questions about powering leds in large matrix like mikes macmux design

Status
Not open for further replies.
Hey deup', looks very nice.

Did my first final today. Got more tomorrow and Wednesday.

Good luck at Old Dominion...

Happy Holidays. Mike
 
Omg wire

Good luck on your exams, I managed to do good on mine last week.

A small update, as my back needs a break. Wires, goddamn wires, and tight spaces:
 

Attachments

  • OMGWIRES.JPG
    1.3 MB · Views: 278
  • OMGWIRING.JPG
    1 MB · Views: 286
  • ALL LIGHTS IN.JPG
    1,009.4 KB · Views: 285
Fluence,

Red, Green, Blue and White.

red and blue run off the same columns, so they share the same resistor setting the current, which allows the red to be driven brighter, as it needs to be to match up with the more efficient, white and green and blue diodes. red will be driven at 80ma and the blue will be driven at 60ma, the green and white share a similar concept in columns but with a much less difference in ma. Since only one row is lit at a time, and each LED has it's own current source, there are no color mixing issues as you would have if you simply wired red LEDs in with green LEDs in parallel.
 
You have somekind of special LED? I never saw 80ma and 60ma LED.

Or you are using they're max peak current?

Do you think (if you're using normal LED), that a "peaked" LED at 60ma is brighter than a "always on" 25mA LED?
 
We are finally getting the columns soldered, so far a 1/2lb of solder has been burned, 4 tips gone, and ~700 feet of 12guage copper wire in place, I'm etching some circuit boards tonight, gonna be fun!
 
Comming along fast, only a little over 1300 soldering points left!

Mike, I was wondering which method you prefer to send data out to the chips? I just don't want to get a whole program running with inefficient code and figured your experience would know which is best.

more teasers:
 

Attachments

  • PC200746.JPG
    1.1 MB · Views: 279
more pics, The testing circuit, the row sinking circuit board, and a shot of the project
 

Attachments

  • PC210751.JPG
    1 MB · Views: 295
  • PC210756.JPG
    1.1 MB · Views: 269
  • PC210753.JPG
    1 MB · Views: 274
deuplonicus said:
... Mike, I was wondering which method you prefer to send data out to the chips? I just don't want to get a whole program running with inefficient code and figured your experience would know which is best. ...
Click to expand...

I suppose it depends on what you're doin' with the display.

If you're building a scrolling message display then you would need a ROM character table aray and a 128 byte "display" array (assuming a 16x64 display). You would place display data into the RAM "display" array in your main program and your interrupt driver would load 128 bits of fresh 'row' data from the "display" array into the driver IC shift registers during each row update interrupt.

If you're just painting large number characters onto the display for a New Year count down then you could probably pull the display data directly from character patterns contained in ROM arrays and skip the RAM "display" array entirely.

Happy Holidays!

Cheerful regards, Mike
 
Last edited:
Actually I'm planning to use SPI for faster transfer, and to change the entire image quite often, hardly ever scrolling. So what I was thinking would be best is to have 2 arrays. One for the data you wish to display soon, and another for the SPI transfer. somewhere along there would be a bit that you set when the data is ready to swap, so in an ISR that data will be swapped out from one matrix array to the other matrix array.

here's yet another teaser, finally got one 64 led section working! now it's crunch time! 7 more days to finish!

Oh yeah here's a weird note, maybe someone can help out with this problem:

In the picture showing running LEDs there appears to be lit Green LEDs elsewhere. Those green LEDs as well as just about all the green LEDs turn on dimly whenever the columns switch, what is perplexing is that at the moment, those dimly lit LEDs only have a ground, their positive anode columns are currently NOT connected to anything those LEDS simply have a ground and are connected to 12" lengths of wire for the columns (anodes). Any ideas what would cause this?
 

Attachments

  • PC220788.JPG
    1.1 MB · Views: 267
  • PC220783.JPG
    1.3 MB · Views: 267
  • PC220780.JPG
    1.2 MB · Views: 284
Last edited:
256 LEDs are working! program is sufficiently fast! pics later, I'm tired, been soldering all morning, now its time for christ mash dinner!

only 1218 LEDs to go
 
ok guys I'm encoutnering data corruption issues by the 4th shift register, it'll start losing data and or the shift registers act like they're receiving a faster clock cycle than I'm sending it, and the data will all of a sudden be a scroll of all 0's or all 1's for up 30 bits....

Any idea how to prevent data and clock signal corruption over 14 shift registers and 10 feet of distance?
 
the distances would still be too long doing that.

I have good reason to believe it is EMI causing data corruption, as the other LEDs that are connected to the ground, but have NOTHING connected to the (+), are flashing with the frequency of the light a foot or more away from them, though really dim, this would mean obvious EMI all around.

is there an easy way to shield these wires?
 
Use twisted pair, and/or shielded cat-5 type cable.

You may try just wrapping them in foil and ground 1 side.
 
Last edited:
sweet, if foil works, I'm gonna make me a foil hat and wear it all day.

Just got back from the doctors, have a horrible 3 week old case of bronchitis and this project has been making me stressed out. So I finally got some antibiotics, now I get to have diarrhea as well! better go buy me some yogurt.

I'm going to put up a video later showing the corruption of 600+ LEDs beeing sent a simple scroll of 0's and 1's, as the move down the scroll they lose data, and or speed up, and by the 5-6th shift register til the 9th it just flashes on and off, no scrolling, all three shift registers together, with no delay, so the clock MUST be speeding up by some sort of corruption.
 
Show the leds.. not the diarrhea!

Actually, I would like to see the resulting corruption from the line length.

Im sure everyone who has ever wanted to build such a beast would be interested in it also.
 
Status
Not open for further replies.

Similar threads

  • Question
Large C connected to CMOS Inputs
2
Replies
26
Views
4K
eTech
E
  • Question
Digital potentiometer problem and questions
Replies
7
Views
7K
GH Crash
I
  • Question
how can i improve the milling of small pad especially for MCU like stm32750vbt6
Replies
8
Views
701
ZipZapOuch
Switching cob leds
2
Replies
25
Views
5K
KevinW
C
  • Question
am my steps about my question solution about bjt correct?
Replies
9
Views
4K
MrAl
M
Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…