by

|   December  2, 2008 at 0:42 AM

Hi,

For a project i'm working on I want to PWM control the LED's. I wrote new firmware for this using WinAVR.

You can download my firmware from:
http://www.microframework.nl/2008/12/02/rgb-led-matrix-now-with-pwmrgb-led-matrix-now-with-pwm/

by

|   January  1, 2009 at 5:04 PM

These are a lot of fun!

http://flickr.com/photos/22886197@N00/3156366100/

by

|   February 14, 2009 at 4:10 PM

Please note that these boards have been revised and now have a separate SPI input & output on the back, which allows for easy daisy-chaining of modules.

Even more exciting - they have updated firmware and use a bit encoding scheme to control the color, allowing for up to 196 colors (3x3x2 bits R/G/B, respectively).

Are there any plans to update the data sheet, or provide sample code for this new firmware?

by

|   February 16, 2009 at 9:22 AM

Oh man, I hate writing datasheets! But you're right; we've updated the code and made the SPI interface and display code much faster.

I will start working on a new datasheet and post some sample code soon.

by

|   February 17, 2009 at 11:35 AM

Please provide some info about the new firmware. And maybe some example code.

by

|   March 20, 2009 at 7:05 PM

Yeah, both the datasheet and the tutorials on this are pretty out-of-date. Would greatly appreciate a new one - I just bought two and was thinking of buying more, but still haven't quite puzzled out the ones I have. :)

by

|   March 23, 2009 at 8:56 AM

Have you downloaded and checked out the 'Sample Interface Code' available in the documents section of the Product Description? This should give you a good idea of how to send data to the matrix controller.

by

|   March 29, 2009 at 3:39 PM

Thanks Ryan, that's definitely helpful.

Could you say something about how daisy-chaining the new version of these modules is supposed to work? I had read something intriguing online (can't find it now) about extra bytes beyond the first 64 received being dumped out the DataOut line while the other lines are being passed through. Also, what's the extra VCC line for? If I just do the naive thing and plug them together, they seem to mirror each other.

by

|   April  2, 2009 at 1:05 PM

The extra VCC line is frivolous, I was going to have a spare unconnected pin on the headers so I just decided to tie it to Vcc.

If you just plug them in and send the boards twice the amount of data, you will see some weird behavior. The easiest way to daisy chain them is to download the RGB Serial Matrix Firmware (recently updated to v3), and open the RGB_Backpack_v3.c file. Change the "NUM_BOARDS" parameter on line 46 to reflect the number of matrices you are daisy chaining together. After you've done this, recompile the code and then reprogram each of your matrices with this new hex file.

After doing this you should be able to just send 64*NUM_BOARDS bytes to the first board to get it to display correctly. Start the sequency by sending the data for the last matrix in the system first.

by

|   April  3, 2009 at 8:20 AM

Hm. So if I understand this correctly, in order to daisy-chain these modules I'd need to buy an STK500-compatible programmmer, solder on headers for the ICSP port for each module, then figure out how to build and flash the firmware on each module using WinAVR. Given the clearance between the backpack and the module, that looks a bit tricky.

At this point it might be easier for me to just drop out the CS pin on the cables that connect the modules and run it separately to my controller.

by

|   May 13, 2009 at 11:10 PM

Just to be clear, it is not necessary to solder the ICSP headers into place to reprogram the modules. You can simply hold it in place, applying good pressure to all six pins. Many of the boards we sell at Sparkfun come without headers, and thus we have to use this method to do our initial programming. There are other ways, such as pogo pins, that can make reprogramming any Sprakfun device solder and hassle free. Although you would still need an STK500 MKII to do the programming, you wouldn't have to disattach the matrix to get to the headers.

by

|   March 25, 2009 at 11:37 PM

I would like to echo the above statements. A new datasheet and some test code really would be awesome. I bought one of these and have not yet been able to address each of the leds and colors properly.

I have found a few different pieces of code to drive the older versions, but they only light random leds on this model.

I certainly don't want to rush great work ryowens84, I am sure you are busy. Any information would surely be appreciated.

by

|   April  2, 2009 at 12:58 AM

The "Sample Interface Code" which is downloadable above shows you how to send color values to the LEDs.

by

|   April  7, 2009 at 9:35 PM

Well, 5 volts all the way works fine.

by

|   April  8, 2009 at 7:26 PM

Is it true that you need to upload new firmware on the new rgb led matrix backpacks in order to pwm the leds? or do they come with this capability?

