HP 46010A Teardown, Restoration, and Conversion

[ScottPaladin]

I've been looking at these keyboards for a little while now and finally bit the bullet.



Got this one from an IT supplier, and I'm gonna admit that I paid too much for it. I've got a familial connection to HP and wanted their logo in my collection in some form and this was about the cheapest way to do it. Since I wasn't able to find a ton of info on these out there, I figure I'll share what I find out as I go along.

The keys are pretty sensitive to off center presses, but feel good when hit directly on. That makes the weird ass profile of the keycaps make a bit more sense and why they went overboard with the stepped keys. The layout is weird, befitting a terminal board, but I do like the color of the legends. I'll get my macro lens out later and get a close up shot of the printing and somebody can identify the method.


Just one big ole foot on the back here. There's a RJ socket on here and the label says it's a 12v system.



If you're going to pull the keycaps on one of these, rock the puller up or down before you apply pressure upward. You run the risk or pulling the switch slider from the barrel, and it's a pain to separate a keycap from the slider if you do that.


That's a hell of a lot of solder points back there. Each switch get's four solder points and there's a diode for each switch. So if I ever get this up and running, it should have NKRO.


Sticker outside says Taiwan, this one says Japan. I'm gonna guess the plate/switches/board were made by Fujistu and then sent out for final assembly. I feel like this is greater than the usual number of ICs for a keyboard. Not a lot of luck getting specs on the chips, just by googling.


These are Fujistu Leaf Spring switches 3rd Gen of the clicky variety. They feel pretty good if you hit them straight on, but bind off center pretty easily. I hope a little lube will help with that, but it will take some experimentation. More info and pictures to come as I get further with this project.

[tp4tissue]

Looks good..

weeb approved..

[chyros]

Very nice! Usually FLS3 are mounted sideways, not upside-down; maybe that's a thing for the clicky version.

FLS are fairly dirtproof, but are not immune to use. Do the caps show sign of having done heavy duty?

[ScottPaladin]

Very nice! Usually FLS3 are mounted sideways, not upside-down; maybe that's a thing for the clicky version.

FLS are fairly dirtproof, but are not immune to use. Do the caps show sign of having done heavy duty?

The caps look pretty good. The board was "restored" so it's possible they were replaced, but I don't think it's likely.

Pulled out my macro lens and got some shots of the keycaps.


I'm not that good at telling one type of printing from another, so feel free to voice any opinions y'all might have.

I don't see any yellowing or shine on these, so my instincts say PBT.




The stabilizer mounting method is a bit novel. This slider clip things are held in place by the keycap mounting to the slider. I bet they're a bugger to reassemble.




With it down to bare board, I've been trying to trace the matrix, but I'm a bit thrown. My multimeter's continuity mode won't pass through the diodes in either direction. This is probably obvious to anybody who's more electrically minded than I. Gonna do a bit of research and think.

[ghostjuggernaut]

Very nice! Usually FLS3 are mounted sideways, not upside-down; maybe that's a thing for the clicky version.

FLS are fairly dirtproof, but are not immune to use. Do the caps show sign of having done heavy duty?

The caps look pretty good. The board was "restored" so it's possible they were replaced, but I don't think it's likely.

Pulled out my macro lens and got some shots of the keycaps.

Show Image

I'm not that good at telling one type of printing from another, so feel free to voice any opinions y'all might have.

I don't see any yellowing or shine on these, so my instincts say PBT.

Show Image

Show Image

Show Image

The stabilizer mounting method is a bit novel. This slider clip things are held in place by the keycap mounting to the slider. I bet they're a bugger to reassemble.

Show Image

Show Image

With it down to bare board, I've been trying to trace the matrix, but I'm a bit thrown. My multimeter's continuity mode won't pass through the diodes in either direction. This is probably obvious to anybody who's more electrically minded than I. Gonna do a bit of research and think.

Diodes can be tough to check when installed in a circuit.  Its possible that you are passing through the diode in the fwd bias direction, but going back through a different path around the diode in the reverse bias direction.  That or the diode is shot, but id be hesitant to think that.



Edit:  Doh, you said it wont pass through.... ignore my post.

[suicidal_orange]

I'm really not understanding these switches - all four pins appear to be one sheet of metal, but that can't work.  Other then that there are some friendly looking columns visible so hopefully the designer wasn't on anything too mind bending making your job easier.

Is it possible the board is old and diode technology wasn't as great back in the day so there's a small resistance instead of true continuity?  Or perhaps there's a waterproof coating as it's an industrial board?  If anything does show continuity it's not this...

As to the caps I'm not seeing any increase in height or loss of texture as expected with pad printing and they're not double shot as the back is all one colour - looks like dye sub PBT.

Interesting board, good luck getting it working!

[ScottPaladin]

I'm really not understanding these switches - all four pins appear to be one sheet of metal, but that can't work.  Other then that there are some friendly looking columns visible so hopefully the designer wasn't on anything too mind bending making your job easier.

That sheet metal is just the leaf spring. The actual switch contacts are under that yellow insulating layer. There's a little convex copper plated disk sitting with it's edge against some pins and it's middle just above another. When the leaf spring compacts it, the disk makes like a snapple lid, which provides the click and tactility (both are pretty light) and bridges the edge pins to the middle.

