Responsive Gel

toward silicone gel that can simulate the tactile properties of arbitrary materials

Progress

On Friday we built five inertial motors with Joe Malloch at the IDMIL in exchange for two to call our own. A few larger vibration motors are in the mail. Two 12mm and one 20mm motor from Precision Microdrives and two large vibration motors from All Electronics (seems we bought the last ones as they're not available anymore). Up next is removing the DC blockers in my Presonus Firebox so that we can send frequencies well below 20hz and building a decent 6 channel audio amplifier for the speakers and motors. 

We've made good progress on the fiber optic bend sensors. The parts from International Fiber Optics (ordered from Digikey) arrived, with the exception of the Eska CK-40 cable we were hoping to order (not stocked by Digikey – we opted out of ordering direct from Industrial Fiber Optics). We bought some jacketed fiber from Abra in Montreal and stripped the jackets ourselves. We found that the best way to strip the jackets is to make a shallow incision down the length of the cable with an x-acto blade, being careful not to scratch the core, and peeling the jacket off with force, being careful not to kink the cable. The emitter/receiver modules are designed to attach to jacketed 1mm fiber, so we had to slip jacket carcasses onto the ends of the cables and screw them down tight. Bill Vorn has a jig for scoring one side of the cables evenly (just a block of metal with a groove in it), but it was designed for a jacketed cable (oops!), so we're doing it by hand using an x-acto. Attaching it to an LED as well as monitoring the receiver with a multimeter gives you an idea of your progress. If they turn out to be too hard to calibrate, we may find some more cable and strip it using Bill's jig from the start. 

Vibrations Speakers are coming in the mail from Namiki in Tokyo. We're curious as to how these will respond to complex sound signals. Will they reproduce low-frequency components as vibration, or will they require a clean low frequency sine-wave? How will the perform below 30hz? 

I spent a good six hours trying to get two XBees to synchronize and send I/O packets through an Arduino to Max. Originally I planned to write a parser for Max, but this proved a waste of time given that a library, xbee-arduino has already been written. Sadly, I wasn't able to get it to work. Elio at the Sensor Lab took the XBees home and got them working with his own hand-written XBee-Arduino library. I proposed that we clean it up for release over Christmas break. Two XBees and a shield should arrive on Monday and hopefully without too much more effort, the sensor portion of the project will be wireless. 

Faiq is planning to buy some plexiglass to make a badass stand for the gel. All of the silicone (Smooth-On Dragon Skin 10 Slow purchased from in Laval, Quebec) has been purchased. We'll be using about 5lbs for an 8"x10"x1.5" slab. 

Now, software!

(download)

Click here to download:
Namiki VS1643-VS1842 Datasheet.pdf (651 KB)
(download)

New Developments

We'll be heading out to <a href="http://www.sial-canada.com/">Sial</a> in Laval to look at different Smooth-On silicone gel products soon. We may go with Dragon Skin at a Shore hardness between 20 and 30. After trying out different Shore hardnesses at Sial, we went with Smooth-On Dragon Skin 10 (10A hardness). We brought our prototype vibration motor and, testing it with samples, discovered that hardness had a negligible effect on the ability of the gel to transmit vibrations. 

I spoke with Martin from Bill Vorn's lab at Concordia and Avrum Hollinger from McGill about fiberoptic bend sensors. They both recommended parts from Industrial Fiber Optics Inc.

I'm waiting to hear from Joe Malloch [meeting early next week] about the feasibility of making a few of his 'inertial motors'. If we use the inertial motors, I'm not sure exactly how we'll drive them. Will line-level be sufficient, or will we need amplifiers? Is there a transistor-like solution that can handle AC so that we can eliminate noise? We will most likely need to use amplifiers. (Time to research low-noise amplification circuits.)

I acquired a license for IRCAM software so that I can start playing with their Modalys physical modeling package.

We'll be using XBee's to transmit the sensor data to the computer. I will need to make a parser for their data packets. I will also need to build some calibration tools for the sensor data.

For output, we will use my Presonus Firebox. To send sub-20hz frequencies, I can bypass the highpass filters DC blockers in the output stage as described by zifcak at em411. Originally I planned to use an AVR for PWM for controlling MOSFETS to drive the motors, but zifcak brought up the fact that I could do the audio->PWM conversion in software on the computer! If I use the inertial motors, then it will be straight audio. 

Vibration Motors

Hi Morgan [...]


In terms of the sensing side of things, you might want to check out:
- Mark Zadel's pliable interface from 2003: http://www.idmil.org/projects/pliable_inertial
- Martin Marier at UdeM was also building a deformable interface ("Martin Marier" <...>)
- Avrum Hollinger in the IDMIL has been building deformable silicone rubber interfaces and using lots of fiber-optic sensing ("Avrum Hollinger" <[..]>)

For vibration actuators you should check out the inertial motors I've been using for the T-Stick and some other projects. They were designed in Vincent Hayward's haptics lab at McGill (he has since moved to France) by his student Hsin Yun Yao. The motors consist of a cylindrical magnetic slug inside coils of electromagnet wire - as AC is passed through the wire, Lorentz force causes the slug to move back and forth. I've posted a few of Yao's publications, and photos of me constructing one of the actuators here: http://www.music.mcgill.ca/~mallochj/morgan/

the eccentric-mass motors you linked to can also be used for more subtle vibrotactile effects if the driving signals are chosen carefully and your circuit allows the motor to be turned in reverse.

