Instructables

DIY Bend Sensor (Using only Conductive Bags and Masking Tape)

Featured
Picture of DIY Bend Sensor (Using only Conductive Bags and Masking Tape)
So, after lurking around in the shadows for years I've finally gotten round to making my first step-by-step instructable.  *yay*

A little background as to how this whole idea came about.

Back in 2009, I needed some bend sensors for a sign language translation glove I was making for fun, and the commercial ones that just came out were over my budget considering I needed at least 10. 

Then I came across plusea's DIY bend sensors on Instructables.
It was what I was looking for but I had one small problem. Well, a few problems actually. (Not with the 'ible of course)
All the different instructables required the use of neoprene, or at least conductive thread and conductive fabric.
I could get the anti static bags locally, but getting the above items would mean the total bill (after shipping and currency exchange rates and tax) would consist of mostly that, and not the cost of the items.

So I decided to see if I could just experiment with the conductive anti static bag(although in all honesty I tried matching it with a whole bunch of other stuff; like carbon paper). It took 2 weeks of playing around during my free time before I finally had a decent build. (Still works to date! )

So for all of you who has a tight budget or like me, can't get easy access to items in the states, this is for you.



 
Remove these adsRemove these ads by Signing Up
MartinVas3 months ago

Hey! I can't seem to find any Velostat or Linqstat or black conductive bags (as you have used here) in any hardware store. I live in Europe and can't find an online shop to order some also. Any help on a substitute material that is easy to find or some shop that ships it worldwide? Thanks!

tonll (author)  MartinVas3 months ago
Hello!

UK Farnell has it.
http://uk.farnell.com/vermason/bp3eg/bag-conductive-100x150mm/dp/522740

You can check to see if your country is listed under "Change Country" - Europe
Hope that helps!

tonll
MartinVas tonll3 months ago

Do you think that this might work (the bag in the ebay link provided) ? I've found some on ebay, and can be shipped to me.

http://www.ebay.co.uk/itm/BLACK-CONDUCTIVE-BAG-102...

tonll (author)  MartinVas3 months ago

It does look similar, and it'll probably work.

But there's no datasheet, so I'm not sure if you'll get the same results since the resistance could be different.

Cheers

tonyd13 months ago

what are the range of values will u get from this sensor

tonyd13 months ago

what are the range of values will u get from this sensor

Robotus7 months ago

Hi tonll / Michelle! Quick question here: I have some of the silver-gray, semi-transparent anti-static bags that are also common, do you know if this sensor would work with those too? You specifically say to use the heavy black bags (Velostat, which I'm not as familiar with), so it sounds like it won't work with the silver-gray bags.
So, I'd need to use something that looks like the image here:
http://sg.element14.com/multicomp/006-0003f/black-...

and the bag in the image here wouldn't work?? :
http://sg.element14.com/multicomp/mc35830/bag-stat...

tonll (author)  Robotus7 months ago

Hello!

Hmm. I'm not sure if it'd work as I haven't tried it on those anti-static bags.

2 main important points to note is that the velostat bag is made of a electrically conductive material (carbon), as well as the surface resistance.

A quick test that can be done to see if the bag you have conducts, is by taking a multimeter and measuring the resistance from one point of the surface of the bag to another.

Changing the distance of the 2 probes from each other should result in a proportionate change in resistance reading. As with the velostat bag, readings tallied with the datasheet's specs of 100K ohm ++ to values in mega ohms.

Which in turn, affected the size I chose for the final dimensions used. (due to the resistive readings)

Although the silver grayish bags have a surface resistive spec, the build of it uses polyester, and the datasheet says nothing about it being one of those conductive polyester films that has carbon black in it.

I'd suggest trying the above test method.

Hope that helps!

Cheers

Robotus tonll7 months ago

Thanks for the quick reply! I figured it had to do with the surface resistance, thanks for confirming and sharing the testing idea. I'll try it out and see if it works.

Wickedweed10 months ago
Hi. Just wondering if the resistance value is more linear than logarithmic?
whr do i will get this conductive bags ?
jestern1 year ago
Hi there just for the sake of trial what is the range of values that you get?
Also i have it working, in the way that if i press the surface it goes up the voltage, but only if i press there not if i bend it... what can it be worng?
Alexisa91 year ago
Hello

With the voltage divider, would you beable to show a picture of the circuit from underneath or tell me, because I do not know where to solder together for current to flow through in the right directions. I can't read the schematics well.
Thankyou
Do you know some material to replace Velostat?
Alexisa91 year ago
Hi guys
I just wanted to ask, what level ohm or kohm resistors are being used for the voltage divider I have 100ohm resistors would they be sufficient for the voltage divider, I will be using arduino to do this project.

Also, I am making a robotic hand which is controlled by the flex sensors connected to a glove, would this type of homemade flex sensor do the job well and be sensitive enough.
It is for a competition and will be doing more than just the glove.
tonll (author)  Alexisa91 year ago
Hi,

Please refer to step 5 for the diagram and values of the resistor for the respective voltage.

