SIDx Not Available (Can Bus Errors) - after wiring MB-6582 CS

[Filch]

I've finally finished getting my CS surface wired up to the main board. When I turned it on the first time, I got an error stating 'SID4 not available' and '(No MBNet Response)'.  I found another thread with the same issue and after reflashing the MIOS, the problem cleared.

I've reflashed the MIOS to 1.9g on all cores. I get READY on all 4 cores. I then loaded the latest MB-6582 SID OS and now I'm getting the following error.

SID4 not available (Can Bus Errors).

R80 is in place, all the jumpers are in place. I wasn't getting any errors until I connected the control surface. Any suggestions of where I should begin looking to resolve this issue?

[Filch]

I reupped the SID OS to all 4 cores again, and now I'm back to SID4 Not Available (No MBNet Response)

[nILS]

[09:32] <nILS> your post doesn't specifically say that, soo:

[09:32] <nILS> did you also upload the mbSID app to all four pics again?

[09:33] <Filch> yah, I did MIOS and Sidapp to each

[09:34] <Filch> wait, I didn't upload the SID app to all 4 pics individually, just the MIOS

[09:35] <nILS> ;)

[09:35] <nILS> well there you go.

66% of the problem was fixed by actually uploading the mbSID app to the 3 slave cores.

Now only core 4 doesn't respond.

Also the first row and first column of the LED matrix is flashing after boot (maybe a hint to Mr. Nice Guy)

[Wilba]

OK... short lesson first:

PICs already labelled with their device ID on them (i.e. with a sticker)... so they're ID=0, ID=1, ID=2, ID=3

They go in Core #1,#2,#3,#4 going right to left (#1 is the master and goes in far right).

SID engines are identified by the device ID, not the Core that it is in. So PIC with ID=3 is "SID4" even if it's in Core #3 or #2.

Test #1:

put PIC ID=2 in Core #4

leave out PIC ID=3.

When running MB-SID, SID3 should be accessible. SID4 should not be (the PIC is out).

Test #2:

put PIC ID=3 in Core #3

leave out PIC ID=2.

When running MB-SID, SID4 should be accessible. SID3 should not be (the PIC is out).

Based on these tests, you'll know whether the problem is with PIC ID=3, or with Core #4.

I am guessing though that since you've already uploaded MIOS and the SID app to PIC ID=3, the problem is with Core #4. Check the soldering of the diode under the PIC's IC socket, and where it connects to the PIC (pins 35 and 36, see PCB diagram for help)

[Filch]

Spoke briefly to Wilba about this problem in the chat room. I was suspect of my CS causing the problem since the mainboard was working okay before I wired up the CS. He suggested removing the Shift Registers and see if the problem goes away.

I removed them all and my error did disappear.  I replaced them one by one to see where the problem began. I started with the 165's and all was well until I placed in the 5th 165 into U20.  I thought I may have a short or bald solder point there. I reflowed all the points and came back with the same problem. To make sure the shift register wasn't bad, I removed it and put the one from U19 into U20 and I had no errors.  I thought I may have found a bad IC, but to be thorough, I put that suspect IC into U19,  removed the IC I had in U20,  and no errors again.

The only time I get this error is when all 5 165 shift registers are in place. The problem goes away when I remove any one of the 165's, it doesn't matter which one. Removing one or all of the 595's does not change anything.

[Wilba]

How exactly do you determine if the error disappears when you have ICs taken out?

i.e. do you see something on the LCD? can you drive SID4 engine?

Can you confirm the CS is responding normally? i.e. all buttons work, especially the SID 1,2,3,4 buttons?

Run the tests I suggested in the previous post.

I am now thinking it might even be possible that both test #1 and #2 will pass (i.e. PIC ID=3 is good, Core #4 is good).

FWIW the 5th 74HC165 handles the switches, the other four handle the encoders. Taking out the 5th one (or one earlier in the chain) is effectively disabling button inputs... which doesn't seem to be related electronically with the CAN bus and some break in the comms between PICs, especially if you're reporting the error disappeared by taking out U20. Bizarre.

Keep at it, it's mostly working... and you might win an award for most bizarre MB-6582 fault. (Maybe I send you a free lazer!)

[Wilba]

Just out of interest, try detaching the LCD from the CS PCB, keep it plugged in, just not mounted and away from any PCB.

A short between the LCD and the CS might be the cause. Maybe. Probably not, but quite a few people forget to insulate between LCD and CS PCB.

[Filch]

I was using the LCD to determine if I was getting the error still or not.  The CS was not responding normally. Most of it was, other parts were acting strange. When I did your test, I was getting the same errors on the LCD.

So, I detached the LCD from the CS and the problems went away!  :D  Looks like I've got a short somewhere. I already had the pads covered in electrical tape. I'll just wrap the rest of the exposed LCD PCB in electrical tape. Thanks Wilba!!!

I'm going to have to remove the front panel to get the LCD attached again. I was trying to avoid taking these apart as I fear I'm going to squish some LEDs when putting the front panel back on.  My panel fitting is a bit snug. I have to nudge a button here and there to get it to slide on properly.

[nILS]

edit:

Tags: MB-6582 [X] lcd [X] Control Surface [X] CAN [X] nILS the Tag Nazi [X]  [Add Tag]

Ja ja, I'm a tag nazi... just trying to clean up the tags so they become somewhat useful. A huge collection of randomness is not very helpful.

[Wilba]

It's good to hear you got it going... still not exactly sure why a short there causes those particular errors...

FWIW I don't use electrical tape, I cut some strips of card (like business card thickness) to fit the area above and below the LCD bezel, with holes at the ends so the screws go through.

[fussylizard]

FWIW I didn't use any insulation, tape, etc.  I just used a small nut on each screw between the PCB and the LCD to offset the LCD below the board far enough so it won't short: http://picasaweb.google.com/fussylizard/MB6582?authkey=Gv1sRgCM7npqbPgvbfYA&feat=directlink#5359892899542622770

Glad you got it working though!

[Filch]

FWIW I didn't use any insulation, tape, etc.  I just used a small nut on each screw between the PCB and the LCD to offset the LCD below the board far enough so it won't short: http://picasaweb.google.com/fussylizard/MB6582?authkey=Gv1sRgCM7npqbPgvbfYA&feat=directlink#5359892899542622770

Glad you got it working though!

I actually noticed your post for that when I searched for LCD Insulation. I think I'll do that as well as taping up the PCB (which I've already done).  My screws are a little small though, so I may stack 2 washers instead for a gap.

My problem now is that with all this futzing around troubleshooting the issue, I've broke a few wire leads on some of the ribbon cables between the CS and the mainboard....  *sigh*

[fussylizard]

I actually noticed your post for that when I searched for LCD Insulation. I think I'll do that as well as taping up the PCB (which I've already done).  My screws are a little small though, so I may stack 2 washers instead for a gap.

My problem now is that with all this futzing around troubleshooting the issue, I've broke a few wire leads on some of the ribbon cables between the CS and the mainboard....  *sigh*

Well then you probably noted my solution there also: use right-angle SIL headers on the top of the CS, and use nice, flexible ribbon cable soldered to the bottom of the main PCB that ends in the mating SIL connectors.  Worked great for me, and I opened and closed that thing a zillion times.  I would have tried Wilba's solution, but (1) I thought a detachable solution might come in handy (it did for me), and (2) I couldn't figure out the magic incantation to get cables like Wilba's out of Mouser, so I just used what I could find. :-)

Good luck sorting it all out- you're almost there!  I love my little MB-6582...