Datanetics DC-50 hysteresis

[Daniel Beardsmore]

Back in April, I inherited Meryl Miller's collection of Datanetics paraphernalia from when he worked there in the 60s and 70s. This included DC-50 and DC-60 switches and components as well as the hard copies of the material published on the wiki. (Edit: Meryl was the engineering manager at Datanetics — I found the vcfed.org topic where he mentioned this.)



I'm not sure where the situation stands regarding the photos — I'm having real trouble getting cardboard dark enough. The tall sliders have to be shot against a comparatively dark background, and two separate packs of card I bought both turned out to be far too light. (It's a particular problem with really tall switches, such as DC-50 and ITT magnetic valve.)

One of the tubs he sent me contained a cut-apart DC-50 switch. Most switches have a top and bottom shell that are totally different, but DC-50 has two identical sides that are permanently attached. This curious design means that you can assemble and operate the switch as a cutaway exhibit.

I've always known that it has hysteresis, but I've never understood how it works. There seems to be some confusion on whether "lost motion" literally means hysteresis or not, and in the case of DC-50, it really does mean lost motion. For me, it didn't make sense until I saw it in operation for real.

The overall principle of operation is basically the same as Alps SKCC/SKCL/SKCM switches. In fact, the "elastic diaphragm" membrane switch technology was licensed to Futaba in 1972 shortly before DC-50 was introduced. A sandwich assembly is closed by a panel bearing a prong, and that is pressed on by a leaf spring.

However, the tiny beryllium-copper folded leaf spring in DC-50 is captive but is free to move vertically. This small amount of lost motion that occurs is what gives it hysteresis. The following photos show the operating positions:

DC-50 operation 1.jpg (150.73 KiB) Viewed 781 times

DC-50 operation 2.jpg (146.39 KiB) Viewed 781 times

Basically, the slider pushes this leaf spring downwards, and it presses against the plastic prong that in turn closes the membrane contacts. When you release the slider, there's a delay before the lower end of the leaf spring recess reaches the slider and pushes it back upwards, so the release lags behind the slider, giving you hysteresis. It's hard to explain, but makes perfect sense when you see it working. There's even a tiny click when the leaf spring clears the prong and it springs back up.

I may have to measure the parts and draw a diagram, as it will be abundantly clear then.

The following photos show the switch more clearly — it's really hard to get a good angle considering that I have to keep my thumb on the slider at all times (otherwise it will catapult out):

DC-50 operation 3.jpg (279.56 KiB) Viewed 781 times

DC-50 operation 4.jpg (220.67 KiB) Viewed 781 times

This is the spring in the slider, in its special recess that is longer than the spring. This isn't even the grey card I bought, but the much darker packing card it came in, and even that is still far too light! (And it looks like there's hair all over it!)

Spring.jpg (77.11 KiB) Viewed 781 times

Getting the right shade of card isn't easy: it needs to be dark enough for the slider to stand out, but not so dark that the shell disappears into it, and actually exist for sale! This is switch pressed together:

DC-50.jpg (137.65 KiB) Viewed 781 times

The actuator leaf is noticeably absent (look at the downwards slot just under the keycap platform), so if I do reshoot this with a less kludgey background I'll reassemble it 100% complete. Getting that leaf spring in without the official Datanetics assembly jig is rather fiddly!