Bongofish
October 16, 2018, 03:30:34 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
News: You can attach photos hosted by the forum rather than using an external image hosting site, this means they will stay forever and not disappear after a year or two.
 
   Home   Help Search Login Register  
Pages: [1] 2
  Print  
Author Topic: Wacom Science  (Read 30021 times)
bernard
Administrator
Hero Member
*****
Posts: 2590


pato mania


View Profile
« on: August 13, 2009, 04:37:35 AM »

I got my hands on a few wacom devices lately and have been scrutinizing it a little in case it could help doing my build (and also because I like opening stuff).

So I have decided to start this thread for discussing our humble findings about the 'science' of the wacom stuff we all love.

Wacom Mouse UC-500-12
One of the early Wacom mouse (works with the first generation UD-xxx series tablets) is original and "simple".

Here's the mouse cracked open showing its inner guts:

The heart of the mouse is not inside the mouse -- it is actually outside! It is in the transparent plastic around the round red "cross". It is the coil that interacts with the sensor board (or the antenna if you wish). If you look at the circuitry it has no transistors, no chips, just a bunch of little capacitors and resistors (one of them being variable) and the micro-switches for the buttons. There is nothing else -- the PCB is single layer (there is nothing on the other side). Electronic-wise, can't have a simpler circuit for a high-precision 4 button mouse!  When I say "simpler" I mean in terms of components -- because the principle and the physics to get there is a bit more on the heavy side. I never saw how a pen of that generation is built, but I would assume it would resemble it: few coils and capacitors / resistors.

I think the way it works is that the sensor board induce power and makes the coil "resonate" -- depending on the capacitors (and resistors) hooked to the coil (micro-switches deciding), the coil will "respond differently" and that will give out the status of the buttons essentially.

Like RFID?

This might resemble how a RFID system works -- basic RFID can essentially transmit a certain number of bits -- like 32 -- to output its ID essentially -- not sure how it works, but maybe wacom uses the same type of stuff - RFID gizmos are batteryless and the early ones were really simple (circuit wise) -- it had to cost near nothing to fabricate a gizmo.  They had like "26 fuses" (for 26 bits) -- I mean visible (if you looked at the circuits) -- programming a card (during manufacturing) meant to blow select fuses to achieve the desired ID.  If you look at the circuit carefully you can see 5 blob of solders numbered 1 to 5. 2 of them are actually "not connecting". I am thinking this is some kind of encoding -- maybe it tries to encode the device ID so the sensor can recognize what type of beast is floating on top?

Wacom intuos 4D Mouse (for intuos1 GD-xxx wacom boards)

This mouse sports 5 buttons plus some kind of "roller" on the side (that springs back in position when releasing it).

Ok let's open this thing -- I had some problems because there were 4 hidden screws behind some kind of "soft pad" under the mouse, had to find them.



Now this circuit is nowhere like the first generation -- it has a few chips -- including a custom Wacom one and tons of variable pots. This little device seems to pack a lot more technology.

One thing to notice is the main coil -- it is vertical (like a pen Smiley ) -- and there are actually 2 of them.  A 4D mouse can sense the "rotation", so I guess this is why there is two -- I think this is similar to the 6D Art pen (intuos3) -- it requires 2 coils next to each other like that to sense rotation -- there is a standard coil on one side, and another special one that have its wires going around both ferrite cores. 

There is a dip switch in there - I wonder what that does? For reference, on that one, only the '1' is "ON" the rest ('2' to '6') is "OFF".
A small 7 wire FFC cable (1mm pitch, more than 10cm) hooks a daughter board that hosts the 5 buttons. The big white gear is the mechanical device that hooks to the roller and that spring is what causing it to "come back in position". The PCB is a standard dual layer with SMT components on both sides.

I do not think a mouse is more complex than a pen (actually pens have an extra 'eraser' tip which doubles the trouble) -- so that probably means that all the electronic that I am seeing here must be "stuffed" into a pen? Quite impressive.
« Last Edit: August 13, 2009, 01:41:01 PM by bernard » Logged
bernard
Administrator
Hero Member
*****
Posts: 2590


pato mania


View Profile
« Reply #1 on: August 13, 2009, 07:30:29 AM »

To ADB or not to ADB

I started looking at the chances that some ADB-based boards be reconfigured for serial (PC) or even USB.

