This equipment (a Microgrid III 36" x 48" graphics digitizing tablet ca 1992) is a tad esoteric at this point..no factory service, tech manual available, etc. I'm a student trying to resurrect this old bird for my GIS classes.
Serial interfacing software (Vtablet) is available for connecting this to my XP Pro Compaq laptop. I bought a serial port to USB adapter (Keyspan) for my MGIII which self-installed at COM5. Since the COM1 port was occupied, I reset the serial Keyspan adapter via the Windows Device Manager to COM2 and set Vtablet to COM2 as well. Following recommended start-up procedure (Microgrid tablet on, then boot) I am unable to elicit any puck output (screen cursor movement in response to moving the puck on the tablet) which is also confirmed as zero port interrupts on the software's Status tab.
I am now concerned that the tablet controller is misbehaving . I also note that the Microgrid start-up only produces two tones instead of the ?all okay? three tone self-diagnostic (don't know the implication of the missing tone).
The unit has a removable 8" x 13" controller board and I was wondering if anyone out there could direct me to an circuit troubleshooter who could look it over and hopefully repair???
(1) that the serial port adapter has installed properly. I've used this or a similar unit and the installation can be buggy especially when switching to a different port. Either connect it to another PC with a known working seerial port or select in a terminal emulator that is known to work with COM5 and see if loopback works.
(2) that the tablet is actually outputting something on the serial port. Make sure it either doesn't need any hand-shake signals (CTS, RTS, etc.) or that they are set properly.
(3) that the cable has pins 2 and 3 swapped, or not swapped as required. When connected correctly, there should probably be a negative voltage on both pins 2 and 3 with respect to pin 5. If incorrect, then there will be a voltage on only one of those pins.
--- sam | Sci.Electronics.Repair FAQ:
formatting link
Repair | Main Table of Contents:
formatting link
+Lasers | Sam's Laser FAQ:
formatting link
| Mirror Sites:
formatting link
Important: Anything sent to the email address in the message header above is ignored unless my full name AND either lasers or electronics is included in the subject line. Or, you can contact me via the Feedback Form in the FAQs.
On Sat, 10 May 2008 14:08:01 GMT, "powerseed" put finger to keyboard and composed:
This is the user manual:
formatting link
Make sure you have matched the communications parameters (baud rate, parity, etc) at both ends of your cable.
Here are some recommended settings when used with Vtablet:
formatting link
To narrow down your problem, you could connect your tablet to a PC with a real serial port and try this suite of test programs:
formatting link
It includes an MGIII test.
Alternatively you could try a DOS terminal emulation program such as Telix or Xtalk which can display incoming data in hex mode. You could also set the tablet switches for ASCII output mode.
The user manual says that anything less than 3 tones means your tablet has a problem. On the plus side, the fact that it produces any tone at all suggests that the CPU and ROM are probably OK.
If the PCB has socketed ICs, then you could remove them one at a time to see if the error tones change. For example, if removing a RAM chip changes your error report from 2 tones to 1 tone, then you can be reasonably confident that your RAM is OK.
Can you upload a photo to your web space? It may jog my memory.
You could try TabletWorks as an alternative to Vtablet:
formatting link
- Franc Zabkar
--
Please remove one \'i\' from my address when replying by email.
I don't have access to another PC w/ a serial port and am not a programmer with familiarity with terminal mode (what can you expect from a clueless mac user :). I am pretty confident tho that this is a tablet problem as you may concur after reading my following responses.
Serial port switches set to mfr's defaults re handshake, etc. I am also non compos mentis when it comes to troubleshooting electronics but it did give it a shot. I have a cheapo R. Shack digital auto-ranging digital multimeter (that produces wandering phantom readings when not connected to a live circuit). I tried to document output from the DB9 female on the serial adapter with the board powered up. Nada when I put the black lead on P5 signal ground and the pos on P2 (during this test I was moving the puck and pressing buttons as well). When I tested P5 and P2 it would momentarily show
0 volts as I was moving and clicking then it begin its up and down wandering again???
I'm not using a serial cable...DB25 female plug to male DB25/female DB9 adapter to DB9/USB adapter to USB cable to PC
See above...and my responses to Frank Zabkar in this thread.
My best,
Bill Curry
--
Message posted via http://www.electronicskb.com
This sounds like a great idea...will see I can find an old PC to test.
This is beyond my PC skills at present;-)
Looks like thru-hole soldering to me.
See
formatting link
for pix you requested.
