors, but I can't seem to find an application for them that pretains with what I want to do.

Here is the rebote for a reference of the pot. http://www.tonepad.com/getFile.asp?id=50  Its called repeats in that layout.

I did a little research on digital pots, but they don't look like they've been quite perfected yet.

[jacobyjd]

4-gang pot? Sometimes the simplest way is best...

Other than that, LDRs come to mind.

[Naz Nomad]

http://www.westfloridacomponents.com/P146/20K+ohm+9mm+4+Ganged+Panel+Control+Potentiometer.html

[mth5044]

4-gang pot? Sometimes the simplest way is best...

Other than that, LDRs come to mind.

I thought about LDRs in the beginning, but can't those only replace a pot when they are being used with two lugs tied together and another to ground or something like that?

http://www.westfloridacomponents.com/P146/20K+ohm+9mm+4+Ganged+Panel+Control+Potentiometer.html

I also saw them, but 20k is a bit low ; ;


Thanks so far 

[earthtonesaudio]

4-gang pot? Sometimes the simplest way is best...

Other than that, LDRs come to mind.

I thought about LDRs in the beginning, but can't those only replace a pot when they are being used with two lugs tied together and another to ground or something like that?

http://www.westfloridacomponents.com/P146/20K+ohm+9mm+4+Ganged+Panel+Control+Potentiometer.html

I also saw them, but 20k is a bit low ; ;


Thanks so far 

You can make a 3-terminal voltage divider with LDRs, but you need two of them to replace one pot, and two out-of-phase lights.  You could meet your requirements with 8 LDRs and a minimum of two LEDs.  Then there's tolerances between the photocells and the matter of making one LED brighter as the other becomes dimmer, not a trivial task.

As long as you're okay with a little inaccuracy and not tracing perfectly together, a quad ganged pot would be the simplest solution (but this is coming from someone with zero knowledge of digital pots, for what it's worth).

[Processaurus]

If you can't find your pot, you could use a quad VCA:
http://www.analog.com/en/amplifiers-and-comparators/variable-gain-amplifiers/ssm2164/products/product.html

[JKowalski]

If I remember correctly, you don't really need a pot for the repeats section.

The feedback path is basically a voltage divider -> into resistor -> into a mixer -> back into the pt2399.

The pot changes the repeats, sure, but the resistor after the pot does as well. I'm pretty sure you can ignore the pot and utilize the single resistor into the mixer which sets the gain of the feedback.

[mth5044]

If I remember correctly, you don't really need a pot for the repeats section.

The feedback path is basically a voltage divider -> into resistor -> into a mixer -> back into the pt2399.

The pot changes the repeats, sure, but the resistor after the pot does as well. I'm pretty sure you can ignore the pot and utilize the single resistor into the mixer which sets the gain of the feedback.

I understand your explanation of what the feedback path is, but I can't really see it in the schematic. Is it correct in thinking that the signal comes out of pin 14 and how much goes to ground and how much is fed back into 16 is determined by the pot?

BUT, your last sentence is the one that is confusing me the most. What is needed here is not an internal trimmer to set and forget the repeats, that is how the rate is going to be done. The feedback needs to be variable for when the system has regeneration switches engaged (if you remember that part of the diagram you drew  ) So substituting in the resistor to determine the feedback won't do.

I am probably misunderstanding your post, I don't know 

[mth5044]

You can make a 3-terminal voltage divider with LDRs, but you need two of them to replace one pot, and two out-of-phase lights.  You could meet your requirements with 8 LDRs and a minimum of two LEDs.  Then there's tolerances between the photocells and the matter of making one LED brighter as the other becomes dimmer, not a trivial task.

As long as you're okay with a little inaccuracy and not tracing perfectly together, a quad ganged pot would be the simplest solution (but this is coming from someone with zero knowledge of digital pots, for what it's worth).

Your first proposal sounds interesting. Would it be as easy as getting a dual gang pot and wiring them backwards to control the LED difference along with lowering the resistance around the pot? I haven't thought it out really, as I'm inbetween classes and checking up on my favorite forum 

I've never really done anything with photocells before, but I've seen them applied in various circuits. Not sure about the 3-terminal voltage divider part of it either. Research tonight I guess!

[mth5044]

If you can't find your pot, you could use a quad VCA:
http://www.analog.com/en/amplifiers-and-comparators/variable-gain-amplifiers/ssm2164/products/product.html

That looks a little... too easy...   I've never seen a VCA, so I'm not familiar with it. I have a feeling it's a trap 

[earthtonesaudio]

Your first proposal sounds interesting. Would it be as easy as getting a dual gang pot and wiring them backwards to control the LED difference along with lowering the resistance around the pot? I haven't thought it out really, as I'm inbetween classes and checking up on my favorite forum 

I've never really done anything with photocells before, but I've seen them applied in various circuits. Not sure about the 3-terminal voltage divider part of it either. Research tonight I guess!

That depends.  LEDs have a non-linear response to a changing voltage, and a somewhat linear response to a changing current.  If you were very lucky, it's possible to pick the "just right" values for everything and get the response you want.  But it's much more likely that you'd need quite a few trimpots to compensate for mismatches between the halves of the dual pot, the LEDs, and the photocells.  If you got it working it would work very well, but the setup would be a nightmare.

[jalien21]

not to mention the task of physically connecting four ldr's to one LED (twice) in such a way that they all get about the same amount of light. so not worth it.

just find a four-ganged pot. they're totally out there- i promise. if you really, really can't find one, a digital pot and microcontroller would work, but it would be a huge pain in the butt.

