characteristics of hardware: testing linearity of the sensor

Post questions, comments and feedback to our 3Dconnexion Windows Development Team.

Moderator: Moderators

Post Reply
harmslijper
Posts: 6
Joined: Thu May 28, 2009 6:08 am
Location: rotterdam

characteristics of hardware: testing linearity of the sensor

Post by harmslijper »

Hello Everyone,

I want to test the the linearity of of the SpaceNav. For Fz translation I am going to place different weights on the sensor en see whether the linearity is good and/or whether there is strong hysteresis. For the other directions and rotations it's going to be much more difficult to test this. Therefore I had the following questions: What kind of strain gauges are mounted in the SpaceNav, are there six of them, and are they all the same. Who makes the strain gauges, and where can I find the specs. of them.

Thank you for your help,

Harm
jwick
Moderator
Moderator
Posts: 3340
Joined: Wed Dec 20, 2006 2:25 pm
Location: USA
Contact:

Post by jwick »

We don't use strain gauges. There are springs in there.
harmslijper
Posts: 6
Joined: Thu May 28, 2009 6:08 am
Location: rotterdam

Post by harmslijper »

Hi Jwick,

Oke, springs, but the springs must have characteristics, that cause some change in voltage, which is translated into a signal send to the computer. So if I rephrase my question, what are the characteristics of the springs and how is thedeformation of the spring translated into the signal that I get as output from the SpaceNav.

thanks for your reply,

Harm
jwick
Moderator
Moderator
Posts: 3340
Joined: Wed Dec 20, 2006 2:25 pm
Location: USA
Contact:

Post by jwick »

No, the springs aren't used that way. The springs hold the cap up. There's an arrangement of LEDs and sensors that do the sensing of the displacement. Take it apart if you like -- it's quite beautiful inside (that will obviously void your warranty and most likely make it inoperable).

To try to answer your underlying question, the device isn't a laboratory measuring instrument; it's a computer input device. Each device will vary slightly. If you calibrate one sample, your calibration will unlikely be absolutely perfect WRT another sample or another model. If you calibrate one, your results should be fairly reproducible with that device. It's a complicated assembly though. I wouldn't expect it to be completely linear in all directions through its entire range of movement. It should be fairly close near the rest point. You'll have to measure it with the experiments you propose.

If you write a paper on it, let us know what you find.
harmslijper
Posts: 6
Joined: Thu May 28, 2009 6:08 am
Location: rotterdam

Post by harmslijper »

Hi Jwick,

Interesting, I just ordered two more spacenavs so I can take one apart. Any suggestions with regard how to disassemble one, without destroying it.

When I calculate the calibration curves, shall I email them to you? what is your email address?

cheers,

harm
MMTE
Posts: 7
Joined: Sat Jun 22, 2013 11:55 am

Re: characteristics of hardware: testing linearity of the se

Post by MMTE »

I'd like to see the graphs :)
Post Reply