Jump to page content

Cherry questions

Contents

The following page details outstanding questions regarding Cherry keyboards and keyboard switches.

Additional contacts

Printed material and parts

General

Country codes

Which country codes were used in custom part numbers? The known or suspected codes are:

N
Japan, known from MX1A-0NNN (the standard Japanese MX type) as well as Hirose M8 (M83A-0N00 and M83S-0N00) and M85 (M85A-0N00); by comparison, the series specific to Japan (MD and MJ) used keystem code 1 instead
G
Germany, known from various MX and M9 custom types (MX1A-1GAP, MX1A-EGEQ, M94A-1GBR, etc)
W
In theory, the custom US-made MX types should have part numbers in the format MX1A-xWxx (where W denotes Waukegan) but no such part numbers have been recovered thus far

Other letters were allocated, such as: D and H for Germany; B, S and P for Brazil; U, K and E for the UK. Were any of these other letters used for switches? U seems to have been used for UK-made keyboards (“UB” instead of “B”), but the Cherry-made Hi-Tek High Profile keyboards were also UB (UB80-01AA being the only known model).

US operations

Early US history

Cherry’s Meet the PRO brochure (for model B70-05AB) contains the following statements:

The company that first introduced gold crosspoint contacts to snap action switches for low energy solid state circuits. Then, applied this same innovative gold crosspoint technology to keyboards back in 1967.

To which product does the second sentence refer? Is the brochure describing Cherry S31 Series? These are dated back as far as 1968. S31 was used in keyboards, although it’s not clear that they were ever intended to be used as such.

The 1973, 1974 and 1979 Cherry keyboard and switch catalogues suggest that the M6 keymodules came in around 1969 or 1970, but the M6 patent was filed in June 1971, and the oldest discovered written material for keyboards also dates back to 1971. So far, it seems that the mechanical keyboards were introduced around 1971. The reed switches are a little older, with the patent being filed in 1970.

Gold crosspoints and Western Electric Alloy #1

Did Cherry license the gold crosspoint technology from Western Electric? Western Electric filed the patents for it in 1930, and these patents are likely to have long expired before Cherry adopted gold crosspoint technology. Cherry did use Western Electric Alloy #1 for the crosspoints, as documented in their catalogues. (Cherry never used the full term “Western Electric”, only “W/E”; a Hi-Tek brochure gave the name in full, allowing it to be identified.)

Reed switches

Cherry briefly advertised reed keyboard switches. The 20x part numbers would mean that these switches were never offered for sale directly. (It is interesting that Cherry would list internal part numbers in promotional literature.)

M4/M5/M6

Keycaps

There seems to be some confusion about keycap production. The pre-DIN (12 mm) keycaps were distinctively Cherry, both the US and the German production. A lot less is known about the DIN-era keycaps from the US: some customers (e.g. Tektronix, Wyse) seemed to get Comptec keycaps instead of Cherry, but it seems there may have been in-house DIN (8 mm) keycaps too.

Do keycap types “KB II” and “KB IV” ring any bells? These series names appeared in a magazine advertisement with neither photographs nor details, and are not mentioned by name in any discovered catalogues.

Solid-state capacitive

When was solid-state capacitive introduced? There is conflicting information that suggests both 1977 and 1979.

The DIN-compliant type was named Series MF. Did this name apply to the pre-DIN types? Did those ever have a series name at all, considering that they were not discrete switches?

Membrane

What was the source of the membrane sheets for US FTSC keyboards? Apparently they were not imported from the membrane production line in the UK.

US-made MX

Consolidation in Germany

Clarification and expansion is still needed regarding the processes and dates of moving to German manufacturing and the spin-off of TG3, including when each specific line (M4/M5/M6, MX, and full keyboards) closed, to help clear up understanding of all the apparent production anomalies that show up from that era. It would be good to corroborate this with Tom Giles.

For example, collector UncleFan has a Tektronix keyboard made by Cherry in the US, from 1987, with German-style switches in it (M-shaped movable contact leaf, and gold-plated wire contact points). Having a clearer idea of the timeline of moving to German production would help clear up whether these examples indicate a change in US production (adoption of German-style parts) or a transfer to importing German-made switches. (Potentially it could be both: German-made metal parts imported for use with US-made springs and plastic parts, or some other combination based around imported switch contacts.) We know that TG3 was making B70-4753 by 1990, and by this point they still had supplies of M5 switches for space bar, while using M7 for the alphanumeric keys. The most recent B70-4753 discovered prior to that was a Hirose-made example from 1985. The most recent Waukegan-made keyboard discovered was a Tektronix keyboard made in 1987, a year after TG3 was founded.