Other points:
**The left LED (yellow) on the puck is supposed to be on when the 16-button cursor is in use, and off when the stylus is being used. It is alway off in all my tests. The middle (green) LED is supposed to come on when the puck has proximity (within the table's electromag field)--it always glows dimly when the table is powered up. The third LED glows dimly all the time also...it's for programmable alerts.
**The tablet came sans power supply. Based on the specs I picked up a 12volt,
1 amp supply. Manual says mfr's p.s. must be used or wtty is void. Is this a possible issue?
**I note there are what appear to be contact tabs (labeled +12v AC, -12v,
+12v, +5v, DIG GND, DIG GND) at the bottom right (plan view). Are these for test fixture powering?
**Misc.--Unknown variable...There are also two blue, 2-receptacle "pin jumpers" (labeled W3 and W5) seen at the bottom center of the board (plan view). Each jumper has two possible positions (left or right) among 3 pins.
I did this a while ago with no response. Current s/w (Cartalinx, Clark Labs) programmed for Vtablet.
Thanks Franc!
--
Message posted via http://www.electronicskb.com
On Mon, 12 May 2008 18:12:53 GMT, "powerseed via ElectronicsKB.com" put finger to keyboard and composed:
Page 15 of the manual states that the tablet needs a 12V ***AC*** 1A power supply. Instead you have provided a 12V ***DC*** 1A power pack.
Maybe. Measure them with a multimeter. I'm betting that the +12VAC (???) and -12V rails will be missing. The -12V is required by the tablet's serial port, and possibly by the analogue circuitry for the digitizing surface, and for various op-amps.
I'm betting that the 2-tones error indicates a problem with the analogue circuitry.
If you get your MG3 working, I would still run the MG3TEST. I've seen lots of tablets with dead spots on the surface, usually caused by dry solder joints, or bad driver chips (yours appears to use the ULN2003 chips).
BTW, if you ever need any spare ICs, keep me in mind. I have a lot of chips of that era ... somewhere. You'll just need to pay for postage from Australia. I also have a new unused 4-button Summagraphics puck, but that cost me money.
- Franc Zabkar
--
Please remove one \'i\' from my address when replying by email.
I'd go with checking the power supply/requirements as mentioned in the other reply. Indeed, if it specifies 12 VAC, it probably won't work correctly on 12 VDC (but won't be damaged either).
The fact that you're getting no change in the LEDs also suggests that it's not a communications problem. Probably not getting through self test due to the power issue.
--- sam | Sci.Electronics.Repair FAQ:
formatting link
Repair | Main Table of Contents:
formatting link
+Lasers | Sam's Laser FAQ:
formatting link
| Mirror Sites:
formatting link
Important: Anything sent to the email address in the message header above is ignored unless my full name AND either lasers or electronics is included in the subject line. Or, you can contact me via the Feedback Form in the FAQs.
Oye-- How dumb is that!!! So this is just a step-down transformer!! I thought all the bricks were AC to DC. Do you think using 12v DC could have hurt the board? Or that it could be the source of my problems??? Not many of these supplies out there. Did find one with a 3' cord though that I will pick up.
Didn't get any current when board was plugged into my !@$#$@% power supply-- perhaps AC will make a difference.
I also saw at p. 40 of the UG that there are THREE MG3TEST diagnostic tests: RAM, X-axis, and Y-axis--perhaps corresponding to the 3 tones??? Or not! If so tho, might suggest Y-axis circuitry.
One ULN2003...along with six UDN2595-A dips which I think are drivers.
Will keep all that in mind!
Thanks!
--
Message posted via ElectronicsKB.com
http://www.electronicskb.com/Uwe/Forums.aspx/repair/200805/1
On Tue, 13 May 2008 01:57:41 GMT, "powerseed via ElectronicsKB.com" put finger to keyboard and composed:
As Sam has said, DC won't hurt your tablet. Using DC instead of AC means that some negative voltages are not able to be generated within the tablet for important functions such as the serial port and for your signal amplifiers (op-amps).
I suspect that the power-up sequence starts with a basic digital test, ie ROM/RAM/CPU check, after which the first tone is output. Then perhaps the next level of circuitry is tested (eg the 2651 serial controller chip), resulting in a second tone. The next and final set of tests may involve the analogue circuitry, ie the X & Y wires, which would fail in this case because of the missing -12V rail. But that's just a WAG ...
- Franc Zabkar
--
Please remove one \'i\' from my address when replying by email.
Well, I got my new 12v AC power supply! The plug was incompatible so I cut it off and replaced it with one that fit. The great news now is that I get all three required diagnostic startup tones and the green puck cursor LED blinks on as it should whenever I move it out of proximity to the prescribed electromagnetic digitizing area. Since it is showing proximity bounds around the board perimeter, I assume it is reporting an x/y coordinate stream to the board controller.
