Go Down

Topic: Only 1/3 of 96x16 LED matrix works..can't figure out why? (Read 1 time) previous topic - next topic


If anyone reads the Sure 3216 topic, you might have seen my recent post about my custom version having issues.   If not, I'll sum it up.    I took the schematics from Sure  Electronics website for the 3216 boards, and basically triplicated it to form one single 96x16 display.   I  spent well over a week checking the design versus the schematic, to make sure I had everything right.    When I was confident I did, I sent the board out for production (which was not cheap).  After receiving the boards in, I started putting them together, section by section, testing EVERYTHING as I went along before final soldering.    I had discovered one issue where I had left out a trace to one of the rows, but this was fixed by a simple wire bridge soldered on.

All was good up to the point i got 8 of 24 LED matrix pieces soldered in (1 whole 3216 board).  When I started working on the second third of the board,  The Matrix pieces did not light up.   I spent a few hours testing, but couldnt figure it out, so I tried the 3rd.  The 3rd also does not work.  I'm at a loss as to what the issue could be, so here are a few of the things I've tried so far:

1.) Check clock in/WR in/DATA in and make sure they're connected to the CLK/WR/DATA out from previous "third" (basically removing the 8x2 connectors from the sure boards and wiring it direct

2.) check voltage to every +5v pin on both the HT1632C controllers and the shift registeres/hex inverters.  Everything had 5V

3.) Hook 3 Sure 3216 displays together, and use a continuity check to test various points (+5v, GND, WR, DATA. and CLK) on the sure boards to make sure the connections were correct, then test same spots on mine

4.)  First hook my arduino to a sure board, then the output of the sure board to my board.   Both the sure and my board (1/3) light up correctly, so input looks fine.

5.) check various pin combinations on the output of the shift registers and other ICs.

I don't have an oscilloscope or logic analyzer, although we have some at work I might be able to use to test.  My thinking is that either something is wrong with the shift register/hex inverters, or I possibly have bad HT1632C chips on the remainder of the board (unlikely).  The only other thing short of sure possibly not doing their schematics correctly, is that by using alternative parts (used a different vendors 74HC164D/74HCT04D chips instead of the base ones, but data sheets were nearly similar) the different inverters shift registers might be throwing it off.

I was able to do some basic tests of the 74HC164D chips to by measuring voltage between ground and the 4 output pins...when the displays were supposed to be showing stuff, the 4 outputs for Chip Select were all high, (3-4V), so as far as I can tell...things are working...except when I measure voltage between ground and the outputs of the drivers that would connect to the matrix, i get a steady 0.9-1.0V, where if I measure the same point on a working module, it fluctuates between 0.6 and 1.5V (0.6 off, 1.5V on to the LEDs)...as if the outputs are stuck or something.

A link to my project is in the signature line, or i can post video/screenshots if anyone prefers, I know it's not the code because if I hook up 1-3 sure displays they work fine.    I've got data sheets for the sure boards in my git repository.

If I can get a logic analyzer or o-scope, what are some things I can check to try and narrow down the problem?

I welcome and appreciate any possible ideas or assistance anyone might have.


The first thing we need to see is the schematic of what you have, I could not see it in that link you gave.


Did you try connecting to the 2nd and 3rd thirds individually, like if they were independent?


Mar 15, 2012, 12:09 am Last Edit: Mar 15, 2012, 12:12 am by evanrich Reason: 1
Attached are the schematic and board, sure's pdf wont fit so heres a link

(Click "raw" to download)

I haven't tried checking each display individually.   I've got to pick up some finer gauge wire than what I have now, but that's one of my goals (to test each section individually).  I've also been testing connections at the top of the pins themselves, to ruin out any possible solder issues (was the cause of a set of red LED's not lighting up in first section)

The likely spots, from what I'm seeing, would be the handoff between sections of data, WR, and CLK, however all connections are correct as tested between 2 sure boards connected together, and then checking the same spots on my board.  I need to check with our engineering department to see if they have a spare oscilloscope I can use to test for the presence of clock, data and WR signals to each chip/shift register.

I'll be able to do some testing with each section seperately this weekend.


I was asking the question because it could be a software issue (and not necessarily hardware, as suspect).


Attached are the schematic and board, sure's pdf wont fit so heres a link

So are you saying that you just made a copy of the schematics in the link?
It would be good to see a schematic what you actually made rather than the schematic you thought you copied. If you had a PCB made then you should be able to post a schematic generated by the CAD package you used.

Go Up