If you can pwm them.. How ?

thanks.. Just want to adjust the brightness while spinning a potentiometer for now...

by

|   April  9, 2009 at 8:09 AM

No, the default FW controls the PWM; currently up to 64 colors can be displayed. All you have to do is send different values to the matrix controller. Each byte that is sent to the controller represents the color for an LED. The byte is structured:

RRRGGGBB

So to change the brightness of any color, just lower the value of the 3 bits corresponding to that color in the byte that you send.

by

|   April 10, 2009 at 10:49 AM

Hmm. Do you know of any code (arduino?) that might be helpful?
I probably should just hire a programmer.

by

|   April 10, 2009 at 12:42 AM

Yes, there's a sample arduino sketch in the 'Documents' section of the product description above.

by

|   May 11, 2009 at 5:36 PM

I think I'm confused, but do you mean 256 colors or 64?

by

|   April 25, 2009 at 2:40 PM

is there a way to talk to this module using two serial lines, TX/RX?

by

|   May 11, 2009 at 5:30 PM

Is it possible to order just the backpack, without the LED matrix?

by

|   June 21, 2009 at 7:34 PM

I would like this as well, I think that this would be a nice controller for the rgb button pad, using 4 of them to make a 8x8 array and this backpack as the led controller.

by

|   June 15, 2009 at 7:01 AM

Could someone give me an idea of how bright these things are?

Could I tell what color the lights were if direct sunlight was shining on the board?

Thanks!

by

|   August 13, 2009 at 10:55 AM

Hi,
I have an 'RGB LED MatrixController v2', I can see only 7 colors. This is normal?

Thank you in advance

by

|   August 29, 2009 at 1:44 PM

Ryan, (I'm a newb) any reason you didn't use the 74HC595 (CMOS) instead of the 74LS595N (TTL) chip? Seems like the 74LS595N is much more expensive! Also, your work freaking rocks!

Jay

by

|   September 23, 2009 at 2:44 PM

Part 1

So this was interesting... I noticed that my animations started to look a bit choppy. When I slowed down my animation it looked like it was updating the screen two columns per refresh, as opposed to 1 (which i programed). After some thought i hypothesized that it was only printing ever OTHER frame. I confirmed this behavior with some short test code.

So i started to audit my code to see if i was doing something wrong, or had hold timers to short or anything like that. I didnt find any error in my code.

After getting slightly confused i stood back and just started to observe again. It was then that i noticed something weird. When the board starts up, it displayed 2 red, 2 green, and 2 blue dots. This is done from the firmware on the device. Usualy this is displayed on LEDs 0-5. However, my board was using LEDs 1-6. I thought this was weird and assumed that i corrupted the boards firmware. But i still had to double check what the boards firmware did.

by

|   September 23, 2009 at 2:45 PM

Part 2

It seems that the placement of the LEDs on that initial screen is based on how many board the system things are connected together. This arrangement means that the system thinks that two board are connected together. This explained it all. Every OTHER frame was being displayed because the second frame was being sent to a phantom board.

Upon further analysis of the firmware, it seems that after 255 times of the board being power cycled, the EEPROM location that stores that there is 1 board being connected gets rewritten. So i just power cycled my board a few time and it started behaving properly again....

I guess during the development of the firmware, they encountered this situation before. That would be the only reason i would think they had the routing to rewrite to the eeprom. Anyone from sparkfun care to comment.

by

|   October 11, 2009 at 2:37 PM

The product I received when I ordered it 12/08 is labeled as "RGB LED Matrix Controller V2". I can't seem to find the datasheet for this version of the controller. For example, what is the CS pin; how is it different from the slaveselect pin on the layout pictured here?

If the V2 requires updated firmware, are there tutorials you can point me to to update it?

Right now when I run the arduino sketch attached here I get one blue LED. I'm guessing this is not what the sketch is supposed to be doing.

by

|   November 15, 2009 at 0:46 AM

whether the tools is ready stock?

if I would make dot matrix 32x32.what avr 8535 can be used as controllers?

how the schematic?

by

|   February  1, 2010 at 11:42 AM

Hi,ryowens84

i have two LED Matrix and want connect togeter,
from datasheet i must use SPI interface and enter '%' in my arduino program to change NUM_BOARDS,
but it's no work,could you write example with how to change NUM_BOARDS for arduino ?

by

|   February  6, 2010 at 7:25 PM

How about some eagle files?