deskthority

I have (part of) an old catalogue from RAFI which covers the RS76M (mechanical) and RS76C (Hall effect) switches as well as the RS74M (short throw mechanical) in my Necam board.

It still doesn't cover "magnetic pulse", nor does RS76C appear to resemble the magnetic pulse switch, though they have the same keycap mount. I've just found a second person who owns a "magnetic pulse" keyboard (kaineko2 — sadly he appears not to speak any English) though he has not commented on the switches:

http://park8.wakwak.com/~kaineko2/keybo ... 10107.html

Is there any source of data on RAFI magnetic pulse switches? I can only extract information from RAFI one sentence at a time :) A model number would be nice. RAFI class the mechanical and Hall effect switches in the same family ("Full-Travel Keyswitches": they use "keyswitch" to refer to pushbuttons as well) — but no mention of magnetic pulse.

The earliest catalog I found, from 1999, only mentions the RS 74, RS 76M and RS 76C.

You have a year on that catalogue? I only have a partial catalogue (just the keyswitches section) with no year on it — I think I did ask about the year, but I overran my sentence quota from RAFI :) Mine calls the 2.5 mm version the RS 74M, not just RS 74. I intend to document the switches later.

Source:
http://web.archive.org/web/200104200148 ... spekte.htm
There are PDFs in English too:
http://web.archive.org/web/200105060712 ... logues.htm
This is from the 2000 website snapshot, I didn't find much interesting in 1999 (first snapshots), but only browsed casually.

The page on the RS 74M is up:

http://deskthority.net/wiki/RAFI_RS_74M

That's the only switch I have. I've mentioned the magnetic pulse switch to RAFI again, maybe they'll remember.

By the way, what is "magnetic pulse", anyway? It doesn't appear to be an industry term — and since Hall effect *is* magnetic, how does magnetic pulse differ from Hall effect, which we know RAFI does still sell to this day? (Although the switches do differ, there are no good pictures of RS 76C to see how it compares with their "magnetic pulse" switches.)

The Hall effect is magnetic, but it isn't pulsed. It's continuous, and actually quite subtle to measure as it is so small.

Magnetic pulse switches sound only a nominally related concept. Their name suggests active strobing. Perhaps they are cheaper. You'd expect more undesirable side effects using pulses than a continuous system like Hall effect.

This may be the patent in question:

http://www.google.com/patents/US4099176

It looks like it's a little transformer, with the "pulses" being the AC required to operate a transformer, and the slider activates and deactivates it. It's hard enough trying to understand the concept with my flaky understanding of electronics without the patent writers' flaky understanding of English and the illustrators' flaky understanding of technical illustration. (Maybe Google Translate needs a "Patent → (your language)" feature. There's nothing patent about patents …)

FWIW, the guy at RAFI is unlikely to shed any light on the magnetic pulse switches. HaaTa is now our only hope :)

Flying back to Canada today...
There are a few patents listed on my keyboard.
*HaaTa goes to read patents, needs a bit of a refresher on this switch*

I hadn't noticed before but I think that patent corresponds to the HP hall effect switches very closely in terms of design (which makes a lot of sense).

Anyways, I think I have a good explanation for how the switch works now. But there's no way I'm gonna type it on some crappy cellphone touchscreen

Nooooooo, just had 2 pages of explanations eaten by Chrome...grrr pictures and everything.

Gotta board my flight now, so will have to wait till tomorrow.

How I know that feeling. Once I've more than two pictures or a few paragraphs in a browser's text field, I get itchy and copy the lot to my text editor. Real process separation.

RAFI's verdict on HaaTa's and kaineko2's photos:

"This is sure no RAFI pushbutton and it is sure no RAFI keycap"

The keycap mount is absolutely identical, so it is either a long-forgotten product, or a clone. That said, it is indeed not their keycap style.

Ooo, neato. This further leads me to believe that it is a switch made by ITW.

