geekhack
geekhack Community => Keyboards => Topic started by: Photoelectric on Wed, 22 May 2013, 01:24:48
-
Just wondering if anyone's done a calculation of the work (Force x distance, in the Physics sense) required to actuate each of the Cherry switches? And maybe 1mm past the actuation point too. It'd be a simple integral if we had the exact force equation, but having some data points along the force diagram curves could still give a decent approximation.
If not, I'll have to think about it and try to approximate it myself based on looking at available force diagrams. Having some actual numbers for those curves would be great though.
(http://peterhallam.com.au/wp-content/uploads/cherry-mx-brown-force-diagram.jpg)
Also, what is the bottom curve--the return path?
It's obvious that the force required at the actuation point does not tell the whole picture. For MX Clears, one has to overcome a bump of over 60cN to get to the actuation point (that requires 55cN or so, right?) So it's quite unlike MX Blacks that start out at 40cN and gradually get to 60cN actuation point with linear acceleration.
(And there are more nuances to think about, but this would be a simple start)
-
Given the unprecise nature of the force graph to begin with..
I'd say a triangle + rectangle is plenty accurate :D
-
lost :confused:
-
Given the unprecise nature of the force graph to begin with..
I'd say a triangle + rectangle is plenty accurate :D
Hm well I'll probably approximate some point values along the curves and make a summation script.
-
It'd be a simple integral if we had the exact force equation
Nothing in nature really ever is.
-
Well yeah... But that's what Physics does best: spherical cow approximations.
-
Well yeah... But that's what Physics does best: spherical cow approximations.
The Koreans will have a better empirical force / travel graph since they invest in the equipment to measure it. Alternatively, go find a prof / lab at your university that has the goods.
I guess you can also do it (more) discretely with some weights and a caliper / ruler.
-
I don't own most of the switches, so I'll have to do with the best diagrams I can find for a first-order approximation.
-
We did this for a couple of Topre switches a while back after we got some lab-quality measuring equipment in, but we haven't gotten around to doing Cherry switches yet. You can use the Rectangle Method (http://en.wikipedia.org/wiki/Rectangle_method) to approximate the integral directly on the graph you posted. Using N between 20 and 40 will yield results accurate enough to compare different switches. This can be done in a spreadsheet rather quickly.
Better accuracy could be attained by applying Simpson's Rule (http://en.wikipedia.org/wiki/Simpson%27s_rule), but it's a little overkill...
Just looking at the characteristics of different switches you can guess that Topres and several varieties of ALPS require more work before actuation than most Cherry MX switches, but MXes requires more work after actuation point. The overall amount of work done while typing on a particular switch in a certain style might be interesting data to consider if you're trying to minimize effort or fatigue. You might find something counter-intuitive. E.G. A 35g Topre switch may end up requiring more total work than a 45g MX switch if you're good at not bottoming out.