In general, stay away from ADB when you can to avoid bad surprises. But if you find a huge deal (or a free one) and you know a little bit about electronics or plain soldering it might be doable. This post is about trying to discover how this could be done.
ADB has a 5v power pin. Serial boards I have seen work with an external 12v dc supply (some stickers says a range of 9v..12v). So that's a difference to keep in mind.

EDIT: I built an adapter (WaxBee) to convert ADB tablets to USB for about 20$+shipping -- see http://forum.bongofish.co.uk/index.php?topic=1738.0

UD-0608-A

The UD-0608-A is a 6x8 first-generation wacom ADB board (the last digit 'A' in the model number means ADB).

in short: no way to convert this tablet to serial or USB. ** EDIT: now yes using WaxBee

Investigation:
I got that board for its pen and was hoping to get more info on ADB and the differences with serial or USB. I opened the tablet, and you could see that the PCB is built only for ADB. All chips are installed and the markings only talks about ADB signals. No trace of serial nor USB. So there is no way to reconfigure that one for sure.

I *think* smaller boards are typically less prone to 'reconfiguration' than the bigger ones (i.e. 9x12 and up). Smaller boards are built for big mass production and every component counts -- so it is cheaper to build a completely new products than "sharing" PCBs between models with "options".

 



GD-0912-A

This is a 9x12 intuos 1 ADB board. The last digit 'A' in the model number means ADB, The letter 'R' is for serial.

In short: Might be possible to convert to a serial one. (GD-0912-R). For sure it can't be converted by just soldering a new cable, it is more than that if at all possible/easy. This is still under investigation - I have both the GD-0912-A and GD-0912-R in my hands.



Investigation:
This tablet I got bundled with another one and was also a chance to learn more on ADB possibilities. The PCB seems to also be the one for serial since there is both the ADB and the Serial pinout described on the board (note: SPLY is for "Power Supply"). So there is some hope to reconfigure this to serial (but not USB -- unless I missed something). *If* it is possible, it will require soldering standard SMT components and a few dollars spent at digikey (probably not much) on the hardware side.



Missing components:
- A RS-232 driver IC (a chip) (did you say RS-232 - wow, that's a surprise!) part number: ADM202JRN
- a couple of what looks like transistors (not sure since I see other transistor with a TR in their name) -- these are named ZD1R and D2R
- a bigger component - probably some kind of triac or whatever (named IC2R). This is most likely involved in the DC-DC conversion 12v -> 5v -- probably a 5v voltage regulator circuit
- a mix of about 15 SMT capacitors and resistors
- ..and of course a cable is missing.

A portion of the missing is for the input voltage conversion -- with some luck, directly supplying 5V through the ADB connection might be doable (unless the RS-232 driver requires the 12v -- which is very possible -- or we could have 2 supplies -- I mean, leave the 12v -- but inject the 5v from another source).

Components only on ADB board (most of them probably won't need to remove):
- a MC68HC05 8bit CPU chip with its 64kBytes OTP EPROM (AT27C512)
- All components with a name ending with "A"

Must be removed (unless trying to bridge a 5V input source)
- R9A - 0 ohm resistor
- R10A - 0 ohm resistor

Now the scary/rough part:
The most difficult part would be the "firmware". There is a chance something in there has some firmware that is configured differently in the -A versus the -R model. If we are lucky, the firmware is 100% the same and it gets "configured" by auto-detecting external components. I know some board designs works this way -- by choosing to install those components it automatically configures the firmware. Hopefully they did that too - it will be difficult to tell without trying it out.

There seems to be a microchip serial eeprom in there IC7 (this is often used to store bootstrap firmware for microcontrollers OR just for plain settings). This part can be reprogrammed (but that's more involved work and more specialized equipment is needed!).

