Is everyone on crack?

[Daniel Beardsmore]

I came across Cherry part MX1A-1GFD for sale, and it got me curious, so I started digging. This is not a complete list of what's out there, but it's enough:

• MX1A-1GFD, MX1A-1GCX
• MX1C-C1NW, MX1C-D1YW, MX1C-F1DW
• MX1D-E1NW
• MX2A-21JN, MX2A-C1YN, MX2A-F1WN
• MX2C-D1YN
• MX2D-21JW
• MX3A-E1NW
• MX3B-D1DW, MX3B-E1RN
• MX4A-D1JW
• MX4D-01NW
• MX5B-11YW, MX5B-11YN, MX5B-11GW, MX5B-11GN, MX5B-11RN, MX5B-01JW, MX5B-01DN
• MX5C-21DN, MX5C-21YW
• MX5D-01DN, MX5D-01JN, MX5D-11JW, MX5D-11RN, MX5D-21RW, MX5D-21YN, MX5D-F1JN

MX1A-_G__ denotes custom parts in Germany, so that G is perfectly legitimate. I may know in the near future what it means.

One site lists all the MX5A types as SPST NO (also known as form A), but does not list this against the MX5B types. The "A" in MX1A officially denotes SPST NO, but it could be A for form A, or A for the first sub-type. I'm not aware of SPST NC MX switches, but nor was anyone aware of the "dwarfism" type until very recently, so oddities are possible.

Even so, are these real? Or do these companies get their jollies from making up nonsensical part numbers?

Cherry US say no, but MX2A is plausible as we went from MX1A (most things) to MX3A (silenced). MX5_ and MX_C are specifically permissible according the 1988 chart, for special parts, but surely there cannot be that many special types? Special part numbers are generally not made publicly known.

I have to wonder if these surplus stockists are completely demented.

[seebart]

Bad thread title Daniel! Otherwise valid content.

[Daniel Beardsmore]

I've looked at the master list of Cherry part numbers from Potis Technologies (theirs is in a PDF) and there are over 2000 MX types, from MX1A-01DN to MX5D-F1YW. It's not gibberish: this is either deliberately and carefully deceptive, or legitimate. The third character stops at 5 (which is legitimate) and the fourth character stops at D.

The rest of the keyboard switch list is short, and they don't mention MX1A-1G* types at all:

M81A-0100
M81E-0500
M81E-05R0
M81E-0600
M81E-1500
M81H-0500
M84A-0100

ML1A-11JW
ML1A-11NW

For Cherry MX, they seem to have written a program to generate every possible code, as there's no way that they all exist. Some of them may be real, but there's no way to know which!

[Engicoder]

Part codes are funny things. Often several different part numbers may refer to the same part. A few of the things I have seen trailing characters in a part number refer to:
- packaging, i.e. a box of 100 would have a different part number than a box of 1000.
- lead material or plating
- a rating or qualification i.e. samples from the production run have been qualified to a given standard
- production plant

[Daniel Beardsmore]

MX[1-5][A-D]-[0-9A-F]1[RGYJWDN][NW] would be 4,480 types. However, there's only 2100 of them. I don't know W means as a feature, but it only appears as WN, never WW.

The real parts are listed separately in the PDF, which is not sorted in order. MX Brown (MX1A-G1*) and Red (MX1A-L1NW) are both outside of the automated list, which appears to treat the type position as hexadecimal, hence stopping at F and omitting brown and red.

These are the real ones so far (with some details from Cherry):

MX1A-1GFD
MX1A-1GAP
MX1A-1GBR — 179 CN PCB MOUNT N/O
MX1A-1GCX — 350CN W/PINS

That last one must be Cherry MX Kal El Black. I wonder how that compares to my lump-of-tar things I got from China (the "WTF switches").

I am hoping to get both of MX1A-1GFD and MX1A-1GCX. Looking at a German supplier, but now investigating their RAFI part numbers.