Retro Scoreboard project....7221, flickering LED's, and problems :(

Projects General Guidance
1

Hello, all.

A few weeks ago, I posted some questions regarding some guidance on a retro scoreboard project I was planning. Well, I'm quite a bit down the road on this, finally down to wiring things up...and I'm having...issues.

The most basic thing is setting up a single MAX7221 to drive four dual-digit displays. After wiring up my host board with the 7221, wiring up the caps and the Rset resistors, and getting everything hooked up, I fired up a simple test to send a sequence of 0-9's to each digit following a half-second initialization pause following a static display.

Problem is the digits flicker briefly once the sketch uploads at the static display, and after a few seconds, perhaps not even three or four, the thing shuts down. I can reboot the Arduino, and the process rep. I can see a "ghost" of the digits I'm sending to the displays.

I've got a good +5V to the chip, Rset of 20kOhms as expected, and I've double-checked the cap values for the flitter end, and all seem correct. I did continuity checks as I soldered up my IC socket and pins, and everything tested correctly.

How critical is the "keep the 7221 close to the LEDs" advisory? What's "too far"? Is that a max wire distance from the cathode pin to the chip?

How might I test my circuit to determine if I've wired up the caps properly? Aside from having too long wire lengths between the LED's and the Max/max to Arduino, I suspect I've done something wrong there that's causing the thing to draw too much current and shut down.

I realize I'm rambling a bit, but I'm frustrated, and kinda searching for a best next step. When I breadboarded this up, my sketch uploaded and it worked perfectly the first time, so I think I got a little arrogant as I pressed on to the "real thing." Now I'm not entirely sure where I've screwed up.

Thanks for any suggestions or pushes in the right direction.

-David

2

Okay, I can add at least a bit more detail.

I modified my sketch to run a cycle of 0-9 numerals to each seven-segment display. That works - but even then, some of the segments flicker very slightly. If I try to light up more than one LED digit at a time, that's when problems occur. I'm looking at a power problem, it seems.

Shouldn't the muxing of the 7221 allow me to drive all six LEDs attached to it? The forward voltage of the LEDs is 2.1V and the max current on the LED data sheet (Lumex ldd-c813ri) is 30mA per chip, which means I should under worst case (30mA x 8 segments) be at 240mA, well under 500mA. Yet it seems my LEDs are starving?

My Uno is being powered by my laptop USB; wonder if it is not negotiating more than one load 100mA load unit?

3

OK, I will repeat a popular meme here. I suspect the problem is in the middle of line 37 of the code you have not posted here (according to the instructions). :astonished:

And it could involve the thing halfway down and two inches from the right hand side of the circuit schematic. A photograph would really help, but only if taken in outside daylight so we can see each part. :grinning:

4

I didn't post code as it was taken from a known-good sketch for a 7221 driving two of these same LEDs on a breadboard, so I had no reason to suspect it. I assume my wiring is screwed up somehow.

I'll take photos and update later when I can.

5

I routinely drive an 8-module ( 64 by 8 ) "Times Square" matrix display from a USB port - albeit not on a laptop but USB is specified for a minimum of 500 mA (you need that or more to power little hard drives or charge phones).

One MAX7219 is chickenfeed!

6

Paul__B:
I routinely drive an 8-module ( 64 by 8 ) "Times Square" matrix display from a USB port - albeit not on a laptop but USB is specified for a minimum of 500 mA (you need that or more to power little hard drives or charge phones).

One MAX7219 is chickenfeed!

That's what I was thinking. I suspect I have screwed up the caps on the ground side somehow. I may just rebuild that board.

7

OKay, I realize I said I'd upload pics, but as I studied my board more and thought more about how it seemed to be failing, I concluded the only realistic probability was that I'd screwed up the anti-flutter caps between vcc and ground. I started over on a new perfboard and fresh socket (even the fresh 7221 I have), soldered it all up - and had nearly the same problem.

The final gremlin? A bad ground wire solder. After moving the setup back to my breadboard with a different LED setup, but seeing the same problems, it led me to go back to square one, remove the chip, and start at step 1 with the multimeter. I finally, finally, finally found that the wire from the caps to the ground hadn't soldered in properly. With that resoldered and presuambly fixed, I put the 7221 back in, put it back on my breadboard LED, and voila, I started having numbers work properly.

I then moved it back to my actual model scoreboard project, and my success persists :slight_smile: I have to deal with a couple of bad solder joints on the digits there, but that's an easily fixable issue.

Thanks for the feedback. I'll post pictures of the finished product later :slight_smile:

8

Hey all...while it is still far from finished here's a pic of my work in progress. First eight (4 x dual digit) LEDs in and working. Long way to go.

1960×4032 622 KB

9

Good work with the diagnosis.
Now you’ve learned how important good soldering is!