Experimenting with the HP 2710p Tablet PC digitizer (SU-12W18A-01X)

[ThrowingChicken]

I would probably have to solder a wire to pin 8 but I can give it a try. 

Did the voltage values help at all?  Is it odd that I am getting 4v instead of 3.3v? 
I still have the laptop opened up and the digitizer hooked up; are there anymore values I should try to get?  I'd rather not having to take the laptop apart again, so if you can think of anything I should test now would be the best time to do it. 

[Aerendraca]

I've just read how you've been getting on with this and it seems to be going quite well so far, i wish you luck!

Am i right in thinking that you measured 4volts until the pen was placed close to the digitizer when you then read 3volts? This sounds about right as it is likely you were measuring the voltages without a load, since the digitizer is only fully switched on when the pen comes into proximity (assuming it operates as per the intuos/bamboo range).

Look forward to see how you get on.

[ThrowingChicken]

Thanks!  Admittedly electronics are not my strong suit so the fun won't really begin for me until I start working on the case; I think that is the part I am most excited about since I like to build crazy things.  Well that and actually using the tablet with my desktop.

I'd say that voltage assessment is partially true.  I had zero readings on all of the pins except for 9 (which isn't hooked up to my laptops motherboard I am a dummy, it is connected after all), 10, 12, and 13 (they all read 4v).  When I put the pen to the tablet I get a 4v reading on pin 6 (which initially ran zero and is also not hooked up to the mobo), and pin 9 changes from 4v to 3v.  None of the other voltages change.  I am using just a cheap analog multimeter from Harbor Freight so I suppose it could not be giving me the best readings, but I would imagine there would be at least some slight movements with the needle if the voltage was changing with the use of the pen but I am just not seeing it on pins 10, 12, & 13.  

[ThrowingChicken]

USPS says the Teensy might be here tomorrow!  Once it gets here, in what order should I hook up the pins to it?  Is there a good tutorial available? 

[bernard]

Warning: Huge post.

Nomenclature:
"high" means closer to VCC (3.3v or 4v in your case).
"low" means close to 0v/GND/ground level.

huh, did you just said that pin 9 (RXD) is *not* connected to the mobo?  According to other serial boards, this is supposed to be the main serial data pin. (From the tablet to the PC). It is supposed to be "high" until there is some data upon which it goes low for split seconds -- your multimeter will "combine" all the high-low levels and thus give you something a little lower than the normal "high" (here your 4v going to 3v).  If you had a "scope" you would see UART data (pulses going down) -- also with a "scope" you could measure the size of each pulse and we could compute the serial port baud rate instantly.  

Hum a (oscillo)scope... Since you like building strange (and real cool btw) stuff, you might get tempted by the dark side of electronics... Anyways.

I was wrong about pin 8 in my earlier post (no need to solder anything there, very sorry!).  It is pin 10 that is the TXD that we need. This should thus be "high" almost 100% of the time. (Until we start sending commands to the tablet, something that almost never occur during normal operation -- only when the "driver" initializes to query the tablet).

For tutorials -- not much for serial ISDV4 actually -- Or you'd have to dig posts on this forums from people that tried this. Most of the tutorials are for interfacing with the non-penenabled wacom serial and ADB digitizers.

First you gonna need to make sure the Teensy you got works. That made me think :  You have a spare mini-usb cable, don't you?   You are gonna need this.   So first thing is to plug it in naked and see its led blink.   Next step (unplug it) is to solder the 3.3v regulator (there is a trace to cut and a solder-jumper to do I believe -- follow the instructions from http://www.pjrc.com/teensy/3volt.html ).  When all done, replug the Teensy should still blink. Use your multimeter and with the black on a GND pin, probe the voltages, you should be able to get 3.3v on the VCC pin(s) (there are two of those, pick the pins away from the connector, it is more convenient). There is enough Amps left to power your digitizer from this 3.3v output.

On the Teensy there are basically 4 wires that are going to be needed. GND, 3.3v (VCC), TXD(D3) and RXD(D2). It is always a bit tricky to know which is which when connecting TXD and RXD. Labels for UARTS are always a mess. You have to know the "direction". Look at our wiki to see the direction (tablet --> host  or  tablet <-- host) and from the Teensy side, "TXD" or "TD" means "Transmit Data" and "RXD" or "RD" means "Receive Data". Make a small drawing on a napkin and double check.  You can check the "tutorial-like" instructions and picture of the Teensy in this post http://forum.bongofish.co.uk/index.php?topic=1930.msg14217#msg14217 to help you out.