Q: Do I have to be concerned about plug polarity with this 12V AC power supply?
I am proceeding to troubleshoot the communication side now that the tablet appears functional. Unfortunately when I plug in my serial to USB connection per my previous post, no COM2 port activity is observed when I move the puck and push buttons.
Switches: Firstly I rechecked my tablet switch settings to confirm they matched the spec given with my Virtual Tablet software.
Interfacing Hardware: As I read it from their manual, the Summagraphics tablet uses a three-wire, null modem interface with loop back handshaking. They recommend a standard DB25(male) to DB9(female) converter/adapter cable. I use a DB25/DE9 converter plug into the female DB25 table port A instead of a cable. I checked male/female continuity on this adapter plug to insure it was a null modem configuration...2 on DB25 to 3 on DE9, and 3 on DB25 to 2 on DE9. The DE9 female of the converter plug is connected to the male DE9 of a Keyspan serial port adapter which is USB cabled to my Compaq laptop.
Output Voltages: With my crappo multi-meter I first checked the DC voltage of the DB25 female serial port A on the tablet, particularly, the TX (2)/RX(3) pins compared to signal ground (7).
Pin 2 to Pin 7: variable 0.7 to 1.4 volts DC Pin 3 to Pin 7: -5.8 volts DC
Q: Does this imply the tablet is streaming data out?
COM port assignment: The Keyspan USB/DB9 male serial adapter has DB9 pin-outs analogous to a DB9 socket on a PC (23=RX, 3=TX, 5+GND, 4&6=DTR/DSR,
7&8=RTS/CTS, 9=ring). One first installs the Keyspan driver software and then the adapter is connected to the PC via USB cable. The doc says that upon plugging in the Keyspan, it will take the first open port after those that have been assigned or set aside. It auto-installs at COM5. Problem here is that the Virtual Tablet software only allows connections via COM1 through COM4. Device Manager only shows COM1 taken for the modem.
Using a Keyspan utility program, I reassigned the Keyspan to COM2 and Device Manager recognizes it at COM2 as available. Keyspan diagnostic software has a port test function and it passed ?transmit only? and ?internal loop back? tests. I assume this means its functioning properly in COM2. I ?ll have to get a DE9 serial loopback connector before I can run the ?external loopback? test.
When I plug the keyspan into the tablet, power up the tablet and then boot the PC, the Keyspan port mapping diagnostic show the adapter on COM2 and ?Busy?. Another diagnostic provides status reports (driver event and data monitoring). It shows streaming driver events and no data activity. The Tablet Manager software was also set to COM2 and it?s status page also shows no data transmission activity.
At this point I am stymied but, per your recommendation, I did get an old PC laptop that has a real DE9 COM port. I need to pick up an adapter cable before I can test the tablet in this manner with the Summagraphics MGIIItest software.
The Keyspan doc doesn?t show any way to set comm params such as baud, parity, stop bits, etc. on the computer end. Since the tablet uses two stop bits would this screw up flow if the Keyspan is operating on a one stop bit default???
Sorry for the data dump but I wanted to give you all my observations in hope that you may have some further insights to assist me. Thanks to all!
Bill Curry
--
Message posted via ElectronicsKB.com
http://www.electronicskb.com/Uwe/Forums.aspx/repair/200805/1
On Mon, 19 May 2008 01:56:09 GMT, "powerseed via ElectronicsKB.com" put finger to keyboard and composed:
No. There is no polarity.
Pin 2 of the 9-pin connector is an input (Rx) to the PC, pin 3 is an output (Tx).
That's the voltage (floating) which one would expect on an input pin.
That's what you would expect on an output pin.
So it appears that your Tx and Rx pins are correctly wired. Your tablet is configured as a DCE (data communication equipment), not a DTE (data terminal equipment).
You need to measure the pins when the puck is off the surface, and then again when the puck is in proximity.
Just short pins 2 and 3 (with a screwdriver?) and access the port using a terminal program such as HyperTerminal. Disable local echo and then type any key. If the port is working you should see your typing being echoed on your screen.
Maybe the RTS/CTS/DTR pins are in the wrong state. You should measure them with your multimeter. An RS232 breakout box might be handy.
Probably.
I wrote the following COM port testing utility to help me test the various RS232C signals:
formatting link
It should also help you determine how your cable is wired, ie which pins are looped back to others. The program runs under DOS and only works with real COM ports.
- Franc Zabkar
--
Please remove one \'i\' from my address when replying by email.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.