[...]

joe

Joseph Malloch
Input Devices and Music Interaction Laboratory
Schulich School of Music - McGill University
web: http://www.music.mcgill.ca/~mallochj

On Fri, Nov 20, 2009 at 4:10 PM, Morgan Sutherland <skiptracer@gmail.com> wrote:
Hello Joe,

I'm doing a project that's based around driving powerful vibration motors (larger than pager motors) with audio signals. Twice now your name has come up while asking about vibration motors, once when talking with Gary and once with Harry. Gary mentioned something about you building your own motors for use in the T-Stick. Harry now just mentioned to me that he was talking with Marije about an unusual design built specifically for handling audio signals, also hand-made. It seems you've put considerable time into this, so I hope you don't mind if I pick your brain a bit. 

I'm looking for vibrations in the dildo-Thumper[1] vibration range capable of transducing complex audio signals (shattering glass, snow crunching, crackers). Ideally I would like to synthesize the response of a brittle object snapping, so transient response is critical. I have been looking at a range of traditional vibration motors like these:

Precision Microdrives sell these, which offer a lot of power (1.6G) or the size: 
http://www.precisionmicrodrives.com/product_info.php?products_id=172&osCsid=1ebea29043ce3c1e1bd440e53bf8ede7 

 Or these which are almost twice the size, but vibrate with 4G of force: 
http://www.precisionmicrodrives.com/product_info.php?products_id=192&osCsid=1ebea29043ce3c1e1bd440e53bf8ede7 

These shaftless motors are not nearly powerful enough

Also this novel design from Namiki, a "vibration speaker":

With the 'shafted' motor designs, however, I'm worried about transient response and, for lack of a proper term, 'momentum smoothing' (or 'resolution'). I was planning on handling this with a multi-modal approach, coupling the motors with small speakers in hopes that the aural feedback would mask the imprecision of the tactile feedback (much like the way subwoofers are used in the cinema to give a convincing-enough tactile dimension to explosions etc.).

So, I'm wondering what sorts of things you've built and whether you can shed any light on the problem. You can follow our research here: http://rgel.posterous.com/ – the project is due in about 3 weeks. A full proposal can be found at http://cart360.msutherl.net/#RGel [2]

Thanks!
Morgan Sutherland

2. [...]

Silicone Gel

I think our best options from Smooth-On are the EcoFlex[1] or the SORTA-Clear[2].

The deciding factor is going to be Shore Hardness. The EcoFlex is "super soft", at 5A hardness while the SORTA-Clear is 18A or 40A.

My guess is that the EcoFlex 5 is going to get us what we want. It's a bit softer than a rubber band and 'water clear'. But, the best thing to do would be to get a sample pack containing a bit of both. 

A useful chart on hardness from Wikipedia:

Bicycle gel seat 15-30 OO
Chewing gum 20 OO
Sorbothane 40 OO
Sorbothane 0 A
Rubber band 25 A
Door seal 55 A
Automotive tire tread 70 A
Soft skateboard wheel 75 A
Hydraulic O-Rings 70-90 A
Hard skateboard wheel 98 A
Ebonite Rubber 100 A
Solid truck tires 50 D
Hard hat 75 D

Vibration Motors

Let's use Posterous for documenting progress. Posterous is nice
because you can post via email. This makes it easy to archive emails
in a convenient-to-access way. When it's time to make an official
progress report, we can review the blog and pull out useful
information.


---------- Forwarded message ----------
From: Morgan Sutherland
Date: Fri, Nov 20, 2009 at 12:11 AM
Subject: Re: Vibration Motors
To: Faiq Hussain


Vibration motors on eBay:

http://cgi.ebay.com/10-Pager-Vibrator-Motor-Bristlebot-Robot-Beam-Part-SCF_W0...

On Fri, Nov 20, 2009 at 12:11 AM, Morgan Sutherland
wrote:
> Just found these vibration-speakers:
>
> http://www.namiki.net/product/sbo/speaker/index.html
>
> vibrator and speaker in one.
>
> On Thu, Nov 19, 2009 at 11:19 PM, Morgan Sutherland
> wrote:
>> Hello Faiq,
>>
>> I just spent some time experimenting with a vibration motor that
>> Shannon lent me. It's this one:
>>
>> http://www.robotshop.ca/solarbotics-vpm2-vibrating-motor.html
>>
>> In my original design, I had planned to use an array of these, but
>> after testing how powerful they are, I don't think they're anywhere
>> near sufficient. I'd say they're slightly more powerful than a cell
>> phone vibrator. This will certainly not deliver the impact I'm looking
>> for. I think we need something in excess of a dildo vibrator,
>> something more like a back-massager.
>>
>> Precision Microdrives sell these, which offer a lot of power (1.6G)
>> for the size:
>>
>> http://www.precisionmicrodrives.com/product_info.php?products_id=172&osCsid=1...
>>
>> Or these which are almost twice the size, but vibrate at 4G:
>>
>> http://www.precisionmicrodrives.com/product_info.php?products_id=192&osCsid=1...
>>
>> They're rather expensive, but I think I can get some funding from the
>> TML to buy them (as others would be interested in using them).
>> Otherwise we could get some motors at Addison and hand-mill the offset
>> cams, but this is probably too much work.
>>
>> I'm thinking that four of those 12mm motors should do the trick. In
>> addition, we could add some of the smaller shaftless motors for
>> transducing higher frequencies (100-200 hz). Think of it like a
>> speaker crossover:
>>
>> 5-100hz > 12-20mm shaft motor
>> 100-200hz > 12mm shaftless motor
>> 300-10,000hz > speaker
>>
>> I think the vibration is going to be the most important part of the
>> project. If it doesn't convincingly mimic other materials, then we
>> have failed completely, so let's do that part first. Will you be
>> around Concordia tomorrow to talk about ordering some parts? I don't
>> have class and it's a good time to catch Elio in the morning.
>>
>