GD-0912-R

I got my hands on a serial version of the GD-0912 (no pen & no supply cable as usual for the real cheap ones). It has indeed the same PCB as the ADB version.





more investigation
First, two big ICs are just.. gone and when comparing I see one (or maybe two) '0 ohms' resistors that are used to jump missing parts - both are more or less near the "5V regulator" - only present in the serial board.

I realized that some component names end with "A" or "R" -- and those are installed on the ADB or serial version respectively. This is true for a lot of them, but not all (some components have no special suffixes).

The 5 Volt voltage regulator - (the part named IC2R) - part number: TA7805F. This is the heart of the DC-DC (12v to 5v) converter.  

I think the whole wacom board works with 5v excluding the serial port which requires roughly 12v to function (I assume this can vary a lot, serial port specs supports a wide voltage range, this would explain the 9v-12v input power range written on the sticker) - I know that PC serial ports will work with even lower voltage. ADB works with 5v so no regulator is required on the ADB version.  EDIT: the serial port is also running of 5v.  If you have a 5v source (like USB) it should be good to power the whole board without any extra electronics.

One possibility might be to avoid building the DC-DC converter and provide a 5v form another source. For the serial port to work, we still require 9v .. 12v source. This will reduce the number of components to add.

Now I can see all resistors values just by reading them. That is not the same story for the capacitors -- I think I would need to de-solder them to "measure" their value with a multimeter (not all multimeters can read capacitance btw, this feature seems rare/expensive) -- that is unless some clever chap tells me a better way of finding out. I would like not to touch my "intuos 1 serial" board too much if possible, but I am willing to desolder a few parts. I do not know how many capacitors needs to be measured there, but there are quite a few. Maybe by looking where they are connected might help knowing the value (especially the ones connected to ICs -- typically those are specified in the manufacturer App notes or directly in the datasheet.

Some parts I do not know what they are:
- ones that start with the letter "F" as in "F2".  (they look like little SMT capacitors but tend to be smaller??).
- transistor-looking parts not marked as "TR" - they are "ZD" and another one "D" -- My thought is "zener diode" and "diode".  Problem is that the package has 3 pins - puzzles me.


UD-0912-A (and bigger sizes like the UD-1212-A & UD-1218-A)
I do not think any UD-xxx series has a USB version (not sure though). So the only choice for UD-xxx is Serial or ADB.

Talking with other folks on this forum makes me think that the ADB model might be easily converted to serial simply by getting a new cable and power adapter (you can buy those or build them -- pin out available in the forum).

But there seem to be a confusion between different type of cables -- some boards come with a serial cable for macintosh (which is also a small Din plug!)  This is not ADB - althought is says "Macintosh" on the box and the plug looks like an ADB one (the serial port has more pins, but I think a 4 pin din might actually "work" too since the txd and rxd are mapped to the 4 primary pins).  More investigation is really needed here to uncover the dust around all this.

UD-1212-R
Below is the control board for a UD-1212-R -- there is a white sticker on a EPROM chip that says "UD-1212" (very pale in the photo). EDIT: it also says 15400 which, turns out to be the ROM version number (version 1.5-4). There are many versions out there and they have different capabilities (tilt, pressure levels, max speed).


Almost all components are present - all but few resistors and capacitors (2 of which are big). There is no mention of the output pinout (like you see conveniently in more recent models).

There is a microchip serial eeprom in there -- maybe this is to store the UDConfig data. Part number: Microchip 93C468

There is a big white 2 pin plastic connector on top -- nothing was connected there -- not sure what it is for? One side is grounded, the other goes into the right pad of the IC102 (TA7805F) -- a 5Volt voltage regulator -- (btw, I see the same part in the serial version of the GD-0912-R). So this really sounds like it has to do with power -- and it looks like it is an alternate input power since it is connected to the input side of the 5v regulator.  Also, the board seems to have a way to remove the regulator -- there are pads to put a 0 ohms (same thing as found on the GD-0912-A).  So the input power should normally by 12v but it could be 5v if the board gets reconfigured (not sure if that's complex to achieve), probably not.

