Author |
Message |
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Fri Nov 09, 2012 11:52 am Post subject:
DIY ribbon Subject description: from scratch |
|
|
Finding materials for making ribbon controllers can be quite frustrating. None of the anti static bags or VHS tapes I could find had low enough resistivity to be useful. So I looked around for ways to make a resistive strip from scratch. It turns out to be quite easy. Most glues and paints can be loaded with powdered graphite to make them conductive.
The materials I ended up using are easy to find locally. They are acrylic ink (or paint) from the art supply / hobby store, fine powdered graphite from the hardware store and a Sketch & Wash water-soluble graphite pencil, also from the art supply store. The Sketch & Wash gives about a third the resistance of a regular soft graphite pencil.
I've done my first experiments using paper substrates. The acrylic ink also seems to adhere very well to acrylic (plexiglass) sheets, so that will be the next thing to try. The paper must be hard and flat -- any waviness will make it impossible to get uniform electrical contact. The best paper substrate I found is called "collage board". I got the Crescent brand medium weight sheets.
To get sheet resistivities down around 1 kOhm / square, as required for potentiometers under ~50 kOhm total resistance, more than one conducting layer is needed. I usually start by using the Sketch & Wash pencil to load up the paper and then add a layer or two of graphite-loaded acrylic ink. This needs to be mixed up with about 3 parts graphite to two parts of ink. It is critical to get fine graphite powder. I was able to find this at Ace Hardware. The other stores sell a kind with large flakes, which doesn't work at all. You can also find large quantity bottles online.
The paint usually needs some careful fine sanding to get uniform resistivity along the strip. To measure the local resistivity, you can do this: Take a small strip of wood, about 10 mm x 15 mm. Wrap small strips of Al foil tape around the ends to define a small square with the contacts along two sides. Then just use your DVM test probes to push the little device against the resistive strip. This doesn't exactly give you Ohms / square, but it's OK for relative measurements to check the uniformity of the potentiometer strip. With some practice and patience it's easy to get linearity better than 5%.
I usually go over the strip again with the pencil pressing hard to get a well burnished shiny surface.
For the conductive strip that presses against the resistive strip, I found a rather remarkable product. It is called "Ni/Cu/Co Conductive Fabric Shielding Tape" and is available at lessemf.com It's very flexible, has a nice fine uniform weave and has metalic conductivity. I just hang it over the carbon strip between a pair of simple clamps.
The attached drawing shows how I used layers of paper and tape to build up a controller. This is a cross section of the central part and doesn't include the details of clamping and contacting. These are done with brass bars that screw down on top of the conducting strip, and Al foil pasted to the end of the resistive strip with conducting glue made from graphite-loaded acrylic paint. The paper tape on top has a feel that I like better that the acrylic of the packaging tape that carries the conductive tape. I comes from the drug store and is called Nexcare Gentle Paper Tape (aka 3M Micropore surgical tape).
I hope all this at least gives a basis for folks who want to try experimenting with DIY ribbons.
Ian
Description: |
|
Filesize: |
8.93 KB |
Viewed: |
42976 Time(s) |
|
|
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Thu Nov 15, 2012 7:04 pm Post subject:
|
|
|
Here's another ribbon version. This one is built on a two foot long base, with an active length of ~530mm. In this version, the resistive strip is painted directly on a plexiglas substrate. It has been carefully adjusted to have better than 1% linearity. The conductive-tape ribbon is suspended over the strip by a piece of rubber cut from an exercise band. This forms a nice and wide and very comfortable, organic-feeling trough for the finger to ride in. The top layer is Nexcare "flexible clear" tape from the drug store. It has a nice tooth so you can feel your finger move without having a lot of friction or bending resistance. This device is an absolute joy to play. It kicks the crap out of the softpot with its narrow, unconfined, hard and sticky surface.
