Help needed - Bluetooth DRO project

[vtsteam]

Help, help! 

I've built the homebrew make-your-own-comparator-board version of the Yuriy's bluetooth TouchDRO.

http://www.yuriystoys.com/p/android-dro.html

and

http://www.yuriystoys.com/p/msp430-launchpad-mixed-scale-dro.html

Hoping to use a combination of digital calipers and an "Igaging" type scale I had on hand.

Everything went well (except for what I don't realize didn't) during construction -- had a blast making the stripboard circuit using a drill bit -- cool process! BUT.......... It didn't actually work with either my scale, or my calipers. 

So to troubleshoot, I removed the homemade board and reprogrammed the TI board for "Basic Igaging" operation. Then my scale worked fine. Switching back to the "Mixed scale" version and attaching the board repeated the problem. They say doing the same exact thing twice and expecting a different result is a sign of insanity. Well I've done this at least ten times, so draw your own conclusions!

To me, the problem can only be in my homemade board, then, or a difference in the firmware, or since different pins on the TI board are used as clock and data connections (for Basic vs Mixed operation), maybe the TI board is partially damaged or defective. Problem is......I have no idea which is at fault. Or where to start.

Unfortunately there are no schematics available. And no clear list of what igaging scales might work with the Basic version and NOT with the Mixed scale version of the firmware. And no self test program for the TI board in hex or txt format that I can load to see if the input pins P2.0-P2.2 are functioning.

Okay I'm going to stop explaining and list what I know here. Any help would be GREATLY appreciated: 

Details:

Voltage Shifter Board:     I did the bare strip board build.
Scales:                               I have an aluminum Digi-mag Igaging type Shahe scale, and a couple of Chinese calipers
TI board:                            EXP430G2 w/ M430G2553 chip
Programmer:                     FET-Pro-430 Software Pkg. - LITE
Firmware:                          DigitalReadout_Universal_V1.0.hex,


Symptoms:

If I flash the TI board for Digital Readout Basic Igaging support firmware - (without using the voltage shifter board) and connect the Shahe scale clock line to P 1.7 and the data line to any of 2.3,2.4, 2.5  the bluetooth tablet reads each perfectly.

If I flash the TI board for Digital Readout Universal, and connect the voltage shifter board with the same Shahe scale (data line to P2.5 and clock line to P2.0), nothing changes on the bluetooth tablet when I move the scale.

If I connect a caliper (clock to P2.1, and data line to p 2.4), the same thing happens. No change on the Bluetooth tablet when moved. (However, the caliper's own readout shows correct measurement moves, and that it is properly powered.)

Checked so far:

1.) All wiring has been checked several times. But we know this is probably the fault, anyway....right?

2.) Voltage to both scales was checked while connected. Shahe is getting 3.5V and caliper is getting 1.5V. This was checked to their individual grounds. Caliper is powered up and shows readings

3.) LM339 pin 3 =3.5v measured against pin 12 (ground). Check.

4.) LM339 pin 6 = 1.7v measured against ground. should be half Vcc. Check.

5.) I replaced the upper LM339. No difference

6.) I have a second identical TI EXP430G2 board. I've substituted that in, and get exactly the same results as the first board. But maybe I wrecked the inputs on both?

7.) Running Android BlueTerm app shows good connection and continuous output of x0,y0,z0,w0 to the tablet. Moving the scales makes no change.

I have tried the 1.0 version of the firmware, as well as the 0.3 version. And also both the .txt, and .hex versions.

I can return to Basic igaging mode firmware and the Shahe scale again works fine. Can't test the calipers in this mode. However their readouts still work fine.

Any suggestions? This is really driving me nuts.

[vtsteam]

The boards:






Notes:

The top connector, left side of board is the Igaging scale input. In the photo the clock and data line (white and green respectively) are reversed. I've tried it both ways. Unfortunately the web instructions don't specify. Power and ground (red and black) are correct (3.5V for this scale).