EDIT:  I was able to successfully inject an external power supply of 5v into the board by connecting it directly to the regulator pad. I do not know if it is "good" to feed 5v like that without removing the regulator in the long run, probably not. There is another post somewhere else in the forum about it.
« Last Edit: January 10, 2012, 07:09:03 AM by bernard » Logged
bernard
Administrator
Hero Member
*****
Posts: 2590


pato mania


View Profile
« Reply #2 on: October 02, 2009, 09:04:43 AM »

UP-703E Pen: Now here's a PenPartner (CT-0405-R) single-button, erasing, pen. Not sure about the model (not written on the pen), but it is probably a UP-703E.

In theory, this UltraPen should work with all the Artz/ArtzII/Digitizer/UD/CT/PL/*SX/DTU/PenEnabled boards.



The eraser portion is a completely separate board not even touching the main board. In other words, this is physically two, totally distinct pens! The eraser has it own little tune potentiometer (the hole in the plastic).

This type of circuitry is almost the same as the UC-500-12 mouse. An antenna (soldered in the back of the PCB) and a bunch of capacitors/resistors. A couple of potentiometer to tune the pen and even the same type of "dots" that are either opened/shorted - like they were options or some sort of id encoding. For example, it looks like 3 parts are not connected on the right because 3 of the "dots" are "opened".

Even if this is an old pen, the circuit board seems to support a duo-switch (parts not installed). Also there seems to have space for a third potentiometer (probably to tune the second button), there is also big pads in the middle, not sure what this is for.

Important notes for disassembly

The pen was already "cracked" when I got it. To separate the 2 parts of the pen, you have to pull apart. There is nothing else to do it seems. Do not attempt to "turn", this will crack the pen.

To remove the little "board" or to access the tuning potentiometers, you have to remove the switch "cover". Look carefully at the picture, I positioned it like it is supposed to fit. You have to pry & lift THE BACK side of the switch cover (e.g. right side in the picture). I tried the front, but I kinda broke it (the plastic tip in the left side has bent and is now very weak).   Note: these notes are only for this specific pen. I have no idea how the other pens are like.
Logged
random-jimmy
Full Member
***
Posts: 79



View Profile WWW
« Reply #3 on: February 07, 2010, 01:45:49 PM »

What a great thread! Kudos on doing the work. There is a lot of interesting stuff happening in these tablets: and it is amazing too to see the difference in quality between the older and newer tablets in terms of design and also manufacture of the boards.

Although I am not too sure about converting tablets from ADB to serial: the tablets themselves would most likely use different firmware (unless somehow that CPU converts ADB signals to a serial connection). Also, it hardly seems worth the work: having to buy the custom parts (and solder them) might cost more than simply buying an RS-232 tablet.

And a note on those components:
F (as in F2) means fuse: these are small SMD package fuses - probably polyswitches (resettable fuses) instead of the normal sacrificial ones.
ZD and D: you are correct, they mean zener diode and diode respectively. They come in a 3-pin package simply because that's the package they come in: I think the two pins on the same side are bridged (probably depends on the diode as well).
Logged
bernard
Administrator
Hero Member
*****
Posts: 2590


pato mania


View Profile
« Reply #4 on: February 09, 2010, 10:56:53 AM »

Converting from ADB: Lots of "if"s.

Yes, there is a chance that extra cpu converts the serial signal to ADB and there is a chance the firmware is the same. ADB is also a serial thingy but there is a fundamental difference:  When sending a message, you have to wait for the bus to be available. So *if* it really "converts" it must know about the protocol somehow to deal when "too much data" is going out (what info should be dropped?)  -- or maybe it does not drop anything special (unless some internal buffer is full) -- they might know how not to exceed the maximum ADB data rate.

But yes, you are right that this seems quite a challenge + risk.  About buying all those components -- what kind of price do you think this could amount to? Since the ADB boards are no longer usable by any modern computer (I think there is still a possibility to use the ADB adapter but it depends on your MAC OS).  Their resell value is thus near zero. Quite a bargain for large, high-quality tablets!
Logged
random-jimmy
Full Member
***
Posts: 79



View Profile WWW
« Reply #5 on: February 10, 2010, 01:25:16 PM »

Buying those parts? Probably near impossible.

