Serial comm testing - cable ok, jibberish on screen

Spittybug · Post by **Spittybug** » Sun Jun 27, 2010 9:51 am

I've tested the cable loopback, per step 26e and it is fine; I get character echo. Unjumper and the echo goes away. Perfect.
However, when I jumper pin 12 to 13 and connect stim, I get pages of jibberish on my screen. Hitting a key doesnt echo back. Sometimes gives my computer blue screen of death. I can't find any bad solder connections (I'm pretty good at them) and component orientations etc. have been checked.

It would appear via the loopback test that I'm fine with the cable (serial to USB) right? Any suggestions???? Thanks.

Spittybug · Post by **Spittybug** » Sun Jun 27, 2010 9:59 am

Not sure what I did, but it's working fine now. I dialed down the pots on the stim and changed the switch to Wide Band. No gibberish on screen and the echo back is functional. Looks like user errror.

Spittybug · Post by **Spittybug** » Sun Jun 27, 2010 12:44 pm

I spoke too soon. I went onto the next step, loading the .S19 file onto the board and it gives me an error. The file reads, but doesn't install. Then it all hangs and I must reboot the computer. I went back to testing the connections again (the last two posts) and I'm back to just jibberish. The second I connect the stim I get screenfulls of characters rolling by. No echo back, but I did that one time as mentioned in the previous post. I don't want to keep building this thing until I know what is going on!! Hopefully someone can help me.
Thanks.

Spittybug · Post by **Spittybug** » Sun Jun 27, 2010 3:28 pm

Wow... screwy. 3.0 board, sorry forgot to mention that. MSII.
Just went through the trouble shooting section and checking DB9 Pin#5 is supposed to have continuity to MAX232 #15. IT does, but also with #s 1,2,4,9,10,13,14 and 16! Is this normal?
I confirmed ~5V on pin #16 and ground on pin #15.
I also checked continuity between individual holes in the DB9 females and the respective back side of the board solder connections. All good.

Matt Cramer · Post by **Matt Cramer** » Mon Jun 28, 2010 9:19 am

Continuity with pin 2 is definitely not normal. Try cleaning the board thoroughly and make sure there are no stray solder blobs.

Spittybug · Post by **Spittybug** » Mon Jun 28, 2010 10:16 am

Matt, thanks for the reply. Is continuity with all of the other pins OK, or are you thinking that they are caused by continuity with #2? I've definite got something not right, I had a bear of a time getting the firmware loaded. I think I may have a faulty 9 pin as odd as that may sound. I've checked the soldering 5 times, all very solid and clean. However, when doing the echo tests I found that wiggling the serial plug a hair made jibberish stop. Either the pins of the cable are screwy or something inside the 9 pin female on the board. Each pin hole shows good connectivity to its motherboard joint. Since I'm borrowing this cable anyway, I'm going to buy one from DIY and see if that was the issue. I was able to load the firmware (it said success) after many tries.

Spittybug · Post by **Spittybug** » Tue Jun 29, 2010 7:12 am

Pins 1 and 9 on the DB9 are VCC pins. The real question is why you have continuity between pins that should be grounded and pins that should be VCC - you have a short between 5 volts and ground somewhere.

Quote Spittybug Re: Serial comm testing - cable ok, jibberish on screen
Sent: Mon Jun 28, 2010 8:03 pm
by Spittybug

Matt; I was able to surgically remove the DB9 from the board without any damage to the board (the DB9 is a little bit bent up however). Interesting results when I use the board holes to perform the continuity check with the Max232:

DB#:Max232# showing continuity
#1: 3,5,6,9,10,11,12,15,16
#2: 14
#3: 13
#4: none (not without the DB9 itself)
#5: 1,3,5,15 with "brief connectivity to" 6,9,10,11,12
#6: none (not without the DB9 itself)
#7: 8
#8: 7
#9: 3,5,6,9,10,11,12,15,16

I don't think this looks right! Bad Max232 chip?

I checked all my solder connections again under a magnifying glass and none are even questionably close to touching. I did the initial test of the board too, the three voltage regulator holes... all proved infinite. I've soldered enough chips & bits over my life to know how to do it quickly and cool, so I'm not inclined to believe that anything has been baked. I don't have the means to check piece by piece on this construct, so I've stopped building at step 62 until I hear what DIY suggests.

Spittybug · Post by **Spittybug** » Tue Jun 29, 2010 8:46 am

Any advice on how to find my short between 5V and ground that you believe I have? I can safely tell you that there are no stray solder blobs or any bridging of leads. I'll wipe the board again with alcohol to ensure no flux bridging, but other than that, is there a way to rule out a bad IC that may be internally shorting? I did test the board initially as instructed and there was NO continuity (between the voltage regulator holes).