Next connector down is the caliper input. That has wires in correct order in this photo. And power to that scale is 1.5V.

Third connector (not connected) also has a 1.5V power pin -- I was going to use another caliper type scale.

I'd like to concentrate in troubleshooting the topmost circuit -- if it's at fault. It is difficult to do this without a schematic. I do know that the topmost LM339 should have half the connector voltage on pin 6 -- and it does (1.7V). And that pin 3 should have 3.5V -- it does.

[vtsteam]

And here are some photos from Yuriy's website (URL in first post per CC license) of how things should look for comparison, in case someone has a sharper pair of eyes than I do (pretty likely):

The top photo shows the main connections, minus the power connections that bring the different voltages needed by various brands and types of scales.

The second photo shows the Power rails (stripes at top), and the "virtual grounds" (stripes mid card). also marked are the locations on the LM339 comparator for virtual ground connections - numbered by scale#.

The next photo shows the power connections set up for a different set of scales than mine (Yuriy's own mix of glass scales, igaging and caliper).

These should be jumpered, per your own mix of scales, to the appropriate pins on the scale connectors. In my case the top scale connector is an Igaging scale and needs 3.5V.

In my case, I need 1.5V on the second scale (Connector # 2) for a caliper scale.

Also the "virtual grounds" should be jumpered to the LM339, again, per your own mix of scales.

If you refer back to my board photo, in the post above this one,  I need 3.5 V on the top scale connector (Igaging scale), so there is a power jumper to the second row on my board (uppermost left, short, red).

Since I also need 1.5V  on the second connector (caliper scale), there is a power jumper to the third row on my board (upper left medium length, violet)

For virtual ground jumpers I have:

Scale #1 a jumper from pin 6 of the LM339 to the "3.5V/2" "virtual ground" - Blue jumper

Scale #2 a jumper from pin 8 of the LM339 to the "1.5V/2" "virtual ground" - Green jumper

 

[vtsteam]

View of the board "traces" and population (from Yuriy's website):

[awemawson]

Steve I can contribute little on the Iguaging or PIC part of your project - never been there, however when I was a callow youth back in the mid/late 1960's I was a Technical Assistant in the Mullard IR devices Application Lab (Multi-element liquid nitrogen cooled arrays for satellite surveillance) . One of my duties was to give physical reality to the circuit designs dreamt up by the clever senior engineers. Usually ultra low noise signal amplifiers.

Most first prototypes were on Vero 'strip board' and like you I used a drill to form breaks in the track. The number of times odd behaviour was eventually tracked down to a hair or whisker of copper still connecting tracks was legion.

I've looked at your pictures and nothing leaps out at me, but I would advise checking each and every break with an ohm meter

[vtsteam]

Hi Andrew, thanks. I actually anticipated that problem, being my first attempt with this type of board, and checked each cut with an ohmeter before I populated anything.

But I can also say SOLVED, woohoo! 

Yuriy pointed out that with the older "Igaging" style scale I was using (Digimag type) the scale expected a clock signal from the remote reader. So an additional wire was needed to the scale's clock line from P1.7 on the TI board. Unfortunately this wasn't shown on the website. But it definitely cured the problem with this scale! So I'm happy.

Well, I'll be even happier if I can debug using some old calipers as scales with this unit as well. But for now the biggest problem has been solved.

[awemawson]

Excellent - glad you found the issue 

I still have a Veroboard 'stripboard cutter' somewhere which essentially is a coarse fluted drill moulded into a blue plastic handle

[seadog]

Me too, and I think I could find it with a few minutes' notice.

[Will_D]

Me too have a stripboard cutter from Vero!!

[vtsteam]

I still have a wire wrapping tool anfd  roll of Kynar wire but didn't know about strip board tools.

A lot less messy than wire wrap. And a lot easier to troubleshoot.

Of course, re troubleshooting, I have tried two different calipers with this DRO project, and still no joy there.

I hope the problem isn't another undocumented wiring requirement