Speaking of wires, I try to use black for ground, red for power (VCC) and other colors for the rest. When I see a red wire I stay away and a black wire is easy to spot when looking for a ground signal. But, of course, for a 4 wire project, it does not really matter, you can even put little stickers onto the wire with names on it. (i.e. GND, VCC, TXD, RXD)

BTW, when you are done, it would be nice to have a connection summary with a couple of pictures of the finished stuff to help out the next folks that want to do this.

Warning: These small wires are not always easy to solder on. If these are the micro-coaxial types you are not going to be able to do anything with them. We tried hard.  Try to skin and solder onto the ones that are not connected before cutting the other ones.  You'll see immediately if it works on not. (use your multimeter to check the continuity).

BTW, your multimeter is analog right? (with a needle): Does it have the continuity test (often with a beep for the digital ones, but it would work with a "OhmMeter" function too).  If unsure, attach a picture (or a link to a picture of a similar model), detailed enough so I can read the letters.  Also can it measure Amps? (in the 20-200 mAmps range -- that's milliAmperes or 0.02 to 0.2 Amps (also written as just "A" or "mA"))-- of course the voltages are below 5 volts.  When doing some tinkering it is always good to keep an eye on the amount of current the board is taking. If there is a short or something is wrong, the Amps will go high -- very high! Looking at the Amps reading can "talk" to you. It tells you what the digitizer is doing. For example, when approach a pen, the amps will go up, when you move the pen away, it will stay high for like a second or two and then go down in "idle" mode.  To measure Amps, you have to connect the leads "in serial" (like the power must go "through" the multimeter). mAmps numbers should vary from 0 (powered off) to under 100mAmps roughly like 80mAmps or even lower. Anyways. Without that, just quickly touch the big black Wacom chips to check if they seem to be burning-hot -- do this quickly after powering it up. You can also just power it for less than a second and check if anything became hot.  Watch out static electricity,  Always "ground" yourself before to discharge the static electricity in your body (like touching the USB connector in the case of the Teensy).

[ThrowingChicken]

Actually you are absolutely right, I have no idea how I got it in my head that 9 was not connected.  Looking back at my previous photo its 5 wires, pins 9-13.  Sorry for the confusion, I will update my chart. 

[bernard]

Pin 13 is definitely the VCC 3.3v power. On the Wacom PCB, the trace is much larger than the others (which is typical for power and ground). You need to connect this to the Teensy VCC (where you previously read 3.3v after the voltage conversion mod).

Pin 14 looks like a ground -- but because the DGND is connected to it -- it might be why it is not connected -- From what I can gather, DGND is a "Digital Ground" and is typically more "noisy" than a AGND (Analog) -- or plain GND. The ground must be going elsewhere.   I see "AGND" on the board. Check on which pin it is connected to (did you said you had a continuity or if not, Ohm-meter capability? -- btw -- *never* test continuity/ohm on a *powered* board. For connecting to the Teensy GND, both will most likely work, I think I prefer AGND, but anything that would already have a connection through the cable is much more convenient of course. -- I think AGND is normally Pin 1, right?  

Then Pin 9 and 10 to the Teensy D2 / D3 (still have to figure out which goes where exactly). If you "reverse the polarity" it is not going to burn anything, but it is not "great" to keep it reversed for long time.

[bernard]

Pin 11 and 12 always been quite mysterious and contradictory. From your readings, it seems there is only 1 pin that is connected that stays low all the time (pin 11)  --  why did you put the letter "G" for pin 1 and 14?  You meant "Ground" I suppose(?).    In the laptop, was there anything else connected to either digitizer boards that would ground them?

To be on the safe side, I would find a good connection to AGND (that's pin 1, right?) and use that as our ground to the Teensy. I do not really trust pin 11 to carry the main ground.

EDIT: I took the liberty of modifying your small "voltage table" with a monofont tag (teletype "tt") -- I used a text editor with a monospace font (courrier) to align stuff and copy-pasted it back. 

[bernard]

Added an entry for your tablet in the wiki page. If you happen to have more technical and physical info about the board to fill in the blanks, it would be great to keep around.
http://wiki.bongofish.co.uk/doku.php?id=bongofish:penenabled

[ThrowingChicken]

Hum a (oscillo)scope... Since you like building strange (and real cool btw) stuff, you might get tempted by the dark side of electronics...

Thanks!  And I will keep that in mind.  I remember seeing one at a science exhibit as a kid and loving the effect.  In the meantime, if you think it would help with the build I might see if I can track down someone nearby that has one.  

I was wrong about pin 8 in my earlier post (no need to solder anything there, very sorry!).  It is pin 10 that is the TXD that we need. This should thus be "high" almost 100% of the time. (Until we start sending commands to the tablet, something that almost never occur during normal operation -- only when the "driver" initializes to query the tablet).

