deskthority

Finally, making progress on my Force Curve Gauge v2!



Not quite complete yet, but the rest of the work is circuitry and firmware (easy parts, for me at least ). I'll be adding in motor control, which will let me control up/down/stop so I can do say 10 measurements per graph but also controlling the speed as well.

More pics of the AEL-200-500N test stand can be found on my Flickr https://www.flickr.com/photos/triplehaa ... 6168116869

v1 had a bunch of issues with it (http://deskthority.net/photos-f62/some- ... t7411.html)

• Deflection error when changing directions
• Not possible to move at a consistent speed
• Slow'ish sample rate
• Pain to assemble

Once I finish putting v2 together, I'll post a list of parts I needed to build it. But the main parts include.

• Teensy 3.1 (custom firmware, modified version of my keyboard controller with keyboard parts stripped out)
• iGaging DRO (12") http://www.igaging.com/page18.html, removed the display and hooked up directly to the Teensy
• AEL-200-500N http://www.ebay.com/itm/280890689041. Yes, the picture doesn't seem to match entirely...
• Imada DS2-1 (these are expensive, any sort of digital force gauge will work as long as it has a rapid data out, preferably analog and has a maximum of around 500 gf; the higher the maximum the lower the resolution)

Hmm getting scientific? Can't wait to see the measurements on some of your rare keyboards. I knew you were into vinatge keyboards, but building your own force displacement measurement equipment?!? Wow. This looks like quite a project. Once this beast is operational and tested we can get specific data for our wiki.

Well, good, I'm glad someone has the time and the means to take this project on. My excuse is "kids".

I wanted at some point to build a precision apparatus that allows:
- calibration of switches across an entire keyboard;
- quantification of lubrication, hysteresis, and other subjective part of "feel";
- automated data gathering;
- effect of switch press rate on force curve;
- repeatability.

I would want at least 0.1mm / 0.5g precision and about 4s/key press and release if I were making this myself. 10x better on precision and 4x faster in time would be even better. I want to be able to compare a smooth vintage Cherry MX Black switch with a brand new one and visualize the "roughness".

Anyway, good luck with this. If you finish I'll send you a free NEC blue oval switch if you don't already have one.

I tried a similar mini-project back in 2012.

http://www.clickeykeyboards.com/model-m ... d-and-ibm/

My rig was borrowed from parts from colleagues at work from an undergraduate engineering lab during the summer of 2012. While the setup had precise distance control to the 0.01mm and a digital gauge to 0.001 Newton, ultimately the test procedure was manually intensive for the experimenter. There was too much noise in the data and the results were not worth publishing.

An automated force-displacement rig with a finer force gauge that recorded all the data points automatically and incorporated the specific moments when the switch actuates and de-actuates would be the best way to go.

A lot of variables to control for.. but with the right gear and technical ability, the final results would be able to quantitatively detail exact force curves for different types of mechanical keyboards. The published results would form the foundations for a legendary reference that would help others understand the differences beyond just the common terms of "clicky" vs "mushy".

Oh - and Happy Birthday.

Well spotted! Pour out a glass of something cold and Canadian and think of us! Here's to you man.

Are you trying to Rube Goldberg with that thing? Or is that what it takes to make a proper force gauge?