Alps Glidepoint Keypad - Backlighting & Matias Switches

User avatar
flowerlandfilms

05 Dec 2018, 13:32

In the 80's and early 90's, it was all about COMDEX.

A massive yearly trade-show where people came from all over the world to showcase hardware, software and everything computers. Careers were made and broken on the show-floor, and handshake deals worth billions were struck over shrimp cocktails in Vegas hotel rooms.

One minute you're a poor audio engineer from Singapore showcasing PC music on your add in "Sound Card", then all of a sudden Michael Jackson drops by your booth, and the Sound Blaster empire is born.

One moment Steve Jobs is blowing you off as just another software hustler who doesn't belong on his precious Mac platform, and the next you're CEO of a little company called Microsoft, and you are keeping Apple afloat by purchasing 18 million shares, for the LOLs. Anyone who was anyone was at COMDEX, and for the players in the Keyboard space, it was no different.

"Everybody is terrified, nobody knows which technology is going to win..."
"Your friends are trying to rip you off, your enemies are buying you drinks, it's capitalism at its finest."
-Donna & Gordon Clarke, on COMDEX '83
Alps Glidepoint 1.jpg
Alps Glidepoint 1.jpg (106.64 KiB) Viewed 5239 times
George Gerpheide felt that he had something special to demo at the upcoming COMDEX, he had grown up in a world where keyboard was king. Programming and typing on keys felt very natural, but having to move your hand away from the keyboard to adjust a "Mouse"? It seemed awkward and inefficient, and after having been repulsed by the usage of the device now included in Mac systems he had invented a technology for controlling the cursor on a PC or Laptop screen without using an external peripheral.

Due to his strong dislike of the mouse, he originally named it the "Cat". To bring it to life he looked at Optical Sensors, Acoustic Wave technology, Resistive Membranes, but in everything he tried he found it was difficult to get a good electrical signal, and at the same time a smooth movement across the surface of the pad with a finger.

Eventually, he stumbled upon the realization that you could distort an electrical field, created by a grid, just by hovering a finger over it, without even making physical contact. That way you could have a smooth surface above the grid that was more conducive to direct touch. He mocked up his prototype "Cat 19" and used some plastic from his sons model kit as the touch surface. The cursor glided smoothly across the screen. Eureka. He had just invented something which could be huge in the burgeoning world of portables and laptops.

"We would often drive to the COMDEX trade show in Las Vegas and stay in a seedy hotel. There wasn't money for a booth at the show, so we carried our GlidePoint prototypes around the convention center making demonstrations to whoever was willing to watch."
- George Gerpheide
Alps Glidepoint 1.1.jpg
Alps Glidepoint 1.1.jpg (70.66 KiB) Viewed 5239 times
George teamed up with some friends and started working the market looking for a buyer.

DEC said no. Compaq said no. Sony said no. Logitech said no. "We've got trackballs, beat it hippy!" (paraphrasing)

Zenith thought it was incredible, and saw the use in it, but told George:

"We don't develop peripherals, we buy them from outside vendors. Have you tried ALPS, or Keytronic?"

Out of patriotism he tried the American company first. They approached Keytronic, but they wanted proof the concept would fly before they invested money. They said:

"Why don't we develop a keyboard with your trackpad in it, and we'll show it off at Comdex 89. If people like it, we'll buy the patents and the concept off of you lock stock and barrel for $75,000." ($153,000 in modern US dollars).

"Gee Whiz!" thought George, "I could pay off the people helping me develop and sell this thing, no more debt for me!"

It was looking promising, but just like the key travel on the Foam & Foil switch for which they are known, the bottom fell out of the deal, and no keyboard prototype was ever developed.
Alps Glidepoint 1.2.jpg
Alps Glidepoint 1.2.jpg (266.47 KiB) Viewed 5239 times
Finally when all hope seemed lost, they took a meeting with Apple. It was now the early 90's and a rudderless Apple was hoping to make a bigger splash with their portable line. They wanted the Glidepoint to eliminate the trackball and make the overall computer smaller. They flew George out to Silicon Valley from Utah, and signed a licensing deal.