Buying the diodes will be cheap, and so will the voltage regulator, but what won't be cheap is buying those IC's, also considering they are probably no longer produced. Although, it would be possible to scab them off other (broken) serial boards i.e. just move the serial circuitry over to the ADB board. However, there is still the problem with firmware.

Also, keep in mind you will need a good soldering iron to do these. Look at around $100 for one - it needs to be temperature regulated, and have a fine tip (0.5mm works well for me). You will also need SMD tweezers - that's assuming you do not already have the equipment. Keep in mind this is very fine soldering work.
Logged
bernard
Administrator
Hero Member
*****
Posts: 2590


pato mania


View Profile
« Reply #6 on: February 11, 2010, 06:45:27 AM »

"those ICs"  -- To convert from ADB to Serial, there isn't much ICs left to buy other than the voltage regulator and serial port driver. Which is pretty standard stuff I think, no?  The serial driver seems to be available (2.15$ for a minimum quantity of 1 at digikey) http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=ADM202JRNZ-ND

Other than that, there are diodes/zener diodes, capacitors, fuses, resistors and the ones I am scared about: TRansistors.  I have the parts in front of me in the serial version of the PCB, but I do not know how to discover/measure what are the part style without de-soldering (and even then, the Transistors I have no clue how to discover what it is).  Do you have a trick?

The firmware is, of course, the most scary part, it is a bet. To simplify "production", the same firmware is sometimes used for the various board options. Especially true for "high-end" stuff whereby saving on the firmware capacity does not make much of a difference. Early Wacom high-end boards shared the same PCBs for all versions at the same time, and I do not see any "sticker-with-a-software-version" on the common ICs (only on the ADB-specific ICs) -- this alone, makes me think there might be hope... But again, this is a bet. The chance of all this working is still slim, I agree.

Yes, this is all crazy stuff. But for the record, I do have a temperature-controlled soldering iron, a fine iron tip, lots of flux, a good-enough (non-SMD) tweezer, flux cleaner, an 'Exacto', a steady hand, a good eye and a crazy mind (that last part is essential).  "Time" is the thing I miss the most I would say, so I steal time from other, much more important activities. I destroyed PCB pads in the past more than I can remember. I also get nice soldering tips from a friend which can solder large, very high pitch SMT IC packages without any specialized equipment -- just plain iron. Again, the chance of this to work is slim if I even do anything (I am blocked at getting the precise part list -- capacitors and TRansistors, etc.
Logged
random-jimmy
Full Member
***
Posts: 79



View Profile WWW
« Reply #7 on: February 11, 2010, 10:10:39 AM »

I didn't know that the serial driver IC would still be available.

As for identifyingother parts, the transistors (and possibly diodes) should have a faint code stamped to the top - I don't know if this identifies them, but it will get you somewhere. Resistors have the normal 4 number code, just in numerals not colours. And capacitors are the hardest: you will need an LCR meter, because to find their value, you will need to desolder them and measure them.

As for tools, it seems you are all decked out! Especially with those tips you were talking about... all I have ever used at work is simply a conical or a chisel tip. Just make sure when you are soldering that you keep the temperature at around 380-400, and you should find it hard to lift a pad. Desoldering parts is the hardest: if you don't want to keep them, just destroy them.
Logged
bernard
Administrator
Hero Member
*****
Posts: 2590


pato mania


View Profile
« Reply #8 on: February 12, 2010, 05:45:48 AM »

Destroy: Dremel to the rescue!  Dremel the pins out of the package and desolder the pins separately.  Grin
Logged
random-jimmy
Full Member
***
Posts: 79



View Profile WWW
« Reply #9 on: February 13, 2010, 11:25:28 AM »

Haha thats what I do  Smiley I think I've learnt here that if you don't do some damage in the course of your build, then something is wrong Wink
Logged
Wilcorp70
Jr. Member
**
Posts: 63


View Profile
« Reply #10 on: February 26, 2010, 08:30:02 PM »