Ian
Description: |
|
Filesize: |
14.98 KB |
Viewed: |
42904 Time(s) |
|
Description: |
|
Filesize: |
112.31 KB |
Viewed: |
934 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Filesize: |
80.42 KB |
Viewed: |
987 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
|
|
Back to top
|
|
|
drapdap
Joined: Oct 11, 2004 Posts: 204 Location: London
Audio files: 1
G2 patch files: 1
|
Posted: Sun Nov 18, 2012 3:16 pm Post subject:
|
|
|
thanks Ian! this is great info. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Sat Dec 01, 2012 11:54 pm Post subject:
|
|
|
Finally got a nice long linear ribbon working. Also improvements to the circuit that make it pretty much bulletproof.
The ribbon is similar to the ones described above. I painted a strip of acrylic gesso right on a piece of wood, then loaded it up using the water-soluble graphite pencil. The strip is ~ 1" wide and 2' long, and its resistance is ~5kOhm. So multiple layers were not needed. This is because the gesso has a lot of "tooth" (that's what it's for). To linearize the resistance I painted acrylic ink loaded with powdered graphite along the edges where needed. I got the final linearity down to around +/- .3% FS. This is 10x better than the Softpot spec!
I got the circuitry tweaked up to where there are very rarely errors with normal playing. It will still gurgle if the ribbon is pressed very lightly, and you can get an error occassionally if your finger is right on the edge of the track.
A block diagram of the electronics is attached. The raw signal is processed in several parallel branches. Complementary comparators detect the finger up and down events and fire a one-shot for blanking the Track/Hold circuitry for a fixed debouncing time.
The raw signal is fed through two lag processors, one "Fast" and one "Slow", each with its own T/H circuit. Lagging produces signals that are smoothed and delayed. The difference between the Raw and Fast lag signals is a measure of the rate of change of the signal and is used to inhibit tracking when it is above a small threshold.
The Fast T/H is the main output when the finger is down. It is sent to the output via an analog switch, provided there is no blanking signal. Upon finger release, both T/H circuits are set to hold, and the output is taken from the Slow lag circuit. The eliminates droop during the switching time.
I'm still thinking about ways to improve the electronics, so there may be further changes. But right now I'm very happy with how it is working.
Ian
Description: |
|
Filesize: |
8.1 KB |
Viewed: |
760 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
|
|
Back to top
|
|
|
LFLab
Joined: Dec 17, 2009 Posts: 497 Location: Rosmalen, Netherlands
|
Posted: Mon Dec 03, 2012 4:08 am Post subject:
|
|
|
Looks great, Ian!
Any worries about the longevity of the masking tape? I can imagine that the adhesive used is aimed at removeability after painting?
I wonder about the blackboard paint that you can also get in spraycans (at least this side of the pond), I know some people use it for shielding guitars, so perhaps it has too low resistivity. Not sure if that lacquer is loaded with graphite or something else. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Mon Dec 03, 2012 2:04 pm Post subject:
|
|
|
LFLab wrote: | Any worries about the longevity of the masking tape? I can imagine that the adhesive used is aimed at removeability after painting? |
True. It's pretty sticky, though. I'm not using it for the active layers anymore -- it wasn't springy enough for that.
Quote: | I wonder about the blackboard paint that you can also get in spraycans (at least this side of the pond), I know some people use it for shielding guitars, so perhaps it has too low resistivity. Not sure if that lacquer is loaded with graphite or something else. |
I went out and got a can of blackboard paint to try, but it doesn't seem to conduct at all. Is there a specific brand I need to look for?
Ian |
|
Back to top
|
|
|
LFLab
Joined: Dec 17, 2009 Posts: 497 Location: Rosmalen, Netherlands
|
Posted: Sun Dec 16, 2012 5:11 am Post subject:
|
|
|
Dug out the tin I bought a while ago, and it is actually magnetic paint. Hadn't tested it myself, so smeared something on a piece of wood.
Not dried yet, I now get 150-200k with the probes +/-15mm apart. I'll do another measurement when it dried, and probably will do another two coats. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Sun Dec 16, 2012 7:58 am Post subject:
|
|
|
LFLab wrote: | Dug out the tin I bought a while ago, and it is actually magnetic paint. Hadn't tested it myself, so smeared something on a piece of wood.
Not dried yet, I now get 150-200k with the probes +/-15mm apart. I'll do another measurement when it dried, and probably will do another two coats. |
Thanks, I'll have a look for some. Is yours in a spray can? The biggest headache with paint is getting it uniform.
