Datanetics DC-50

Contents

• Overview
• Design
  • Mounting and alignment
• Models
• Specifications
• Variants
  • Alternate action prototype
  • NCR adaptation
  • Switchcraft adaptation
• Keyboards
• Documentation
• See also

Overview

Datanetics DC-50 is a family of tactile mechanical keyboard switches introduced in 1973. DC-50 was designed by Mike Muller of Datanetics, who filed a patent for them in 1972; this was granted as US patent 3777090 in 1973. DC-50 is a continuation of their elastic diaphragm system that began as the array keyboards, yet it was introduced in the same year as the much more conventional DC-60 pure mechanical switch series. Meryl Miller contributed the illuminated version and the original alternate action design. The goals for DC-50 were to have a discrete switch (in place of their existing batch-fabricated array system) with hysteresis and minimal contact bounce.

In his notes, Meryl placed a handwritten note to indicate that Datanetics was selling three million (presumably) DC-50 switches per year at $1 each. This is implied to be at a point in the late 1970s, possibly when they were manufacturing all of Apple’s keyboards. According to a letter from Datanetics to Joe Torzewski in February 1978, these switches were 50¢ each, or just over $2 each in December 2020 prices. The original press release for DC-50 gave a price of 31¢ each in orders of 1 million pieces.

DC-50 Series (as “DC50”) appears in ITT Schadow’s advertisement in Electronic Engineers Master catalogue for 1983–84. It is absent however from the 1985–86, 1988–89 and 1989–90 advertisements, and Apple’s technical documentation for the Apple II from September 1985 states that “Datanetics keyboards and keyswitches are no longer supported and replacement parts can no longer be ordered”.

Design

DC-50 is a discrete switch that uses a hybrid arrangement of both membrane and mechanical switching. Flat metal strips form the switch contacts and terminals, but these are glued to Mylar membrane sheets that provide the release action for the contacts. The sandwich of metal and Mylar forms a sealed contact assembly. The diagram below depicts this assembly:

The connection between the plunger and contact sandwich is made by an arched leaf spring that operates a plastic prong assembly. This design may have inspired the Alps “switchplate” system, which functions in a similar manner. Although most “switchplate”-based switches (from not just Alps but also Omron and other brands) use a different type of actuator spring (specifically a vertical leaf spring attached at the top), Alps SCK switches copy the Datanetics design much more closely, and as with DC-50, provide hysteresis.

Standard tactile switches are formed from two identical shell halves, ultrasonically welded together. Single and double pole operation is supported. All known types are normally open.

Kapton was used for the membrane material in the original prototype contact units, following on from the Kapton membranes in the batch-fabricated array keyboards. The use of Kapton survived to the DC-50 press release in 1972, but the production switches used Mylar membranes instead, which were considerably cheaper.

Mike Muller also designed the manufacturing system for making the contact assemblies, production of which was automated. The material was supplied on rolls and machine-assembled into finished contact assemblies. The berillium copper contact strips were affixed to the two outer layers, then the pre-punched spacer sheet was inserted in between. Following lamination, the contact assemblies were rapidly die cut, and the exposed the solder terminals were formed and dip tinned. This process was turning out around 50 units per minute. The assembly of the switch itself was conducted by hand using a jig to hold the parts. Following this, the switches were fed into a custom-designed ultrasonic welding machine to join the housing together, and then placed into an electronic tester that ran each switch through several cycles to verify correct operation. The tested switches were then placed into styrofoam packaging with a single slot for each switch, rather than tipped into a box or bag as seen with some other brands.

Initially the internal actuator was moulded from black plastic. To make the task of assembling the switches easier, the actuator colour was changed to red.

Mounting and alignment

DC-50 switches are conventionally secured to the PCB with a screw. The patent notes that the stepped shape of the housing allows for a “clamping” bar to secure the switches against the PCB or other mounting panel. The ridge along the side also allows for a “stiffening bar” to be used. Plate-mounted switches did already exist (going at least as far back as Elec-Trol Datakeys from 1968) but not all manufacturers chose to adopt the idea. DC-60 however was indeed plate mount.