     I have been trying to do some basic research on the UD series tablets.  Tablet PC digitizers use this same series, if the different UD series tablet boards used a common control board to standardize parts, it makes me wonder if one of these controller boards couldn't be wired to the daughter board of a tablet pc digitizer. 
     Does anyone know how many pins are on the connector to the UD 1212 board and info on the pinout?  There was a post from a couple of years ago about tablet pc digitizers having a 10 pin interface with 3 for power 7 for data I have been meaning to confirm this using my own toshiba tablet but haven't gotten around to it.  I know its a long shot, but if these two could interface even if it requires new firmware it would be a huge benefit to those hoping to add tablet capabilities to laptops.
Logged
bernard
Administrator
Hero Member
*****
Posts: 2590


pato mania


View Profile
« Reply #11 on: February 27, 2010, 12:46:19 AM »

I am not sure I understand the "benefit" if we find a way to connect a PC Tablet "digitizer" board to a UD-* daugther board?  Do you mean to be able to reuse just the screen LCD and wacom sensor portions and throwing away the laptop motherboard?

The "Penabled" PC Tablets indeed share the technology with the UD-series (pens are compatible for once). And there were "kits" made by wacom specifically targeted for people to create PC Tablets ,these kits had 2 sections, and yes, it seems "probable" that it could be done.

The daugher board (for which there is a close-up picture in this thread) uses the FFC connector labelled "CN001" -- I do not have it handy to check (buried in a box in the basement) but according to the picture, I see ... 17 pins  Huh -- which is a rather unusual number Huh -- I must be missing something.

If someone can open its Penabled Tablet PC and look at the wacom sensor connection (with detailled pictures of the boards), we might work together to investigate this possibility. Smiley

Edit: another thread now contains good information about this topic  http://forum.bongofish.co.uk/index.php?topic=1659.0
« Last Edit: August 02, 2010, 04:32:17 PM by bernard » Logged
Wilcorp70
Jr. Member
**
Posts: 63


View Profile
« Reply #12 on: February 27, 2010, 02:55:44 AM »

Ok, I can see how my previous post wasn't particularly clear.  So, as for advantages, Im not talking about throwing away the motherboard of a tablet pc, there are many lcd + digitizers available on ebay, etc.  These can be picked up fairly inexpensively and the digitizer can be easily removed from lcd and placed behind any screen (ignoring the need for shielding of course).  It really just makes it easier to install directly into a laptop since the regular tablets are a little bulky for the conversion, additionally, the antenna area along the side is smaller compared to the active area.
      When I said the daughter board, I meant the daughter board on the tablet pc digitizer.  The different definition here of daughter board gets confusing quickly when comparing the two.  Unlike the "regular" tablets, the pc uses a daughter board attached directly to the digitizer and then interfaces that via a cable to the controller electronics on the laptop mother board.  You mentioned the kits made by wacom, which they still sell by the way (if you are an oem), but after looking at the kit's part numbers on their product page they are the same tablet pc digitizer attached to a "standardized" controller.
     I counted 17 pins in that picture too, I was hoping I was just looking at the wrong part, cause you're right, that makes little sense.  I will be taking my tablet apart in the next week, Ill snap some pictures or find some on the web.  Ive been thinking about this for a while, but until your thorough investigation of the UD series tablet I didn't have much to go on.  Thanks for an awesome post.
Logged
bernard
Administrator
Hero Member
*****
Posts: 2590


pato mania


View Profile
« Reply #13 on: February 27, 2010, 05:42:45 AM »

Let me get that UD-* controller board in real, I'll take a closer look at the connection. I tried to look at a (tiny!) picture of the oem kit -- true that there is what they call a "baby board" sitting between the sensor and the more complex controller board.  Would like to see a closer picture.
Logged
Wilcorp70
Jr. Member
**
Posts: 63


View Profile
« Reply #14 on: February 27, 2010, 05:52:49 AM »

Yeah, I tried staring at that interface board too, made my eyes hurt.  You're right they called it a baby board not a daughter board, I hadn't looked at it in a while.
Logged
Pages: [1] 2
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Valid XHTML 1.0! Valid CSS!