Dead MS2 Daughterboard?

[1herrick1]

Due to an unexpected and hasty job change forced by the economy, my MegaSquirt project had to be quickly packed up and mothballed 2 years ago shortly after I got it built. It's an MS2 daughterboard on a V3.0 mainboard running V2.88 code. I also have a relay board and a JimStim. At that time, it checked out and worked great and I had just started feeling my way through the tuning software and the JimStim. Now I'm trying to get back into the project so I can get it installed this summer, but I seem to have ran into some trouble.

When I power it up, the 3 onboard LED's on the MS and the FP and injector LED's on the JimStim are all lit at about 50% power all the time and it doesn't seem to be doing anything. Neither MegaTune nor TunerStudio want anything to do with it, either. They won't even acknowledge that it's plugged in. I pulled the daughterboard out and checked the 5V supply and grounds at the various socket pins called out in the manual and they all check out. I also jumpered pins 12 and 13 and did a loopback test and my cable, port, and all the serial components on the board are functioning exactly as they should. And, as one would expect, none of the LED's are lit at all with the daughterboard out. This all tells me that the main board is OK. I thought that maybe reflashing the firmware would straighten it out, but it won't have any part of that. I get a message about no processor found or something like that from the downloader program. It's not completely dead, though. If I use the "Read Signature" function of the downloader program, the 3 onboard LED's on the main board and the FP LED on the JimStim will briefly go on 100% and the injector LED's will go completely out before they all return to their 50% on state. Sometimes when I "Read Signature" it will report "MSII V2.88 Embedded Code", sometimes it will only report some fragment of that message, sometimes it returns a few characters of jibberish, and sometimes I get the message about no processor found. Sick, but not completely dead.

Oddly enough, my JimStim isn't working right anymore, either. I think the wheel mode processor on it is hosed, also. When I tested the tach signal per the JimStim build instructions, all I get are a few random flashes of the LED. As much as I hate to, I must admit that in my haste to pack everything up during the job change, neither component got properly stored in a static bag like they should have been although they were separate in different places and not plugged together. I'm leaning towards static electricity damage to both the JimStim processor and the MSII daughterboard right now. Everything was stored in a dry, climate controlled building with no carpet, but I can't think of any other cause for these issues. Any other thoughts or insight before I suck it up and drop a C-note on a new MS2 daughterboard?

[trakkies]

It's very unlikely there is static damage on assembled components. The risk is small even when handling semiconductors during assembly, unless all your clothes and everything around you is made of artificial fibre. ;-)

[msiddalingaiah]

Oddly enough, my JimStim isn't working right anymore, either.

This is your clue. Even if you blew up the daughtercard, it's very unlikely you blew up the JimStim as well. How are you powering everything? Are you sure your power supply is working properly?

Try to get the JimStim working by itself first. A flashing LED would be enough.

[1herrick1]

It's very unlikely there is static damage on assembled components. The risk is small even when handling semiconductors during assembly, unless all your clothes and everything around you is made of artificial fibre. ;-)

I completely agree. I'm an electrical tech by trade and I deal with boards, components, and IC's of all types on a daily basis. I've never had a static-related failure that I can remember. I just can't think of anything else that would be causing these problems. I'm absolutely positive that neither the MS board/processor nor the JimStim have been wet or subject to any mechanical damage or other harm. The MSII daughtercard was stored plugged into it's socket on the main board. The main board and processor assembly spent the last 2 years laying on a piece of pink static resistant bubblewrap in the bottom drawer of a metal rolling filing cabinet and the JimStim was in a cardboard box in a drawer of a stationary metal filing cabinet. Not ideal conditions, but not like it was wrapped in a wool blanket and drug around on the carpet, either.

This is your clue. Even if you blew up the daughtercard, it's very unlikely you blew up the JimStim as well. How are you powering everything? Are you sure your power supply is working properly?

Try to get the JimStim working by itself first. A flashing LED would be enough

Once again, I agree. Very strange that they both quit. When I originally built it, I powered it with a 9V wall wart power supply plugged into the JimStim jack. It worked fine then. When I started messing with it again the other night, I started out using the exact same supply. I immediately suspected the power supply when it wouldn't work right, so I switched to a regulated benchtop power supply. I checked the voltage at both the JimStim and MS board and they were around 11.5 volts. The JimStim manual instructions for testing the processor and tach output are as follows:

Testing of the tach signal can be done using the LEDs. This test will also verify that the Jimstim CPU is working correctly. To do this you need to do this:.

Power off the JimStim and disconnect it from the MegaSquirt
Remove the pull up jumpers
Remove the primary tach jumper
Set the wheel mode to distributor mode (all DIP switches off)
Connect a wire jumper from the primary tach header (where the jumper usually goes) to an IAC led. Use the primary tach header pin closest to the pots (away from the PRIMARY TACH text) and the last pin of the 10-pin header (pin closest to the pots)
Connect a wire jumper from the other side of the IAC led to 12V. Use the next to last pin on the 10-pin header (just besides the pin connected in the previous step) and one of the pins on the 3-pin header besides the voltage regulator.
Turn the RPM pots all the way counterclockwise.
Once the connections above are done, power up the JimStim (with the MegaSquirt still not connected) and start slowly turning the RPM FINE potentiometer clockwise. You should see the led start flashing and it will flash faster as you turn the pot more.
Warning: With the JimStim not connected to the MegaSquirt, the bottom of the board will rest on a surface with all the exposed leads and solder pads. Make absolutely sure that the surface is non-conducting and that there is nothing that could make a short also on the surface. A short may cause permanent damage to the CPU and possibly other components..

