MBI5030 starter board — build instructions

IMGP7748

Now that the ‘customs-guys’ have finally released the LED-driver ICs, I can finish this project. 4 Weeks is a looooong time for ‘air-mail’.

If you’re thinking about 100% hand-assembly, start with the LEDs first! Then add all the other small parts, then the potentiometer and as the very last part the IC. Use plenty of flux (paste preferred), clean your tip often and go easy on the LEDs. They burn easily. You’ll probably want a magnifier and lots of light as well.

If you use solder paste, you probably know how to deal with it anyway. So I’ll just shut up. But please make sure you put the chip and the LEDs on the right way. Your multimeter surely has a diode tester ;-)

Design-files and code can be found here.

This entry was posted in Soldering & PCBs. and tagged , , , , . Bookmark the permalink.

2 Responses to MBI5030 starter board — build instructions

  1. Robert,

    I had trouble joining your forum. I found the ATTiny code but by then I was locked out.

    I was looking at your MBI5030 board, and noticed that the code only supports 1 chip. The MBI5030 would be more useful if chips could be cascaded, and I tried to extend your code (using the Fzero – clock & latch within the data stream) formatting, but have not been successful. Sending 32 PWM words with various data/global latches both chips always display the same 16 PWM words. I found another code fragment from a project called RGBClock, but this does not work for 2 chips either.

    Before I re-write from scratch using F1 (clock & latch AFTER the data stream, I was wondering if you have looked into this and have any insights?

    • robert says:

      You can connect as many chips as you like, just not on one pin. Just create new instances for the 2nd, 3rd… chip.

      I was thinking about cascading these chips, but due to the not-quite-SPI protocol (special timing for the latch pin) the AVR hardware-SPI cannot be used anyway (which would have been plenty fast for cascaded chips).

      The current code does bit-banging on a GPIO pin and there are plenty of these :-) As of now, one per chip is required.

      So far I have not looked into cascading at all. I’m pretty sure it should work [somehow]. I like the chip per se, but this special timing requirement gives me the creeps, so I’m not motivated to fight with it.

Comments are closed.