With this money George founded 'Cirque' to develop future touchpad products. However Apple had exclusivity on the technology for 3 years, and while Apple had the resources and connections to make integrated circuits to run the tech in a small package, getting a chip maker to deal directly with Cirque was another hurdle. They had designed the chip, they just needed a foundry to make it for them. And on their small scale, it would take the entire length of the Apple exclusivity period to develop it in time to have any products to sell.

Meanwhile at Apple, infighting about product launches and a greater than expected difficulty in IC design led to delays, and the first trackpad product was not released until 1994 with the PowerBook 500. Around this same time, Glidepoint was ready with its first products.
Alps Glidepoint 1.5.jpg
Alps Glidepoint 1.5.jpg (213.28 KiB) Viewed 5239 times
ALPS however, had been watching from the wings since the very first COMDEX where Glidepoint was shown. As a traditionally cautious Japanese company, they didn't come out and state their interest right away. They knew trackballs were going out of fashion, and they wanted in on a touchpad play, but they were hedging their bets for what would catch on, also investing in Thinkpad style rubber pointing devices. With Apples exclusivity period over, ALPS made a licensing deal to produce Glidepoint trackpads for a variety of products.
Alps Glidepoint 2.jpg
Alps Glidepoint 2.jpg (2.05 MiB) Viewed 5239 times
Cirque was making money hand over fist, & although the money from ALPS licensing was helpful, the margins were much higher for Cirque to sell their own retail products. In 1994 'PC Magazine' listed Cirques Glidepoint Trackpad as a "Best product of 1994". And sales went through the roof. Cirque was all of a sudden one of the fastest growing companies in America, but the ALPS deal was giving them headaches.

They had assumed that a deal with an OEM company like ALPS would just be free money, ALPS pays a license fee per product they make and sell with Glidepoint tech, and Cirque doesn't have to do anything. But ALPS had all of a sudden decided to try dipping their toes in the retail sphere. For the first time they were making and branding their own products in stores, with both ALPS and Glidepoint branding, on the same shelves as Cirques products. Cirque found themselves in the position of being a source of major competition against themselves, which was eating into their profits.

This led to some animosity amongst the partners.

STORY TO BE CONTINUED...

But there is a far worse consequence of this corporate skullduggery, than hurt feelings and lost profits. This is an example of one of the ALPS retail products which got Cirque so worked up. In an effort to compete in the new and lucrative trackpad market, ALPS was forced to keep costs low, resulting in this missed opportunity... the ALPS Glidepoint Trackpad.
Alps Glidepoint 3.jpg
Alps Glidepoint 3.jpg (4.07 MiB) Viewed 5239 times
This thing is an abomination in a snazzy package. A full numpad and trackpad in a discreet portable unit with a serial connection, the cover slides out and folds under to become a stand. Doesn't sound so bad right? But...

Crappy feeling SKFS Switches and Chintzy non-ALPS mount, pad printed keycaps make this numpad a disgrace. It sickens me. And yet, the trackpad is really cool tech. This product could have really been something with a little more cost and effort. George revolutionised the computing world with over 90 percent of laptops using some form of trackpad to this very day. His invention deserves better than this tomfoolery. It's worth saving. Let's see what we can do to fix it...

Then after I'm done, we'll finish the story of what happened to ALPS, and Cirque, and COMDEX and George.
Will little Timmy get out of the well?! Stay tuned to find out!

Step 1: You know me, I'm gonna backlight this sucker - CHECK!
Step 2: Needs better Switches - CHECK!
Step 3: Needs better Keycaps - CHECK!
Step 4: Needs conversion to USB

[steps may be enacted in no particular order]
Last edited by flowerlandfilms on 06 Dec 2018, 05:51, edited 3 times in total.