[sean k]

Why not use a 4 x 3 rotary switch. That allows three different voltage divisions. The problem is, though, that the total restance accross the output resistor is half the total resitance of each divider... If you buffer the signal before going into it you might get away with something like that... but could you live with only three settings?

[JKowalski]

If I remember correctly, you don't really need a pot for the repeats section.

The feedback path is basically a voltage divider -> into resistor -> into a mixer -> back into the pt2399.

The pot changes the repeats, sure, but the resistor after the pot does as well. I'm pretty sure you can ignore the pot and utilize the single resistor into the mixer which sets the gain of the feedback.

I understand your explanation of what the feedback path is, but I can't really see it in the schematic. Is it correct in thinking that the signal comes out of pin 14 and how much goes to ground and how much is fed back into 16 is determined by the pot?

BUT, your last sentence is the one that is confusing me the most. What is needed here is not an internal trimmer to set and forget the repeats, that is how the rate is going to be done. The feedback needs to be variable for when the system has regeneration switches engaged (if you remember that part of the diagram you drew   ) So substituting in the resistor to determine the feedback won't do.

I am probably misunderstanding your post, I don't know  

Yeah, you are. Here's the feedback path (in green):



That trimpot that is suggested limits the maximum amount of repeats. (it does the same exact thing as the repeat pot, except that you set and forget it) Keep that, for sure. Remove the repeat pot, instead of tapping it off at the center lug tap it off at the top lug. Then put your VCR in series with the trimpot. This will give you voltage controlled repeats, that are limited by the trimpot. I showed that in the boxed off copy in the left.


It's true that you are going to have some non-linearity as far as how four of these things will behave, but I don't think it will be too bad. I mean, if one of the PT2399's has a little bit more repeat then the others, are you really going to notice it? If you were doing speed control or something, then I would worry about it, but for repeats it doesn't seem like it would matter.








TAKE A LOOK AT THIS REBOTE. It is the 2.5, and it shows the PT2399 pins a little better. You can see how pins 15, 16, 13, and 14 are op-amps inside the chip. You can see the feedback path from the last op-amp to the first one very clearly, too.

http://tonepad.com/getFile.asp?id=98




Okay, im off to make a vid showing how it turns out as I mod my breadboarded PT2399 mess to do this. Will post in about 30 min.

[mth5044]

Dear JKowalski,

You are a gentleman and a scholar. Although you have explained it a good four times to me already, you didn't give up and even drew a picture 

My next sentence was about to 'while I don't quite get it yet...', it has changed to 'something just clicked' and it makes sense.

IT MAKES SENSE.

Can't wait to hear the results.

[JKowalski]

Dear JKowalski,

You are a gentleman and a scholar. Although you have explained it a good four times to me already, you didn't give up and even drew a picture  

My next sentence was about to 'while I don't quite get it yet...', it has changed to 'something just clicked' and it makes sense.

IT MAKES SENSE.

Can't wait to hear the results.

My pleasure  

Alright, well youtube is taking way too long to upload (I think it's my computers anti-virus stuff, I just reinstalled windows and its been giving me hell with all kinds of sites like yahoo mail and ebay)

So forget about the vid. I'll just explain how it turned out for now.

It works... Using a variable power supply for the LED control, a 1k resistor, and a white superbright LED, the usable range was from 1.5-4V, 4 being infinite repeats (that never go crazy because they are still being limited by the trimpot) and 0-1.5 being about single one. The range is kind of finicky, but you can mess with the resistances around the LDR to make it much better in terms of voltage control range and linearity based on the LED.

The shortest time is limited to the maximum resistance of the photocell with the LED off, so you probably want to either package your LED/LDR well so as to block ambient light from outside/get a vactrol/get a LDR with a high OFF resistance. Even at 0v, I still had about 1 1/4 repeats. If you have a very high on resistance this should go away no problem.

Having said that, this still is not your best choice, though it is by far the easiest. The cons about it are:

1. Only decent, not perfect, linearity
2. Difficult to match four LDRs for the same voltage/resistance ranges (vactrols would help in this, better matched but expensive)

[mth5044]

So I'm finally getting into the subsection development of this (to me) gigantic delay project, which is where I develope a parts list.

Do you remember what number your photocell was? Or the min/max resistance for dark and light? The highest resistance one I can find is 20M in darkness 5 - 20K min/max light resistance. That is the 9203 from small bear.

Thanks man 

[Fp-www.Tonepad.com]

i hope you can post pictures and sound samples.

[JKowalski]

So I'm finally getting into the subsection development of this (to me) gigantic delay project, which is where I develope a parts list.

Do you remember what number your photocell was? Or the min/max resistance for dark and light? The highest resistance one I can find is 20M in darkness 5 - 20K min/max light resistance. That is the 9203 from small bear.

Thanks man  

I don't have any idea what any of my photocells are in range anymore. I just toss them in a box whenever I get a new batch. I package them like this now:

http://www.diystompboxes.com/smfforum/index.php?topic=78945.0

Which gives me ranges from 50 ohms to 50 meg pretty much regardless of stated specifications!

Besides, they vary so much between batches that it really doesn't matter which ones you get.

There is also the fact that for your project, the usable range for the repeats is very small. That range pretty much falls in the range of most photocells.

I'll tell you what, though. I will check on my breadboard (yes, it's still there) what the resistance range for repeats is.



Have you drawn up a preliminary schematic yet? You should definitely do that before you source parts, or it's guaranteed that you'll be regretting forgetting something later.  




EDIT: You know, you should really consider finding yourself a quad potentiometer. It would make this whole thing 100X simpler, more precise, and more reliable.