If the LED does not flash, there is a problem with the JimStim CPU or the board. Go to the next section for troubleshooting. If the LED does flash, the board and CPU are fine. Remove the wire jumpers and put back the needed jumpers before connecting to the MegaSquirt..

It won't work with or without the MS plugged in to it. I've tested the LED and it's not the problem. All I get is an odd random flash now and then when the FINE tach pot is turned all the way down. If you turn it up at all, it doesn't do anything at all.

[Matt Cramer]

Could you please post a screen shot of what exactly happens when you try loading firmware? This will give me a bit more to go on about what the problem might be.

[1herrick1]

Here's what I get when I try to load firmware:

Reading C:/Documents and Settings/Joe/Desktop/Monitor_v2.891.abs.s19:
Detected MS-II (HCS12) code.
Entry point: 0x0000
Code range : 0x4000-0x3db420
File read successfully.

Opened comm port 7 at 115200.
Erasing main FLASH:
ERROR: Unknown error status response, e=0x00
Elapsed time: 0.36 seconds

Here's 3 samples of what I get when I use the "Read Signature" function:

Opened comm port 7 at 115200.
Reading signature:
Signature: "MSI"
MegaSquirt Download 2.00

Opened comm port 7 at 115200.
Reading signature:
Signature: "** V2.8ø"
MegaSquirt Download 2.00

Opened comm port 7 at 115200.
Reading signature:
No signature available, possibly could not connect properly.
MegaSquirt Download 2.00

I never get the same thing back with "Read Signature" twice.

[1herrick1]

One last note - I had serious doubts that anything on the main board itself was causing the problem, but just to rule it out, I plugged the MSII daughterboard into a bread board and supplied an external 5V to the necessary pins from a regulated power supply. The only jumpers to the main board were for Tx and Rx. It still does the exact same thing. I've been through the JimStim board from front to back checking voltages and looking for shorts, etc. at the processor socket and there's nothing wrong with it either. Unless Matt or somebody has another idea, I don't think I have any choice other than to condemn the MSII daughterboard and the JimStim processor. I still don't understand what could have happened that would brick both of them, but that's sure what it looks like.

[Matt Cramer]

Here's what I get when I try to load firmware:

Reading C:/Documents and Settings/Joe/Desktop/Monitor_v2.891.abs.s19:
Detected MS-II (HCS12) code.
Entry point: 0x0000
Code range : 0x4000-0x3db420
File read successfully.

Opened comm port 7 at 115200.
Erasing main FLASH:
ERROR: Unknown error status response, e=0x00
Elapsed time: 0.36 seconds

Here's 3 samples of what I get when I use the "Read Signature" function:

Opened comm port 7 at 115200.
Reading signature:
Signature: "MSI"
MegaSquirt Download 2.00

Opened comm port 7 at 115200.
Reading signature:
Signature: "** V2.8ø"
MegaSquirt Download 2.00

Opened comm port 7 at 115200.
Reading signature:
No signature available, possibly could not connect properly.
MegaSquirt Download 2.00

I never get the same thing back with "Read Signature" twice.

This may be a serial communication glitch. Can you try loading the firmware with a different computer - maybe a desktop with a real serial port?

[1herrick1]

I've got another computer I can try tonight, but the laptop I'm using is the same one I used before without a problem. It does have a real serial port and I've tried using both the onboard port and a USB/serial converter. Exact same thing either way.

On another note, I DID manage to get the JimStim working! As much as I hate to admit it, I hadn't bothered to clean the flux off yet after I built it as I was anxious to start testing and it didn't seem to bother it then. Then it got put in a box for a couple years and completely forgotten. Um, oops. I had nothing else left to try, so I scrubbed it up with some acetone, rinsed it with hot water, and reinstalled the processor after the board dried. Works great now! That is certainly NOT the problem with the MegaSquirt as I scrubbed and rinsed it immediately after it was completed, but at least one of them is functional again.

[msiddalingaiah]

I DID manage to get the JimStim working! As much as I hate to admit it, I hadn't bothered to clean the flux off yet after I built it as I was anxious to start testing and it didn't seem to bother it then. Then it got put in a box for a couple years and completely forgotten. Um, oops. I had nothing else left to try, so I scrubbed it up with some acetone, rinsed it with hot water, and reinstalled the processor after the board dried. Works great now!

That's interesting. What kind of solder/flux were you using? This is a V3.0 board right?

Here's something else to check: measure the voltage on pins 2 and pin 6 of U6 with respect to ground. Pin 2 should be something like +10V and pin 6 should be around -10V. If that's not there, you won't have any RS232 communication. Also make sure pin 16 has +5V and pin 15 is zero volts (good connection to ground). Post your measurements.

[Matt Cramer]

If you remove the MS2 daughter card and jumper pins 12 and 13 of the 40 pin socket, will the MS2 main board pass a loopback test? Try this several times and see if it sometimes loses characters or echoes back the wrong thing.

[1herrick1]

Thanks for all the help and suggestions. They are much appreciated. I finally found the problem and all the components are OK after all. It wasn't the solder flux on the JimStim. I had switched over from the wall wart to my regulated DC power supply immediately after cleaning the flux off the board because I had gotten it out to power the MSII daughtercard when I was testing it on the breadboard and it was handy. It was at that time that the JimStim started working correctly and the rest of the MegaSquirt began working right also when I tried it again. I never really made the connection that it was the wall wart until I attempted to use it again last night. Everything started freaking out again. I hooked the wall wart up to the scope and it looks like part of the bridge has failed or something because the output is a square wave instead of straight DC. It worked fine when I built the thing but it sure doesn't now. Oh, well. It's in the trash, MegaSquirt is happy again, and I gots no worries. Thanks again!