User avatar
flowerlandfilms

05 Dec 2018, 13:59

Figuring out where the Teensy and LEDs will go with the Mainboard and switches removed, the whole thing has to go I think to make room for full travel switches and nicer caps.
WP_20181204_16_10_52_Pro (2).jpg
WP_20181204_16_10_52_Pro (2).jpg (2.94 MiB) Viewed 5225 times

User avatar
flowerlandfilms

05 Dec 2018, 15:06

Decided on a mounting point for the Teensy and wired up the lights.
It's going half under the trackpad, plenty of space to lie down.
WP_20181205_19_09_07_Pro (2).jpg
WP_20181205_19_09_07_Pro (2).jpg (2.16 MiB) Viewed 5203 times
Last edited by flowerlandfilms on 06 Dec 2018, 00:17, edited 1 time in total.

User avatar
flowerlandfilms

05 Dec 2018, 15:11

Testing the lights.
WP_20181205_19_04_01_Pro.jpg
WP_20181205_19_04_01_Pro.jpg (3.27 MiB) Viewed 5200 times

User avatar
flowerlandfilms

05 Dec 2018, 23:47

For the plate, I ordered this steel one from KPRepublic and Sawed off a Numpad sized chunk.
WP_20181206_09_37_46_Pro.jpg
WP_20181206_09_37_46_Pro.jpg (3.83 MiB) Viewed 5161 times

User avatar
flowerlandfilms

06 Dec 2018, 00:16

At first I was trying to figure out how to add space for the longer keys. Then I realised if I just ignore those I can have more buttons. This is mostly going to be used for Maya AutoDesk Hot Keys.
WP_20181120_16_38_11_Pro.jpg
WP_20181120_16_38_11_Pro.jpg (4.32 MiB) Viewed 5155 times

User avatar
flowerlandfilms

06 Dec 2018, 00:56

20 Matias Quiet Clicks, Linearized and un-dampened.
I took apart a spare Display Port cable to get all the wires.
WP_20181205_12_50_13_Pro.jpg
WP_20181205_12_50_13_Pro.jpg (4.42 MiB) Viewed 5146 times

User avatar
flowerlandfilms

06 Dec 2018, 01:42

Adding wires to the mouse buttons daughter board.
WP_20181206_11_34_41_Pro.jpg
WP_20181206_11_34_41_Pro.jpg (3.54 MiB) Viewed 5136 times

User avatar
flowerlandfilms

06 Dec 2018, 03:02

Switch array soldered and ready to go. Hopefully I didn't damage them while removing the tactile leaves and rubber dampeners.
WP_20181206_11_13_00_Pro.jpg
WP_20181206_11_13_00_Pro.jpg (3.99 MiB) Viewed 5124 times

User avatar
flowerlandfilms

06 Dec 2018, 03:22

Putting down some plastic insulation to prevent unwanted touching. These copper contacts need to leave some room for Jesus.
WP_20181206_10_51_01_Pro.jpg
WP_20181206_10_51_01_Pro.jpg (3.92 MiB) Viewed 5115 times

User avatar
flowerlandfilms

06 Dec 2018, 05:49

Wires laid out ready for soldering.
Temporary layout, but I will be using these nice ALPS doubleshots from a Toshiba laptop. They have the same colour scheme as the originals, but are much better quality. They should have used these in the first place.
WP_20181206_15_43_41_Pro (2).jpg
WP_20181206_15_43_41_Pro (2).jpg (1.97 MiB) Viewed 5091 times

User avatar
flowerlandfilms

06 Dec 2018, 06:24

All the switches are now connected, but I've got a bit of an issue with the trackpad.
I've found V+ and Ground, but poking around with the multimeter I can only identify either Clock or Data, but one of them is missing. I've identified the one by watching the voltage drop to half as the trackpad is being used, but their should be a second. The only other pins registering voltage are quite low, and they connect up to the left and right mouse click buttons. I'll have to just make some educated guesses here.
WP_20181206_16_17_26_Pro.jpg
WP_20181206_16_17_26_Pro.jpg (3.81 MiB) Viewed 5083 times

User avatar
flowerlandfilms

06 Dec 2018, 09:25

Up to the programming stage.
The A and S keys are misbehaving, not sure why.