Most arduinos are 5V systems. So you'll probably need 2x 10k Ohm resistors.

Yes. It's sensitive enough. If you read the introduction, these sensors were originally made for use with a glove for a sign language translation application.

Hope that helps!


Cheers
Great DIY ...i def want to make these ..... 1 question tho would i need to use the voltage divider circuit you used for an arduino application of these sensors???
tonll (author)  ghostridernyc1 year ago
Hello!
Thanks! =)
Yes, you do. Depending on the arduino you're using, please following the resistor values for a 5V or 3.3V arduino version.
Cheers
mhussar2 years ago
sorry for the newbie question. Just getting started. I'm using arduino. I'd like to try this with it. How do we know which of the wires from the flex sensor goes to VCC and which goes to GND?
tonll (author)  mhussar2 years ago
Hi, you'll need to connect it to a voltage divider before you can connect it to a microcontroller.
Please read through the entire instructable, you'll find helpful schematics on step 5 that states that the sensor is just like a variable resistor, and how you can perform the interfacing.

Hope that helps!
mhussar tonll2 years ago
Hi Tonll,
Thanks for your reply. I did read the through the entire instructable and I saw step 5 and the illustration for 5V. As I said I am a newbie so perhaps I'm misunderstanding something (although I've been reading a ton on electronics as well as beginning the MITx class EE). I used the Flex Sensor from Spark Fun and it too needs to be wired through a voltage divider. Your schematic in Step 5 looks like the schematic I used for the wiring of the Spark Fun Flex Sensor. So I thought I knew what I voltage divider was. In the Spark fun Flex sensor there is a lead that goes to VCC and one that goes to GND. My question was which of the leads in your Flex Sensor goes to VCC and which goes to GND ? Maybe it doesn't matter or maybe I'm missing something?

Thanks again!! and great instructable
tonll (author)  mhussar2 years ago
Hi mhussar,

It's great that you're picking up electronics and programming. I hope you have fun with the many adventures I'm sure you'll have. =) Makes life all the more interesting.

From what you have expressed in your post above, it seems to me (I could be wrong) that perhaps there's been a fundamental misunderstanding of what resistors or maybe just the flex sensors are, and its role in a voltage divider. If I may try to explain as plainly as possible from now on, so hopefully it clears some misunderstandings and make you go AH-HA! or OOOOH! out loud as you're reading.

For that, here's some background references for what I'm about to explain.
Since you're coming from a previous understanding of the Flex Sensor that Sparkfun carries;
◆Tutorial for the flex sensor I assume you're referring to:
http://www.sparkfun.com/tutorials/270
◆There's a great link on explaining what voltage dividers are here:
http://www.sparkfun.com/tutorials/207

"I used the Flex Sensor from Spark Fun and it too needs to be wired through a voltage divider."
Actually, The reference circuit given in the  datasheet uses the sensor as a resistor to form part of the voltage divider. Because the flex sensor is just that. A resistor that varies when it bends. The same concept goes for the DIY Bend Sensor.

"In the Spark fun Flex sensor there is a lead that goes to VCC and one that goes to GND."
Back to my previous point. It's a resistor. Resistors have no polarity. The flex sensor's (carried by Sparkfun) leads are actually interchangeable. You don't actually connect one to VCC and the other to GND. It's used as R1 of a simple voltage divider circuit. So one goes to VCC, and the other lead to another resistor of a chosen value (pre-calculated) that connects to GND. In which case you then take the point in between the 2 resistors, and connect it to an ADC pin. (Refer to the voltage divider equation at this point)

To help visualize and maybe simplify things.. I made an image.
I really hope that helps.
http://www.instructables.com/files/deriv/FFQ/NG9W/H5TUCAFL/FFQNG9WH5TUCAFL.THUMB.jpg

So when you asked this before..
"My question was which of the leads in your Flex Sensor goes to VCC and which goes to GND ?"
Actually... I answered your question.
"the sensor is just like a variable resistor"

I really hope at this point, you've understood.
If you still don't, please point out which part you don't understand, and we'll try again from there.

Cheers.
MISC.png
mhussar tonll2 years ago
Oohh and Ahh.
Totally get it. Not sure where I picked up that Flex Sensors have polarity. Maybe I mistook the brown bit on the right wire in the Spark Fun Flex Sensor as indicating polarity. However, as I am beginning to understand in general, once one grasps these concepts more fundamentally, mistakes like that should be less likely to happen. I already knew that resistors don't have polarity and also how a POT works as a variable resistor creating a voltage divider. Just didn't get that the Flex sensor was essentially the same thing until you explained it. Thanks so much!!!
tonll (author)  mhussar2 years ago
Hey! YAY! You got it! Hahahah. That's great!
Glad I could help. And yeah.. gain as much experience as possible while learning. You'll find that doing will let you appreciate all the theory even more. It did for me.