So, bored on my flight so doing the write-up again

I recently did electromagnetics research in my final year of undergrad (earlier this year), so I'm actually sorta qualified to explain this
(This should go on the wiki after a bit of editing/comments, and getting/making properly licensed pics)
(I really should do this for the HP inductive switch as well as I have full schematics for my keypad )


While I regret a little bit naming this switch mechanism (will be explained) as it is not integral to how the switch works. I didn't fully understand how the switch worked. I knew that it was like the Honeywell Hall Effect sensors, but not quite the same, and that for some reason the signals were pulsed. But using the name magnetic pulse is not completely incorrect, as calling it a magnetically pulsed keyboard is accurate. (And magnetic pulse just sounds cool )


Before I get into explaining the magnetic pulses, I need to explain how the ferrite works in relation to the magnet and the driving current used for the sense line, which are electrically disconnected.


This switch, according to wikipedia, is a true hall effect mechanism. It is Ferrite toroid Hall effect current transducer. But it differs from Honeywell Hall Effect switches, as controller does not measure the strength of the magnetic field coming from the magnet. If I were to give this a name, I'd call it a magnetic valve switch (I'll be explaining why after I get through the technical details).



But for those of us with high school physics (and especially those of us who didn't pay attention; don't worry I didn't either ), hall effect doesn't mean anything really.
In laymans terms, a hall effect sensor is a device that can measure the strength of a magnetic field, outputing a higher voltage with a higher magnetic field.
Honeywell Hall Effect switches use a semiconductor transducer (i.e. microchip) to output the voltage. Obviously for the RAFI switch, there is no semiconductor, just a ferrite with two separate metal "loops" (electromagnetics term).


So, the RAFI switch has three key magnetic components: wire loops, magnet, and a ferrite toroid.
The wire loops are responsible for carrying electrons (current), the magnet for providing a strong magnetic field, and the ferrite acts as a magnetic field transfer medium (magnetic field flows more easily through than through air).

It's necessary at this point to understand the relationship between the wire loops magnetic fields.
A power generator uses a "changing" magnetic field to "induce" a current on a loop of wire. This is done by having a loop of wire rotate in the prescence of a magnetic field. The magnetic field causes electrons to flow in specific direction (remember the "right-hand rule"?), this flow of electrons is called a current and for a current.



The inverse is also true. Any wire that has a current moving through it will generate a magnetic field, the strength determined by the current (more right-hand rule ).



The RAFI switch utilizes both of these effects in order to build a magnetic "circuit".
When a current is passed through the drive loop (line) of the switch, a magnetic field is generated around the loop and since magnetic fields flow through the mediums of least work, flows around the ferrite (direction determined by the right-hand rule).
This magnetic field causes a flow of electrons to occur on the sense loop, creating a current from the displacement of electrons (giving rise to a measurable voltage).


This is how the magnetic circuit works using the ferrite toroid as a magnetic field transfer mechanism. Yet there are still two more things to cover. Why is the signal pulsed, and what is the purpose of the magnet.


The magnet is the reason I'd call the mechanism a magnetic "valve". When the magnet is far away from the ferrite (or the field is not directly ideal), the circuit works as explained above.
Yet, when the magnet is moved close to the ferrite the magnetic field of the magnet "overpowers" the magnetic field generated by the single loop. This is important as the magnetic field from the magnet is different in this case. It is much larger, encompassing the width of the ferrite and thus does not flow completely around the toroid circle.
Since the magnetic field is not flowing in a circle around the toroid it does not encircle the wire loop, and thus does not displace electrons (i.e. no current).
Thus when the magnet is brought close enough (and in the correct direction) it completely nullifies the current from the drive line (loop), acting like a magnetic "valve" for the electrical signal.

(I really should have a pic here, but I'm pretty sure I'd have to draw one...Daniel bug me later, once I have a real mouse, I'll make an SVG)

Now, from the discussion so far, we having everything needed to make a switch for a keyboard. What's the point of the "magnetic pulse" (or electrical pulse)?
Actually this part makes lots of sense
Just think of what happens when you scan a keyboard matrix
Each key is scanned separately, and this is exactly what's happening here, and thus the keyboard is "magnetically pulsed" in order to read each of the switch states.


And there you have it, an explanation of the ITW (RAFI) magnetic pulse keyboard patent. ITW has a bunch of different variants, I can think of at least 3 right now, I might have 1 or 2 more examples but I'd need to double check my pics/records.

