Problem

[rm357]

Add on to my previous message - if you do start probing around with a multimeter, please be very careful. There is live 120VAC on that end of the board!!

RM

[LondoB5]

It looks to me like your 485 chip is in backwards. Pin 1 the one with the circle, should be facing the top of the board.

EDIT - Nevermind. Day late and a dollar short. Looks like everyone else caught that too.

[mnarel]

I touched up the solder joints for the 15K resistors and the 4700uf.  I also tried the 'hold the reset' without effect.  I examined the reset (and other buttons closely, and they are all in the right way.   My daughter was looking at it next to my functioning board and noted that the center resistor next to the 3.3 and 5v lights do not match between the boards.  My other boards all have a wide brown, thin blue, thin brown, thin gold stripe.  The problem board has a Wide brown, wide grey, wide brown, wide gold pattern.   All the other resistors on the board appear correct.  Is there a chance I have the wrong resistor for this application or is it just a matter of different manufacturers?

[mnarel]

At a more complete look... all the 180 resistors are the new pattern on board 12, so I believe it's just a manufacturing difference (unless they are all wrong!?!).     I looked as closely as I could for any 'bridging' inside the sockets and didn't really see anything amiss, though I'm not sure what exactly I would find if it were 'bridged' in there.

I am also in joyous receipt of 4 more Lynx express kits today compliments of the J team in FL (Thanks guys), so I now have spare parts galore if it's time to start pulling and replacing things on board 12.   I'm not (yet) an electronics guru, so testing via DVOM etc is beyond my current knowledge level... and I wouldn't know what I was looking for anyway...   so I ask:

Is it time to start pulling and replacing components?  If so, what's the best order to move in?    I don't mind buying a few replacement parts if necessary, I just am really reluctant to scrap the whole board.   Given that the chips all work in board 11,  I have to believe it's either my error, or a minor component failure.


Thanks also to everyone who's offered suggestions so far.   I didn't mean to tie up so much collective brainpower on my piddly problem.   I'm at a loss here and I really do appreciate all the help.

[KeithTarpley]

Greetings,,,

Can you double check, in case I missed it, and see if you have triacs in all 16 locations, and not a switched regulator in one of those locations, just in case?

Keith

[mnarel]

I checked and all 16 are triacs.  I aslo pulled both heat sinks and physically checked the regulators to verify that they were in the right slots, and that I didn't get accidentally shipped 17 triacs or another kind of regulator -  LD33CV in the lower and the 78s05 in the upper.

[CaptKirk]

Did you ever go back and reflow the solder connections on the PIC socket?  There is nothing obvious that I see from the pictures but the board certainly is acting brain dead and you KNOW the PIC is ok from swaping.

[mnarel]

I did reflow the connections at the PIC and at the numeral control chip.    Since we know the chips and numerals themselves are functional, that seems to leave the sockets themselves, or the power supply to either the PIC or the Numeral chip.   I've been trying to trace the board (and in way over my head here) but it seems to me that the power to the PIC is 3.3 V coming in via the reset button, and that power to the numeral chip is 5V coming via the R485 chip next to the DMX in.  5.5V is then supplied as required to the LED screens via the numeral chip.

We know via the working leds that we have good voltage at both the 3.3V and 5V origins (do we?) so the problem ought to lie between the LED lights after the regulators and the two chips.    Does this make sense or am I stabbing wildy in the dark?   Everything from the AC input to the regulators should be out of the diagonstic loop, and everything from the PIC downstream through the channel LEDs to the output sockets should also be irrelivant at this point.  

I confess that trying to follow the board through the PIC and numeral chip is losing me so I can't tell which of the resistors at the top of the board are in the pattern or which of the 1ufs or the 2700pf could be factors as well.   Am  I wrong at this point in excluding the power tranformation/regulation, and output streams from consideration?

[mnarel]

I didn't replace the fuses, but removing the left fuse from the board results in no power at all.  With the fuse in, I get good LED's at the 3.3 and 5 V regulators.  I think this rules that out.

 Could a bad fuse, or more likely a bad or partial solder joint at the fuse holder causing irregular AC result in a constant green LED at the regulator but still  create a problem further downstream?

[LondoB5]

From the way it looks, I doubt it.

Just so we're all on the same page, and just to make sure of the symptoms:
Can you redescribe what happens when you power up the board (what lights up, etc)?

Also, can you list all of the steps you have taken thus far (since programming the pic)?

[rm357]

Resistors - The width of the stripe doesn't matter. If I remember my color codes correctly Brown-Blue-Brown is 160 ohms while Brown-Grey-Brown is 180 ohms. That particular one is used to limit the current to the LED, so 20 ohms is not going to make or break the system. Gold is the tolerance, I think its 5%.

RM

[LondoB5]

According to the Express wiki page, if you have 3.3v and 5v lights on before installing the chips, everything with the board is good.
It sounds like you've replaced all of the chips with chips from another express. You don't have a multimeter so we can't read voltages or currents across resistors or caps.

Could it be a short across pins on one of the chip sockets on the top side of the board (rather than the underside, which is easier to see)?
Reflowing would also make a short like this worse.

[mnarel]

Here's the board's behavior to date:

When I was building the board I conducted the recommended power up test prior to installing the chips and final wiring.  

When I plugged it in at that time, both the 3.3V and 5V LEDs came on and stayed lit.  

After installing the chips and wiring I then plugged the board back in.  The 3.3V and 5V LEDs come on and stay on.  The LED from Channel 5 also came on, and there was no activity on the LED numerals.   I had experienced this problem on a prior build and knew that the PIC needed reflashing (that solved the problem in the earlier instance).  

I pulled the PIC and reflashed it and another PIC using a Pickit 2 and the current firmware.  

The PIC I placed back in my board 11 worked fine after the reflash.  

I placed the flashed PIC in board 12 and plugged it in.  The 3.3V and 5V lights came on, no other LED's illuminated, but there is no activity at all on the LED numeral readouts (other than a very very faint light in the far left numeral screens decimal point).  

I have since replaced every chip on the board with chips from a working board and continue to get the same results.  The chips I removed from the problem board I installed in the working board and they function fine (so the chips are good, and the problem board is not frying anything).

I have also revisited the solder joints on the entire board a couple of times without luck.  

I hooked up the problem board to my laptop and ran a channel check accross all 512 channels of DMX with no output from any of the 16 sockets on the board so there it appears to be more than just the display aspect.

[mnarel]

I beleive I do have a multimeter lying around (I was an amatuer electrician in a former life... just have never applied it in instances less than 120V AC or wiring less than 14 gauge).  If someone can let me know where to stab and what equals a pass or fail reading, I can start working that.

There are no obvious front side shorts on any of the resistors but I can't really see the socket pins on the front end.  Is there a way to check this effectively without removing the sockets and redoing?

[CaptKirk]

Do you have a friend who knows how to use a multimeter/DVOM?  The actual power levels could be at play like a bad 3.3V (it has happened before).  You should get 3.3V at the PIC on the pin you traced from the regulator.