Cheers
BudBump2 years ago
Your design would have a limited use-life and the fatigue reading would degrade after a threshold number of bends is reached. Better materials (plastic dip perhaps) to take the place of the tape would increase lifetime use of the sensor. I really like the simple approach you used to address a need. You prove we need not be slaves to commercial suppliers when we really CAN do it our self. I hope to see more instructables authored by you.
tonll (author)  BudBump2 years ago
Hello. Hmmm. Maybe. I don't exactly have a technical specification and test results that prove that point about the fatigue, but it has done its job for the time frame of at least 8 months that the project lasted. (the occasional taking out and playing with it)
I do like your idea of plastic dip. And yes, I agree that better materials will really take this sensor's life time and perhaps performance to the next level. But at that time, I really had to make do with what I had lying around. And this instructable was the result =)

Thank you for the motivation. I really hope to be submitting more instructables very soon.
it doesnt matter if the life is low atleast its like 10 times cheaper!!and could ya tell us ur idea(with plastic dip?)
walshlg2 years ago
THis is really cool. One thing I noticed is what looks like some "fatigue" where the readings seemed to drop with time. Was that a problem in regular use?
tonll (author)  walshlg2 years ago
Hello! Thanks!
Perhaps you're referring to the part where I tried bending it further in the middle and the values didn't change. The thing is, there's a sweet spot to this bend sensor and that is where the jumper wire is.
It responds to the bend at that junction the most.
If there really is a fatigue.. I wouldn't be able to hit the max value and that was kinda based on the min and max values I was getting.
I've had some bend sensors that I made 2 years back and it still works with the same program and gives out the same reading. Perhaps there may be chance of fatigue reading on very constant use over a long period of time. But as of yet, I have not had the chance to do a thorough test like that.

Thanks for the keen observations. You made me ponder about the possibility of fatigue over regular use. =)

And thanks for the kind words dude!
Elgonzalo2 years ago
I'm spanish and my english is not good enough.
That's a nice instructable, congratulations!!, but can you tell me an example where you use this sensor? I have a microcontroller and i'm experimenting with it.
Thank you very much
tonll (author)  Elgonzalo2 years ago
Hello! Your English is fine =)
Thanks for the compliments!
I initially used it for a sign language translation glove. (alphabets) along with an accelerometer for some letters that required motion. (like the letter 'j')
There is an accelerometer on the OLED board that was used so I could just use that. Training (for the alphabet recognition) is stored in integer arrays and downloaded to the onboard Dataflash chip, and then just use the screen and audio as the User Interface. (Results on screen, audio reads out the words) Audio comes with the OLED board too. I used 10 of these in the end so I could have 2 gloves. One for each hand. They're all wired to the BCore100 (it's got 16 ADC channels so it's more than enough).

Does that help???
You could use it for simpler projects. like for gestures. 2 bend sensors is enough. (forward, backward, left, right.) things like that.
Elgonzalo tonll2 years ago
Yes, very good help, thanks!!
I use a NXP,LPC-1768. I'll use the sensor in a simpler project (jaja, i'm leaning now to program) in summer.Thank you for the examples and sorry for the late reply!!!
*learning :)
Ralphxyz2 years ago
Great, wonderful, marvelous, loved it, thank you. Just what I need for a upcoming project.

Ralph
tonll (author)  Ralphxyz2 years ago
Thank you!, Thanks, That's nice of you, Great!, You're welcome. x)
I'll be interested to know what you did with it!

Cheers
Michelle
chrwei2 years ago
Could you post some theory on how/why this works, or link to one of the instructables you mention that explains it?
tonll (author)  chrwei2 years ago
Hello chrwei, hahhaha.. I'll try to explain best that I can.
The concept is really simple actually. These bend sensors are essentially variable resistors. Like potentiometers, they have a range of resistance values.
If you check out the schematic that I did in one of the steps, it should help explain what the sensor really is.

The conductive bags that I used have surface resistance. (Surface Resistivity Max: 0.1Mohm/sq)
If you were to just take a bag, and use a multimeter on resistance measurement mode, try placing the probes with different distances from each other on the bag, you'll notice different resistance values.

The middle piece ( the larger one) that was sandwiched between the 2 thinner ones serve 2 functions.
1: It's only conductive on one side, so I had to fold it to get the total relative resistance of both sides of the sensor. During a bend, there will definitely be a difference in value. (one side bends more than the other)
2: It helps provide the difference in resistance compared to the thin ones, which helps with the overall bend value. I really should be able to explain this point but somehow, I can't. At least not in a way that would make electronic sense. Perhaps someone else would be able to explain it simply.

I hope this helps a little bit.

Cheers
Michelle
This is excellent! Great build of a useful sensor out of some parts most of us just have within reach (except for the Blazing Core).
tonll (author)  piles.of.spam2 years ago
Thank you so much!
Hahhaha yeah. But the BlazingCore is great for writing up the U.I in a jiffy to show the test results.
Again, thanks for reading and for bothering to write something nice. Much appreciated! =)