HaaTa wrote:(I really should have a pic here, but I'm pretty sure I'd have to draw one...Daniel bug me later, once I have a real mouse, I'll make an SVG)

Humbug, I did all my force curve diagrams with my laptop trackpad while trying to resurrect Windows 8, which took days.

Just kidding (though it's true) — I have quit bugging you, or anyone in particular. I'm supposed to have quit anything non-Alps, but look how well that turned out …

I did ask RAFI for a good photo of the RS 76C, but the guy didn't co-operate. I wanted to get a detailed view of RS 76C just to see how different they really are.

If these are not RAFI switches, why do they use RAFI-mount keycaps? I guess at this stage we know too little about RAFI, and whether any other companies made compatible keycaps, and whether some other company decided to make switches compatible with RAFI-compatible keycaps …

RAFI do make keyboards with more conventional keycaps (link 1, link 2).

I suppose you could also describe this switch as a transformer with a magnet-intolerance.

Daniel, it may also be possible to get some information out of Cortron concerning the switch that's in the Xerox keyboard.

Are you saying that you will, or are you trying to tell me to do this? If the latter, it would be helpful if you could provide some details that I can pass onto Cortron, instead of "you may have had something to do with some Xerox keyboard once" … (it's likely they won't recognise the photos any more)

Another thing I just realised — these "magnetic pulse" switches are plate mount, while all the RAFI switches in their catalogue are PCB mount. RAFI keyboards with a plate use a "coupling" piece to extend the slider height up to reach the keycaps.

Cortron has used this type of switch in a number of keyboards I have. Not the exact same kind, but they use the same ITW mechanism as I described above.
I can follow up with them, but it'll have to wait until I find a place to live (end of the month?).

https://plus.google.com/photos/11384566 ... banner=pwa
https://plus.google.com/photos/11384566 ... banner=pwa

It was a while back, but I have seen this switch (the one in the Xerox) used in other keyboards.

The link in the other topic has a lot more details that I can pass on, so I'll e-mail them tonight; I've just given Stackpole a poke, too, as it would be great to get that mystery solved.

I wouldn't want to bet on it, but my suspicion is that RAFI did make the switches and simply forgot about it. The bloke I was in contact with wasn't overflowing with joy and knowledge.

E-mail sent off to Cortron. I notice that that Xerox PCB is clearly labelled "ITW Cortron" — seems ITW and Cortron were at some stage related. That gives credence to the suggestion that ITW made the switches — it would be odd to licence the design to RAFI and then buy back RAFI's switches … So why change the design radically and do their utmost to resemble RAFI's? Maybe Germany were able to make them cheaper?

Not unless RAFI styled their switches—but not any of the mechanics/electronics—on ITW's switch?

Possibly the RS 76C is in fact a knock-off of the ITW shell/slider/mount design using more Honeywell-style electronics, and then the RS 76M was a cheaper alternative.

See, my Necam keyboard has 1984 on the PCB, and your Xerox keyboard came out around 1982. And RAFI tell me that the full-travel series was introduced in 1975. The ITW patent was filed 1974 — would they have really changed the design so dramatically in a year? After all, you've got a keyboard that has switches pretty much as given in the patent.

It's like Alps clones all over again!

Incidentally, Stackpole's switch division likely became Ark-Les, who are part of ITW! I've put both switch types to ITW, in case someone still remembers.

Very cool, did not even think to consider that

I got a confused and very cautious reply from Cortron — apparently the switches are "Cortron" mount (an industry standard term at the time, apparently), and they're definitely nothing to do with RAFI at all. Not sure whether we'll learn anything more officially; we'll see.

Just found this page when searching for "buckling rubber sleeve":

http://deskthority.net/keyboards-f2/rev ... -t449.html

1998 date on the PCB! I've dropped Devlin an e-mail too, to see what they remember.

I just now read through this thread. That is a very interesting switch mechanism. Having finished German Gymnasium, in a class with focus on physics, this year, I had no problems following your explanation. I thought that the Hall effect is basic knowledge… (all our physics courses were taught about it, not only our science class)

It's necro time.

[wiki]ITW magnetic valve[/wiki]

Confirmed as ITW, with later production by Devlin until 2005.

I'll let HaaTa finish off that page — I've just thrown something up with all the details I know.