WP_20181206_18_10_03_Pro (2).jpg
WP_20181206_18_10_03_Pro (2).jpg (1.88 MiB) Viewed 5057 times

Anakey

06 Dec 2018, 09:40

Your A key looks like the pins are shorted together with the pink wire being soldered to both legs

resources/image/50449

For the S key again visually check the wiring then use a multimeter to confirm the switch is functioning

User avatar
flowerlandfilms

06 Dec 2018, 10:39

Anakey wrote: Your A key looks like the pins are shorted together with the pink wire being soldered to both legs

resources/image/50449

For the S key again visually check the wiring then use a multimeter to confirm the switch is functioning
AH! good catch!
The simplest explanation is usually correct, in this case that I am shit at soldering.
I fixed A but S is still aberrant. Multimeter says it has continuity only when pressed.
I might try swapping the switch anyway.

Anakey

06 Dec 2018, 11:20

Hang on, are you wiring up the numpad as a matrix or as single switches? If running as a matrix then you would need 4 columns and 5 rows so 9 pins. If doing individual switches then each switch would need to be connected individually on both pins. It looks like you mare making things more complicated then they could be.

I find it is best when soldering onto switches is to wrap the wire around the switch leg so it is physically anchored and in contact with the leg before then soldering in place. Also after soldering, cut any tails that remain after the solder joint as short as possible to prevent shorts. I would also recommend using slightly less solder on your Teensy some of your joints look like they are balling up on top of the pad like on D3 whereas it should be like on B1 that would prevent the pads near each other being occidentally bridged. I hope that you manage to figure out the correct pins for the trackpoint and can get that last switch working.

User avatar
flowerlandfilms

06 Dec 2018, 11:54

Anakey wrote: Hang on, are you wiring up the numpad as a matrix or as single switches? If running as a matrix then you would need 4 columns and 5 rows so 9 pins. If doing individual switches then each switch would need to be connected individually on both pins. It looks like you mare making things more complicated then they could be.

I find it is best when soldering onto switches is to wrap the wire around the switch leg so it is physically anchored and in contact with the leg before then soldering in place. Also after soldering, cut any tails that remain after the solder joint as short as possible to prevent shorts. I would also recommend using slightly less solder on your Teensy some of your joints look like they are balling up on top of the pad like on D3 whereas it should be like on B1 that would prevent the pads near each other being occidentally bridged. I hope that you manage to figure out the correct pins for the trackpoint and can get that last switch working.
Thanks for the tips! I'll keep those in mind for the future.
I got S working!
No matrix just individual pins, the Teensy++ has more than enough.
I think it was a short originally, but then when I moved it to another pin to troubleshoot, I forgot to add the pullup resistor for that pin into the code. I just told it to look for S on that pin, but without the pullup it couldn't find it.
All sorted now, onto that trackpad.
WP_20181206_21_40_07_Pro.jpg
WP_20181206_21_40_07_Pro.jpg (6.58 MiB) Viewed 5014 times

User avatar
flowerlandfilms

10 Dec 2018, 11:52

Success! I have X & Y coordinates printing to the serial monitor.
However I don't know how to alter the code to make the cursor actually move.
Does anyone have any handy links?
WP_20181210_21_42_05_Pro.jpg
WP_20181210_21_42_05_Pro.jpg (10.77 MiB) Viewed 4957 times

User avatar
flowerlandfilms

11 Dec 2018, 11:23

I've got mouse movement, but it's erratic. It's quite frustrating. I can't map to an absolute position because only later Teensy have Move.MouseTo.
It shouldn't be a problem because I can use relative movements with Mouse.Move but the mouse movement is all wacky.

Post Reply

Return to “Workshop”