Oh good, I was worried about having to solder something so small!  And yes, pin 10 reads high the whole time, at least during the operations I was performing.  

First you gonna need to make sure the Teensy you got works. That made me think :  You have a spare mini-usb cable, don't you?   You are gonna need this.   So first thing is to plug it in naked and see its led blink.   Next step (unplug it) is to solder the 3.3v regulator (there is a trace to cut and a solder-jumper to do I believe -- follow the instructions from http://www.pjrc.com/teensy/3volt.html ).  When all done, replug the Teensy should still blink. Use your multimeter and with the black on a GND pin, probe the voltages, you should be able to get 3.3v on the VCC pin(s) (there are two of those, pick the pins away from the connector, it is more convenient). There is enough Amps left to power your digitizer from this 3.3v output.

Great pictorial!  Looks simple enough.  

It is always a bit tricky to know which is which when connecting TXD and RXD. Labels for UARTS are always a mess. You have to know the "direction". Look at our wiki to see the direction (tablet --> host  or  tablet <-- host) and from the Teensy side, "TXD" or "TD" means "Transmit Data" and "RXD" or "RD" means "Receive Data".

But because we see highs and lows on pin 9 with pen activity, this is most likely RXD because this pin is sending data out to the computer, right?  

You can check the "tutorial-like" instructions and picture of the Teensy in this post http://forum.bongofish.co.uk/index.php?topic=1930.msg14217#msg14217 to help you out.

Oh yes, it looks like you are working with a whole different beast there but it helps me understand what the different pins are doing on both Teensy and the tablet.  

BTW, when you are done, it would be nice to have a connection summary with a couple of pictures of the finished stuff to help out the next folks that want to do this.

I will supply you guys with everything I have.  Even though my digitizer is not the same as others, just the fact that people were doing this with others is what got me to take up the challenge (and for me, this is a challenge).  This place is great and I'd love to contribute.  

If these are the micro-coaxial types you are not going to be able to do anything with them. We tried hard.  Try to skin and solder onto the ones that are not connected before cutting the other ones.  You'll see immediately if it works on not. (use your multimeter to check the continuity).

Never heard of these before, but it got me worried enough to test it.  I cut off a segment from one of the unused lines and soldered the ends to another wire.  We are good to go.  

BTW, your multimeter is analog right? (with a needle): Does it have the continuity test (often with a beep for the digital ones, but it would work with a "OhmMeter" function too).  If unsure, attach a picture (or a link to a picture of a similar model), detailed enough so I can read the letters.

I think?  It's the thing I use to test for a closed circuit, right?  This is mine right here.

Also can it measure Amps?

I believe so.  It has a DmA setting.

To measure Amps, you have to connect the leads "in serial" (like the power must go "through" the multimeter).

So while experimenting, it would probably be best to use some sort of connector instead of soldering the lines directly to the board, that way I can move things around and splice in the multimeter easily.  Do I need a multimeter for each of the four wires to watch for spikes or do I only need it on one wire?

Pin 14 looks like a ground -- but because the DGND is connected to it -- it might be why it is not connected -- From what I can gather, DGND is a "Digital Ground" and is typically more "noisy" than a AGND (Analog) -- or plain GND. The ground must be going elsewhere.   I see "AGND" on the board. Check on which pin it is connected to..... .....For connecting to the Teensy GND, both will most likely work, I think I prefer AGND, but anything that would already have a connection through the cable is much more convenient of course. -- I think AGND is normally Pin 1, right?  

I just tested it and the results are that both pin 1 and pin 14 are connected to AGND and DGND, which I guess would also mean AGND and DGND are connected.  I scanned the board looking for another pad but I am not seeing a plain GND.

Pin 11 and 12 always been quite mysterious and contradictory. From your readings, it seems there is only 1 pin that is connected that stays low all the time (pin 11)

That is correct, I didn't get any readings on it at all.  

why did you put the letter "G" for pin 1 and 14?  You meant "Ground" I suppose(?).

Yes, G meant that it was connected to the DGND pad (and now we know, the AGND pad as well).

In the laptop, was there anything else connected to either digitizer boards that would ground them?

Hrmmm, there are these pads were the screws go on the connector that plugs into the motherboard.  I can get a connection between them and the GND pads.  



I played around with it a little more and I know pin 1 and 14 make contact with the silver housing the connector slides into.  That conductive fabric is attached to the housing, so I initially thought this was how the ground was transferred, so to test this I pealed it back so it no longer touches the metal at all, yet I still get a connection between this housing and the end of the cord (where it connects to the laptop).  So now, really, I have no idea how the connection is extending the length of the cable.  Maybe once I strip the wires something will start to make sense.  What do you think?  

Thanks for the help, I will follow the Teensy instructions when it gets here.  