Ian |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
|
Back to top
|
|
|
Benjamin AM
Joined: Nov 04, 2010 Posts: 83 Location: Boise
|
Posted: Sun Dec 16, 2012 10:45 am Post subject:
|
|
|
I'm dieing to to see/ hear this beast. |
|
Back to top
|
|
|
LFLab
Joined: Dec 17, 2009 Posts: 497 Location: Rosmalen, Netherlands
|
Posted: Sun Dec 16, 2012 11:16 am Post subject:
|
|
|
frijitz wrote: | LFLab wrote: | Dug out the tin I bought a while ago, and it is actually magnetic paint. Hadn't tested it myself, so smeared something on a piece of wood.
Not dried yet, I now get 150-200k with the probes +/-15mm apart. I'll do another measurement when it dried, and probably will do another two coats. |
Thanks, I'll have a look for some. Is yours in a spray can? The biggest headache with paint is getting it uniform.
Ian |
It is dry now, but the weird thing is that some place do have a finite resistance, while other locations do not conduct at all. In a thicker layer I got about 300-500k with the probes at 50mm apart. But surface resistivity is spotty, you need to push the probes in a bit to get a reading.
I just put on the second layer, we'll see tomorrow.
The paint I have is pretty thick, don't think you can use this particular brand in a spray can.
As for uniform layers, you almost need to improvise some kind of doctor blading setup to get it really uniform. And it seems you need a bit of layer thickness to get good electrical conductivity.
Quite a bit of circuit needed to get useable results for a ribbon it seems, pretty nifty block diagram.
With respect to the switching time you are using, does that need tweaking often? Is it a dependent on playing technique or the ribbon used? |
|
Back to top
|
|
|
LFLab
Joined: Dec 17, 2009 Posts: 497 Location: Rosmalen, Netherlands
|
Posted: Mon Dec 17, 2012 11:56 pm Post subject:
|
|
|
Update:
Completely dried now, I get no reading whatsoever, unless the probes are really close together (500k @ 1.5mm).
This is not going to work. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Tue Dec 18, 2012 7:22 pm Post subject:
|
|
|
LFLab wrote: | Quite a bit of circuit needed to get useable results for a ribbon it seems, pretty nifty block diagram. |
The complexity is strictly from the requirement for the output voltage to hold on finger-up. Otherwise you can just use the ribbon as the R in an RC oscillator:
http://torigoya.main.jp/RibbonController_en.html
Quote: | With respect to the switching time you are using, does that need tweaking often? Is it a dependent on playing technique or the ribbon used? |
The time constant needed for the Slow lag seems to depend somewhat on the ribbon mechanical characteristics, ie how much bounce and springback. There is also probably a need to adjust the gain of the derivative circuit for different ribbons, but I haven't looked at this in detail yet.
I have several ribbons built now and am working on yet another. At some point I will looking at optimizing the circuit for each one. Probably I'll get some PCBs made so I can have a dedicated board for each ribbon.
Ian |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Tue Dec 18, 2012 7:24 pm Post subject:
|
|
|
LFLab wrote: | Update:
Completely dried now, I get no reading whatsoever, unless the probes are really close together (500k @ 1.5mm). |
Thanks for taking the time to look at this and for the update.
Ian |
|
Back to top
|
|
|
Benjamin AM
Joined: Nov 04, 2010 Posts: 83 Location: Boise
|
Posted: Fri Feb 01, 2013 10:29 am Post subject:
|
|
|
Hey Ian,
Now that you've spent some time with the new ribbon design, any new thoughts? I would love to try it out myself. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Sat Feb 02, 2013 8:04 pm Post subject:
|
|
|
Benjamin AM wrote: | Now that you've spent some time with the new ribbon design, any new thoughts? I would love to try it out myself. |
Thanks for asking. I'm still working hard on this project, and have a new incarnation just about finished. I like the mechanical parts -- dimensions, materials -- but getting a good linear scale is still challenging. I'm taking more snapshots as I go along this time, and carefully writing down all the dimensions. So I should have more detailed info available to post soon.