Matt Cramer · Post by **Matt Cramer** » Tue Jun 29, 2010 9:58 am

To check for a short in the Megasquirt PCB, power up the Megasquirt on a Stimulator or on the car and check the following points for voltage with a multimeter.

You should find the same voltage as the battery voltage on the following points: S12, S12C, the center legs of Q9 and Q12, the left (non-banded) end of D3, the left leg of U5, and the left (banded) end of D9.

You should have 5 volts at the following points: S5, the two +5V holes in the proto area, the right (non-banded) end of D9, the right leg of U5, the left (banded) end of D19, and pins 1, 20, and 31 of the CPU.

Let me know where you find the correct voltages, and where you find the wrong ones.

Spittybug · Post by **Spittybug** » Tue Jun 29, 2010 6:02 pm

Matt;
I thoroughly cleaned backside with acetone and gently scraped any resin away with dental pick. Minor quantities. I have very clean and distinct gaps between nice cone solder joints.

Using your list provided above, I have battery voltage confirmed on ALL POINTS LISTED: S12, S12C, Q9 center, Q12 center, left D3, left U5 and left D9.

I have 5 volts on: +5V holes in proto, right D9, right U5, left D19, 1,20 and 31 of CPU. I DO NOT HAVE ANYTHING ON S5. Did I miss an instruction somewhere to jumper this??? I checked 3 times tonight but couldn't see anywhere where it told me to do anything with S5.

I again checked the MA232 to DB9 holes (plug is gone now).
#1 still is common to 3,5,6,9,10,11,12,15,16
#5 is common to 1,3,15,16
#9 is common to both the right leg of the voltage regulator and the center.
All others are as should be.

Hopefully this helps you diagnose this?? Thanks.
Owen.

Matt Cramer · Post by **Matt Cramer** » Wed Jun 30, 2010 4:09 am

Your continuity tester may be a bit too sensitive. How many ohms are there between the center leg of U5 and DB9 pin 9?

Spittybug · Post by **Spittybug** » Wed Jun 30, 2010 4:45 am

Matt Cramer wrote:Your continuity tester may be a bit too sensitive. How many ohms are there between the center leg of U5 and DB9 pin 9?

I was thinking the same... I don't have it in front of me, but I do recall noticing it was materially different.
What about the JS5? Is that supposed to be zero or did you mistype what I should be checking?

Thanks.

Matt Cramer · Post by **Matt Cramer** » Wed Jun 30, 2010 7:53 am

S5 - not JS5 - is connected to the 5 volt plane by traces in the PCB.

Spittybug · Post by **Spittybug** » Wed Jun 30, 2010 8:18 am

Matt Cramer wrote:S5 - not JS5 - is connected to the 5 volt plane by traces in the PCB.

My bad, sorry... where is S5? I don't see it on the v3 components.gif. Is it a labeled hole on the backside? Next to what? Thanks.

Matt Cramer · Post by **Matt Cramer** » Wed Jun 30, 2010 8:31 am

It's next to the DB9.

Spittybug · Post by **Spittybug** » Wed Jun 30, 2010 2:47 pm

S5, got it, it has 5 v on it.
I think I have good news when I look at actual resistance not just continuity...
DP9 #1 pin versus each MAX232 pin sequentially: 1,1.25,1,varies,1,1,1,1,1,1,1,1,1,1,1.32,0 with the meter on the 2K setting. So yes, it looks like I have a highly sensitive continuity checker!! Pin 4 doesn't seem to settle...it keeps incrementing. But, I think these look as they should, do they not?

Thanks, Owen.

Spittybug · Post by **Spittybug** » Wed Jun 30, 2010 5:43 pm

Sorry, forgot to answer your question...... ~870 ohms between #9 and center leg of U5.

Matt Cramer · Post by **Matt Cramer** » Thu Jul 01, 2010 4:50 am

Ok, sounds like the short isn't there, you just have an overly sensitive continuity checker. Does your board pass the loopback test?

Spittybug · Post by **Spittybug** » Thu Jul 01, 2010 4:21 pm

Matt, this thing is killing me.... despite all the positive diagnostics of the past couple of days that would indicate all is alright, tonight I failed again.........

Step 26 e. passed with flying colors... loopback of cable. Everything echos OK.
Step 26 h. connect the stim, jumer 12 to 13, connect to DB9............rolling screens of jibberish until it leads to a blue screen of death. No means to stop the jibberish, no loose cable connection, no bad solder joints, nothing.

What the F? is going on here!?!?!?!? This is getting really frustrating that the voltages and continuity look ok, but the data connection does this

I sincerely appreciate your help in figuring this out, and hope it's just me doing something stupid, but my gut is telling me that Max232 is somehow bad. I'm happy to send a hi-res photo of the backside of the board if you think it would help. Next step?

Owen.

Matt Cramer · Post by **Matt Cramer** » Fri Jul 02, 2010 3:51 am

My first guess would be a bad serial chip, too.