Keyboard part numbers

Is anything still known about the format of Cherry keyboard part numbers, e.g. KXN3-8451, KFN3-8358 and B80-40AB? The X in KXN3 seems to mean MX, and the F in KFN3 seems to denote MF solid-state capacitive (possibly “foam pad” or “foil disk”). Some models had schematic codes, but the rest are not clear.

For all the details acquired so far, see the Cherry keyboard codes page.

German operations

Model numbering

Does the “G” in the model numbers of German keyboards (e.g. G80-1000, G84-4400) denote “German-designed” keyboards or “German-made” keyboards?

Part numbering

What was the assignment method of German custom part numbers, e.g. M94A-1GBR, MX1A-1GCX (MX super heavy, last made in 2016)? Was this arrangement schematic (where letters designate specific variations), mnemonic (the letters stand for the customer name or project) or simply sequential (e.g. AA, AB, AC, AD etc)? The tendency for the first letter to be low in the alphabet suggests sequential assignment, but it is also strange that—of the few codes recovered to date—there is both an MX1A-1GBR and a M94A-1GBR.

Switch contacts

Thus far, we know that the switch contact process was as follows:

  1. Cherry in the US implemented gold crosspoint contacts using Western Electric’s design and alloy, with contact prisms with shoulders to support the assembly machinery
  2. Cherry in Germany could not find a factory to produce such prisms, but they did find a company (whose name has not been provided) who could manufacture a similar design, triangular in cross section, using a European alloy; these became contact types 1 (gold alloy: M7, M81, M91) and 2 (silver alloy: M82, M92); possibly the alloys were supplied by Wieland Edelmetalle, as M9 used Wieland copper alloy for the terminals
  3. Switches made with Western Electric Alloy #1 contacts (as used in the US) were designated contact type 3 (German-made M93, and Japanese-made M83)
  4. To save costs, Cherry in Germany then introduced gold-plated wire for the contact surfaces, with a corresponding reduction in capacity (current capacity reduced from 100 mA to 10 mA, and maximum voltage reduced from 28 V to 12 V); gold-plated wire contacts were ready for MX, and designated contact type 4 (M84 and M94).

However, there have been two examples of switches discovered with contact designation “1” and gold-plated wire contacts. Specifically, M81F-1500 (stored in unofficial packaging dated to 2016, several years after M8 was discontinued) and M91A-11NN, found in Lebanon in official Cherry Mikroschalter boxes, albeit with corrections made to the part numbers.

The question thus is: were M81 and M91 switches ever made with higher-rated gold wire contacts? In other words, if we see wire contacts, does that always mean contact type 4 (M84, M94) or were there also M81 and M91 production runs with wire contacts?

Keycaps

Were German double-shot keycaps made in-house or outsourced? Who would GMK have acquired the tooling from, when Cherry ceased production of double-shot keycaps: Cherry, or the contract manufacturer? [Günter seemed to indicate that Cherry formed a joint venture with another company to produce keycap tooling, and bought out this company after five years, becoming the owner of the keycap tooling. Presumably this is what was sold to GMK after production was discontinued some time after 2011, although there is no record anywhere of these announcements.] GMK have utterly refused to respond, and GMK customers also refuse to help.

M7

The numbering process for Serie M7 starts with the illuminated type (M71). Double-pole illuminated (M72) was skipped (it may have existed, but not in the 1982 catalogue). Standard single-pole momentary was assigned M73. The patent for the corner-illuminated US keyboard switches (M41) was filed in May 1974 (there is no known M42 either). The drawing for the German centre-illuminated type (M71-0100) is also from May 1974.

From this, there is a suggestion that Serie M7 started out as a German-made illuminated extension to the US-made M6 switches (or to the US M4 corner-lit design) that grew into a full series. The M73-0110 (standard 10° type) only goes back to 1976, and the 0° T-stem type (M73-0120) to 1979.

