Haptic device. any idea ?

[snowman]

dear friends,

I was trying to make a pantograph using my sherline cnc mill. the idea was moving a pointer over the object and meanwhile let cnc to carve it in any material. usually its mechanical stuff but I wanted to make electronic one.
while googling I come up with haptic devices. I always needed something like this. its a kind of mechanical arm. when you move its end point it calculated the coordinate of end point. all the mechanical stuff and electronis are okay but I wonder how can I sense the position of joints in a very sensitive way ? imagine your own arm. I need a sensor or sensor mechanism to sense position of every joint.

any idea how can I achieve this ? what kind of sensors can I use ?

best.

[Signal]

AykutCanturk, give us more input data. Are we talking about 2D? What sizes? Precision? For 2D think about pen tablet digitizer.
For arm expected by you I can not imagine solution simpler than axial potentiometer. Although I have no experience in robotics.

[pickit2]

for 2d search pantograph engraver these also have limited z movment.
main problem would be speed, you need to be slower than your cnc motors.

another thought you might try is convert movement to g-code via a mod to a jog pendent.

[snowman]

hi,
I'm talking about an like this to get coordinates.

http://www.canadianmetalworking.com/wp-content/uploads/2013/02/outsourcing-pic-1.jpg

another example is surgery robots. how do they sense small coordinate changes ?

I found someting called kma199 angle sensor but there's no sample application on google.

https://www.digikey.com/product-highlights/us/en/nxp-semiconductors-kma-sensor-systems/1168

also I read about interference issues on readings that makes everything complication. briefly I could not find any good source about that.

Is can get readings by usb port to computer I can write an application to combine them and even generate g-code

still curious about the sensors on surgery robots it must be very clever and simple technology...

[CocaCola]

Digital linear and rotary encoders, they come in all shapes and sizes and many very accurate ones can be had for cheap now...  Consider you cheap made in China digital calipers, they hardly cost any more than analog ones and are accurate to +/- a few 100th of a mm...  The Law of Cosines will come in handy, if you make each joint a triangle using fixed distances or linear encoders, for example...

For many applications like surgical robots there is generally no secondary location feedback, it's simply calculated based on the steps of the actuators after it's been zeroed out prior to use...

[snowman]

Digital linear and rotary encoders, they come in all shapes and sizes and many very accurate ones can be had for cheap now...  Consider you cheap made in China digital calipers, they hardly cost any more than analog ones and are accurate to +/- a few 100th of a mm...  The Law of Cosines will come in handy, if you make each joint a triangle using fixed distances or linear encoders, for example...

For many applications like surgical robots there is generally no secondary location feedback, it's simply calculated based on the steps of the actuators after it's been zeroed out prior to use...

there are angular versions of them on ebay. thanks my friend. that was helpful thought.

[CocaCola]

there are angular versions of them on ebay. thanks my friend. that was helpful thought.

Yeah there are 'digital protractors' that can be had for about $15 delivered on Ebay that resolve to 0.1 degree, not extremely accurate but fun to experiment with...  Many of these will also have a digital out port that is sometimes hidden on the board (one size fits all models PC board) under the plastic housing, other times it's just a slide cover if they want to advertise it has a digital out port, makes interfacing them to computers or micros dead simple...

[LabVIEWguru]

I don't know if this will provide the resolution you are looking for, but I used this exact circuit a few years back because it was $CHEAP$

[solutions]

Hey Labview - can you microstep that encoder? LOL

I like it.

Seriously,though- for a 200 step motor,does that mean 400 PPR?

[snowman]

I don't know if this will provide the resolution you are looking for, but I used this exact circuit a few years back because it was $CHEAP$

my friend I already forgot this step motor option. that is great for my problem. I believe resolution would be more than enough my using gears already exists in old printers, which I have a lot. also using step motors allows easy mechanical assembly.

not I need to find a way to convert this information to coordinates by writing an application.

thanks for great ideas
 

Posted on: June 29, 2014, 11:25:02 11:25 - Automerged

Hey Labview - can you microstep that encoder? LOL

I like it.

Seriously,though- for a 200 step motor,does that mean 400 PPR?

gears my friend.. gears. if you consider helical gears in printers resolution would be wonderful. of course there will be a backlash problem in mechanical design but software can compensate.

[solutions]

Gears are not linear devices.

They also create a LOT of inertia - multiplies by the ratio, in fact. It'll also multiply the cogging torque.

For a pantograph, get one of these: http://www.ebay.com/itm/Microscribe-3dx-/121373560368?pt=LH_DefaultDomain_0&hash=item1c426d8e30

[Ichan]

AS5048 14-bit rotary position sensor with digital angle, the datasheet say the maximum system accuracy is 0.05 degree.

-ichan