Ian |
|
Back to top
|
|
|
garcho
Joined: Sep 28, 2011 Posts: 41 Location: Richmond, VA
Audio files: 2
|
Posted: Sat Feb 23, 2013 12:31 pm Post subject:
|
|
|
Any word? How's it coming along? _________________ "The life of a repo man is always intense." |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Sat Feb 23, 2013 3:06 pm Post subject:
|
|
|
garcho wrote: | Any word? How's it coming along? |
Thanks for your interest! That's what, four people now? I'm still trying to figure out how to get better linearity on the resistive strip. I get it perfectly linear running a test probe up and down directly on the element, but then it is off after assembly. Take it apart again and it looks OK. So that's what I'll be working on the next couple of weeks. Everything else I'm happy with.
Ian |
|
Back to top
|
|
|
BananaPlug
Joined: Jul 04, 2007 Posts: 307 Location: Philly
Audio files: 5
|
Posted: Tue Apr 09, 2013 2:39 pm Post subject:
|
|
|
Make that 5 people. I'm also curious what became of this. Have also been seeking something simpler than Appendage but with a held voltage after finger-up. |
|
Back to top
|
|
|
RingMad
Joined: Jan 15, 2011 Posts: 427 Location: Montreal, Canada
Audio files: 4
|
Posted: Wed Apr 17, 2013 3:08 am Post subject:
|
|
|
Thanks for posting all these details of your research, frijitz... I was never quite happy with my rough experiments involving broadcast videotape as a resistive strip, and your approach looks much better. I have a few projects to finish before I get around to trying this stuff... hopefully soon.
Thanks,
James. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Thu Apr 18, 2013 9:10 am Post subject:
|
|
|
Afraid I've been sidetracked from working on this as much as I would like. But I just finished another build and finally was able to get really good linearity. Note positions are within 1/8 in of ideal! This version has less travel and actuation force than the previous. I'm going to try to tighten it up a bit, because it has some issues capturing at the low voltage end of the ribbon.
Ian |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Fri Apr 19, 2013 9:13 pm Post subject:
|
|
|
Benjamin AM wrote: | I'm dieing to to see/ hear this beast. |
Brief demo now up at u-tube:
http://youtu.be/th9v66KzUCI
Ian |
|
Back to top
|
|
|
brock
Joined: May 26, 2011 Posts: 112 Location: Canada
|
Posted: Fri Apr 19, 2013 10:14 pm Post subject:
|
|
|
Nice to finally see and hear it in action.
What is the shortest lag you have found you can use and still have it hold pitch on release? |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Sat Apr 20, 2013 11:32 am Post subject:
|
|
|
brock wrote: | Nice to finally see and hear it in action.
What is the shortest lag you have found you can use and still have it hold pitch on release? |
The fast and slow signal lags are produced by passive 2-pole RCRC networks. The slow lag time constant is 1.6ms (82kOhm / .02uF). Full settling (ie, catching up to the input) takes ~30ms. The only situation where the 30ms would matter would be if you lifted your finger right at the end of a very rapid slide. But you can't stop and lift that abruptly! If you lift during a rapid slide, it's hard to tell where exactly you lifted. In either case you can't really detect the lag.
Of course, this assumes you lift your finger cleanly and rapidly. Sometimes I have had trouble with the pitch shifting no matter how clean the lift is. This looks to usually be a mechanical problem, for example a flaw in the strip or ribbon. A flaw can actually cause the pitch to shift before the circuitry detects that the finger contact is broken. On my most recent build, notes at the low end of the resistance range (highest pitches) had a tiny bit of shift. I was able to get rid of this by shimming the end of the ribbon up a bit, and by increasing the minimum finger-up signal by shifting a reference level in the circuit. This also got rid of a small voltage shift when pressing below the selected pitch.
The fast lag circuit -- which provides the output while the finger is down -- is 10x faster than the slow lag circuit. This allows clean trills. Well, they glide a tiny bit for really large intervals, but this isn't much of a problem.
I hope to put together a pdf on this project, rather than post all the details on the forum. Too hard to get the attachments in order, for one thing.
Ian |
|
Back to top
|
|
|
KnobHell
Joined: Jan 28, 2012 Posts: 56 Location: SLC
|
|
Back to top
|
|
|
|