[bernard]

Ground: The outershell of the cable?  It does not really matter how it works in the laptop. What matters is that you find a good ground for your purpose that is well connected to the AGND/DGND. Maybe soldering onto the pin 14 ? It looks like a nice spot to solder a wire.

Scope: You do not really need it. Only if we run into trouble and we try to diagnose what is happening. We'll know only after you power up the digitizer using a 3.3v source.

But because we see highs and lows on pin 9 with pen activity, this is most likely RXD because this pin is sending data out to the computer, right?

yup.

I think?  It's the thing I use to test for a closed circuit, right?

yes -- I call this a "continuity test" (to see if something is connected with another thing). I looked  at your multimeter and you have the Ohm meter (that Omega symbol on the top left). Again, never use this mode with a powered circuit. Probably any value is fine, but I would pick the highest setting 1K. Touch the probes to see the effect of a "closed circuit".

Amps: yes, you have a DCmA -- which means "Direct Current milli-Amperes" -- this only works on DC current -- which is what we are dealing with here.  Pick the 500 setting if you make that test (as in "500mA").

So while experimenting, it would probably be best to use some sort of connector instead of soldering the lines directly to the board, that way I can move things around and splice in the multimeter easily.  Do I need a multimeter for each of the four wires to watch for spikes or do I only need it on one wire?

I use colored crocodile clips wires to perform temporary electronic setup. Like tie wraps, I never have enough of these. They are easy to find even in local electronic shops (if you know one nearby, call them to see if they have something in stock)  You are probably going to be fine with a single multimeter. The only thing would be to have another one to watch for the amps at all times as a "safety", but it is not really necessary.  If you have a really bad short and you send a lot of current, it could burn before you can notice it anyways.  The much better safety is to use a "lab" power supply -- which you can limit the current to the minimum.

[ThrowingChicken]

Alright, I am all wired up.  Do I need to load something onto the Teensy before plugging it in? 

[ThrowingChicken]

I am not so confident I have found my ground.  The amps seem off the chart on the meter.  If I had to, what if I soldered a line directly to the DGND pad?  It seems to be more out of the way than the AGND one. 

[ThrowingChicken]

So I am pulling 1.1Amps between the Teensy and the digitizer, which is a far cry from the 80mA we were told to expect.  I tested the digitizer and Teensy and both to seem to function just fine.  Are we sure 80mA is correct?  If 1.1Amps is possible, what kind of stress will this be putting on the Teensy board/regulator? 

[ThrowingChicken]

Boy what a rough day.  Here is the rundown. 

- Teensy regulator is in place, wires have been soldered.
- Wires connecting to pins 9, 10, 13, and ground have been soldered or stripped, connected to the Teensy with clip wires.
- Multimeter is connected in series between VCC and pin 13. 
- No matter what I do the multimeter reads ~1.1Amps. 
- Took an extra interface cable and spliced in the Multimeter on the line going to pin 13, plugged the cable into the laptop in efforts to figure out the native amp measurement.
- Multimeter read around 1Amp, the wire ran hot.  In the brief moment it was on we were unable to get the tablet to recognize the stylus.
- Soldered back together the wire going to pin 13 in efforts to retest the display.  Wire ran hot, melted the electrical tape. 

And on a slightly unrelated/related note:
- Removed the experimental digitizer and reconnected the permanent assembly only to find that my laptop would not boot up.
- Windows repair disabled the digitizer driver.
- Windows boots up, no pen abilities. 
- Reinstall drivers, no installation fails
- Try again
- Try again
- Try again
- Try again, installation successful.  Pen abilities back.
- Open Photoshop, digitizer driver crashes.  No pen sensitivity. 
- Try again
- Try again
- Reboot, try again
-  Try again
- Reinstall drivers again, successful, Photoshop and pen sensitivity successful.  Everything seems to be working now.

So what do I make of all this?  I think the experimental digitizer is probably fine.  I think the short was taking place somewhere between laptop and the digitizer connector, somewhere in the cables, based on the heat coming off the cable instead of the digitizer board, and the way my laptop was acting after the test.  I won't be taking my laptop apart again.  Besides sacrificing the cable in hopes to measure the digitizers native current; I don't want to risk messing up my laptop again.  My old laptop digitizer is better than no digitizer at all. 

But I'm not done yet.  I believe there is a short somewhere, some how, between these supposedly insulated cables.  The next thing I plan to do is put shrink tube over all of the wires coming from the pins, the full length of the wire.  If that is where the short originating, as I believe, then that should take care of it.  If that doesn't work then I am going to lose the wires and the connector and solder leads directly to the pins.  And if that doesn't work, then, well, I'm at a loss. 

Thoughts?