Hi everyone!
Sorry for the long silence. I received news from my supplier today.
Below is a link to a testing report from Cherry.
https://drive.google.com/file/d/1-5qTDjfSvq3vW2gFCH6C3P6Z3woIvOUH/view?usp=sharing
Apparently the spring weights are within acceptable limits.
Doesn't seem too good. I am not sure how to proceed form here. Any advice is helpful.
TL;DR (on top since is a long read)
Cherry should provide the tests they did and the results MUST include the curve graphs measured on different switches of the batch, since the graph is actually part of the specs. No actual proofs provided by cherry, except a pre-compiled paper.
Oh they want to play with specs, alright! (their response is a joke btw, it's a range of 40g of tolerance)
So their point is that they are in the -/+ 20cn for actuation so this switch respects the specs. Imho they are wrong.
So assuming these are their SPECS of Silent Black:
http://cherryamericas.com/wp-content/uploads/2017/02/EN_CHERRY_MX_SILENT_BLACK.pdf1) THE GRAPH OF THE CURVE, this is INSIDE the specifications file under the
Operation characteristics attribute, it clearly shows around 80/90cn at bottomout so since this graph is in this file, the switch curve clearly has to respect that graph curve.
So even if they wrote -/+20cn for actuation, they didn't specified the rest of the curve, because the actuation is just a point, or said better a short segment on the line/curve, based on their specification the actuation has a range of travel of 1.2mm.
So they didn't specified any -/+20cn for the rest of the curve, so their switch can do -/+20 for the actuation range, but must respect the curve on the rest of the travel because is not specified otherwise (pre-actuation travel/after-actuation travel), so
OUT OF SPECS.
I know this is illogical because this is a linear switch, but their specifications are not considering this case, since they want to play with the specification document, based on their specifications the curve is out of specs.
2) Now let's assume point 1 is not valid for any reason (which i doubt) let's still talk about
THE GRAPH OF THE CURVE, see this image:
Since this is a linear switch and since this is the graph on their spec-file, i can assume that by translating the curve of the graph by -20cn or by using the silent red graph, the switch keypress curve should be over that curve, but is not:
case 2A) MOVE THE CURVE OF -20cn: We just
translate the curve at a -20cn from the graph curve on the spec file. If we assume an Actuation of 40cn, the bottomout should be 70cn, which is around 71.4g ( in fact their curve shows around 92grams at bottomout).
Many of us calculated the bottomout in poor ways yes, but our result was definetely lower than 71.4g for bottomout and i'm pretty sure that there is someone that can accurately calculate the bottomout of that switch (maybe Haata?). Pretty sure the curve of our switch will be out of specs once you measure it correctly, even after taking into account the -20cn.
case 2B) SILENT RED CURVE: Since you usually can't just translate the curve, let's take into account the
silent red curve. I toke the silent red graph (which is probably the spring we have inside our blacks). That switch bottomout is marked at around 65grams (63/64cn), so this is around 6/7 grams lower than the minimum that cherry can do on the silent black if they want respect their specs.
So VicaryG you should ask for the result of the tests which MUST include the graph absolutely, the test they did must include the CURVE of our silent black switch, since again the curve is inside the specification graph.
Please correct me if you see something wrong, not your fault on this VicaryG, nothing you can do to directly to fix this clown fiesta between Arrow and Cherry, but at least try.