M8

M9

FTSC (MY)

M10

What was M10? Cherry Germany has suggested that it exists, but nobody since has ever responded to further questions about it. “MX” means “M10”, seemingly because the redesigned part number schema still did not permit a double-digit series name. M11 series existed as a custom type produced specifically for HP as a clone of Datanetics DC-60 (although this is what Cherry Germany thought M10 might have been), so where did M10 go?

German-made MX

Origin

For MX notes to date, see the Cherry MX page.

The original patent for Cherry MX was filed in August 1982, and it described and depicted a switch with the following characteristics:

Günter’s initial explanation for the origin of the movement differential type was that Olympia requested from Cherry a click-tactile switch priced more competitively than the Marquardt switches that were using (likely the type that we call Marquardt “butterfly” for lack of a better name: Marquardt no longer have any details about it). (He later noted that the Marquardt type was patented, but no patent has been found for this Marquardt type.)

This resulted in MX gaining a click-tactile system, as depicted and described in the patent. However, Olympia declined this product (for unclear reasons), and thus Cherry were left with a feature that interested nobody else. They then decided to apply grease to it to prevent the click sound, and market it as a movement differential type. (Actual marketing shows that it was offered as a tactile type.)

However, there is now growing confusion as to whether this is really what took place, and whether this story took place with MX or M9. It would be good to get clarification that Olympia’s request, and the design of the click-feedback type, really did all take place before 1982, before the patent was filed on the MX design. MX did not become commercially available for another year after the German patent was filed, and when it did, it was redesigned as a single-pole-only type with support for fitting an LED, jumper or diode. With that said, US advertisments in 1983 and 1984 listed double-pole and double-contact configurations, although these were not illustrated, and the single photograph shows a keystem more suited to taking 6 mm keycaps.

General

Variants

For MX variant notes to date, see the Cherry MX variants page.

“Vintage Black”

The keyboard enthusiast community maintains this idea that MX Black (MX1A-11xx, the basic linear type) was much smoother in the 80s and early 90s, than it was after that. Was there any major change around 1992–1995 that would have resulted in a change switch feel, such as a change in plastic material, or a change in contract manufacturer for the injection-moulded parts?

Tipro

Did Cherry in Germany ever supply parts to Tipro in Slovenia? There is a strange Tipro T141A keyboard using Tipro’s own MX-like switches, housed in a Cherry G80-1000 or G81-1000 case date-stamped in German from 1988. The keyboard itself is from circa 1989. Tipro are still around, but this keyboard pre-dates the current Tipro company’s incorporation in 1991.

UK operations

FTSC

Japanese operations

General

See the Hirose page for general notes about Hirose.

M5 and M6

Apparently Hirose started producing M5 from 1983 and M6 from 1985. This has not been verified, however.

Some M6 switches have a pair of trapezoidal holes in the base. There is tenuous evidence to suggest that this is a characteristic of Japanese-made switches, but this remains unproven, and there is growing evidence to suggest that US-made switches were also made in this way. So far, no explanation has been put forward for such a design, and it would risk fluid penetration into the switch during keyboard assembly. It would, however, allow any plastic fragments formed from the ultrasonic welding process to be shaken out of the switch.

M8, MD and MJ

Did any knowledge exist outside of Japan about Hirose’s custom keycap mount, and of their MD and MJ series? MD and MJ are variants of M8 with increased travel (3 mm and 4 mm respectively).

So far, little in the way of details of these switches has been recovered. Hirose did provide datasheets for the basic models (for M83A, M83S, MD, MJ and MX), but forbid them to be shared online. (Seems to be Japanese paranoia. Frustrating.)

The reason for “M83” is also a mystery. Hirose used its own contact system. There was no M83 in Germany but there was an M93. It seems that “3” refers to the contact alloy, not necessarily the contact form, as Hirose M83A claims to use the same alloy as German M93, even if the contact construction is completely different in every way. See Cherry common part number schema for notes on the meanings of the part numbers from M8 onwards.

What was the reason for Hirose designing M8 to be single-pole only? Was this done to provide greater keycap stability (and avoid binding) on the basis that they did not need double pole configuration? This design was also found in a German keyboard under the unstabilised wider keys, although those were clearly German-made switches (from the use of German contact prisms).

MX

Hirose MX switches began with “HCP” branding, before changing to standard Cherry branding. Was there a reason for this change? How much re-tooling was required to achieve it?

Was Hirose responsible for the MX switches that took Alps keycaps? The people who own these never photographed the switch contacts, which would have told us which factory they came from.

Other operations