Here's a disassembled switch from DT wiki which I think makes things clear.

Is it possible the board is old and diode technology wasn't as great back in the day so there's a small resistance instead of true continuity?  Or perhaps there's a waterproof coating as it's an industrial board?  If anything does show continuity it's not this...

After a little research, my guess is that whatever number of millivolts my multimeter's continuity setting uses isn't enough to overcome the knee voltage required to activate the diode. (I hope I'm using those terms correctly.)

As to the caps I'm not seeing any increase in height or loss of texture as expected with pad printing and they're not double shot as the back is all one colour - looks like dye sub PBT.

Thanks, PBT was my guess, which implied dye-sub. However, I know I'm no expert on this subject so it's nice to hear other's concur.

Interesting board, good luck getting it working!

[ScottPaladin]

A bit of experimenting at lunch has shown that the sticking/binding is greatly improved by a bit of silicone lube applied to the barrel around the slider, so that's the plan going forward. I need to break out some stiffer cleaning agents. There's an irregular layer of sticky brownish yellow material on the PCB. I'd suspect it was the ancient a soda from about twenty years ago, but I couldn't find trace of it on the switch plate or case. So either somebody cleaned those really well after the spill, or the source is still unidentifiable. Isopropanol makes a bit of a dent, but hasn't removed it completely. If anybody has a suggestion for something stronger that won't damage the PCB, I'm all ears.

[ScottPaladin]

How do I spend my Friday evenings? Manually following keyboard traces.


This is done in photoshop, so each column/row is on its own layer. I hope I mapped everything right. It's a single sided, single layer board. Luckily for whomever designed it, each switch has three bridged legs, which can act as a jumper on the other side. (Don't know the correct term).


Here's the overlay of the layout. There are some extra footprints for two more switches. I wonder if they made a variant that populated those. There are holes in the plate too, so if I had extra switches and some compatible keycaps, I could add key there.


Columns only.


And the rows.

Looks like this is 8x14 which is smaller than I feared. 22 pins means I think I'm out of Micro territory and need to get myself a teensy. I've got one or two in the post from china, but waiting sucks.

[ScottPaladin]

I'm typing this post on the board right now.

I found a compatible micro controller in my stock and got a little excited. I did what research I could on the HIL protocol that this board is suppose to use in hopes I could convert it, but it seemed like too much effort just to keep this thing unmodified. These aren't rare or valuable boards and this one isn't pristine, so making it USB compatible with an arduino is a net gain. Using my key map, I made up a QMK firmware for it and flashed a Arduino Micro. I also spent a couple of hours lubing the sliders and barrels of all the switches with dielectric silicone. The sticking on off center presses is pretty well taken care of, although if you try you can still make them stick.  I was right about the stabilizers, they are a pain to put back on.

Typing experience is pretty good. These are pretty smooth since they've been lubed, and the weight is quite light. There's a slight tactility on both up and down strokes and a pretty quiet click. No pinging or ringing from the mounting plate, except for the space bar which has a sling metallic sound. This wouldn't get you lynched in an open office. Layout is a bit weird of course; I'll have too keep tweaking the exact features if I want to use this thing for more than a few paragraphs of typing.

Next order of business it is to make the wiring permanent and stuff it all inside the case. There's plenty of room, but not quite enough vertical space to handle the headers and jumper wires, so I'll have to trim those down and solder the wires directly to them.

[ScottPaladin]

I'm gonna call this thing done. When I was trying to prep for the permanent wiring, I had four solder pads lift and start pulling their traces. I decided that I needed to stop ****ing around with this thing it would just get worse. I compacted some of the wiring and can everything electrical isolated, then tucked it all away. Like an idiot, I'd misplaced the original case screws. Using my soldering iron, I sunk some M3 inserts into the case (only do this in a well ventilated area if you ever try it, plastic fumes are no joke). So now it's held together by some proper machine screws.



The way things worked out, you can see the Arduino through the gaping hole that HP/Fujitsu left at the back of the case.

If I was going to do another one of these, it would be a lot simpler and I think I'd end up with a better product. If I spot an new old stock or in box one, I might just do that. But for now it's good enough to use and it makes me happy to have it.

I did a quick video of me typing on it, just to give people an idea of the sound. I've a bit of a caveman when I type, so it's loud. In the last third of the video, I try to compare it to my Lexmark era IBM Model M, both with bottoming out and pressing only to activation, which probably gives you a better idea of how it could sound if a proper typist was using it.

[Daniel Beardsmore]

You know, that's one of the best-sounding switch types I've heard. Very deep, much more so than Alps!

Curiously, I see that you could add extra modifier switches where the blockers are placed, which would improve its usability.

[ScottPaladin]

You know, that's one of the best-sounding switch types I've heard. Very deep, much more so than Alps!

I was surprised when I got them but I think these might be the best sounding switches I've heard in person.

Curiously, I see that you could add extra modifier switches where the blockers are placed, which would improve its usability.

Yep if you had a couple of switches and some 1.5u caps, the traces are all there. I personally don't find myself wishing I had some additional keys in those spots, but it wouldn't be too hard to do.