Models

The forces given for the alternate action models are the separate latch and release forces, with the latter being higher.

Specifications

The specifications below are from the DC-50 Series Key Switches brochure from November 1974 except where noted.

Variants

Alternate action prototype

In April 2016, Meryl Miller provided package of sample parts and brochures. Included in this was a prototype DC-50 alternate action switch that he designed, that “worked well but did not do well enough in life testing to warrant putting it into production.” Nonetheless, this is the alternate action type illustrated in the DC-50 Series Key Switches brochure from November 1974, albeit with a black latch arm instead of white.

Both the standard type and Meryl’s version can be seen in the photos below.

Surprisingly, the original type did indeed see commercial use, in the ARP Omni. The ARP Omni Model 2480 Service Notes from 1977 list ARP part 1903002 as “SWITCH, ALT ACTION, SP”, manufacturer part DC-51-03. Rich Diemer of BustedGear.com has serviced the Omni for decades, and thinks that the Omni used both the plastic arm and wire arm DC-50 alternate action switches, but all spare parts have sold out and this cannot presently be confirmed. The first photograph below is from the BustedGear.com site listing for UOMNI-SW (not in stock at the time of writing), while the rest were provided on request and show his one remaining switch that does not latch. These switches have a thin piece of metal on the top of the plunger that has not been explained.

Meryl does not know how these came to be sold. One possibility is that these were accidentally sold off as production switches by an employee who did not realise that these were not production switches. The answer will likely never be known.

It is interesting to note that his example was marked “PAT PEND” on both sides, while the example taken from the Omni has “PAT. 3777090” on the side that shares the shell half with momentary types. That is, the switch in photographs 2–5 above may be from a later batch, after US patent 377709—filed in 1972—was granted in 1973. The ARP Omni itself was made from 1975–1977.

All illuminated switches observed to date have the slot in the plunger to take the alternate action arm as found in the prototype switches.

NCR adaptation

Datanetics supplied keyboards to NCR. NCR required these keyboards to use their existing keycaps, so Meryl adapted the switch accordingly. This required replacing the standard cruciform mount with a flat mount. Two such Datanetics-made NCR keyboards were placed for sale on eBay; of these, he notes (in correspondence in November 2019):

I was very much involved in that particular keyboard from the beginning. Nice to see that a couple of them survived all these years. That project has to date back to the 1970s. NCR was unique as a customer. They insisted on Datanetics using their keycaps. So I designed a version of the DC-50 with a keystem having a blade type of top that would engage the slot in their keycaps. In addition, some of the larger keycaps, especially the “L” shaped ones required a dummy support under the keycap in addition the the keyswitch. That allowed the operator to depress the key from anywhere on the top surface.

Having spent the first 15 yrs of my product design career at NCR, I was worked closely with their design team on that particular project. As I recall, the end product was an accounting machine as opposed to a computer terminal.

Meryl previously provided his single sample switch of this type:

A similar keyboard using the same keycaps, and with complicated-looking reed switches is also advertised on eBay. John Culver of The CPU Shack reports that the 73* codes are the production dates, so this keyboard would have been made somewhere around the time that DC-50 was introduced.

Similar keycaps can be found on the NCR Class 18-22 desktop calculator, although that is made in Japan and the keycap step profile is less prominent.

Switchcraft adaptation

Switchcraft 315900 is an adaptation of DC-50 with a non-keyboard button mounting. These are marked “ITT SCHADOW” and were thus made in 1979 or later following ITT’s acquisition of Datanetics. These switches have a higher operating force than the keyboard switches and are double pole.

Keyboards

Documentation

Meryl Miller provided two separate DC-50 brochures. These have been donated to the National Museum of American History Library who provided higher quality scans.

• DC-50 news release, Bumpas & Jones, September 11 1972 (scanned by Meryl Miller)
• DC-50 Series Key Switches
• DC-50/DC-60 Key Switches, ITT Datanetics, 1981
• ITT Schadow advertisement, Electronic Engineers Master 1983–84 Volume 1 (includes DC50 and DC60)

See also

• Datanetics DC-50 series on the Deskthority wiki for prior notes and illustrations