Doesn't work with Dragonboard 410c

[lmsb]

The display is all garbled. Library and python code examples run fine but actual LEDs display garbled.

[rm-hull]

Can you post details of:

• What command or program you ran
• How many LED matrix units you have, and how you have them wired up
• A picture of the garbled LEDs
• A url link to to the device you have purchased

[lmsb]

Hi @rm-hull ,

Thanks for your help,

• I ran the matrix_demo.py and also tried the basic lines as next:

import max7219.led as led
device = led.matrix(cascaded = 4)
device.show_message("Hello world!")

• I have 4 8x8 LED matrix units cascaded. I bought the one from Amazon, Wangdd22 - http://a.co/iX4wSah

• Please find a video of when running the line "device.show_message("hello")": video on you tube

• The Dragonboard 410c board I have with Debian Linux is the next: DB410c, 96boards

Regards,
Luis.

[rm-hull]

ok, things I would do:

• Remove all the wires, and run an SPI sanity check - see https://github.com/rm-hull/spidev-test
• remove any other wires from the board other than those needed for driving the LEDs
• check that there are no other programs/processes running that also use SPI
• check/double-check my wiring between the board and the LED matrix

[lmsb]

@rm-hull ,
I'm now suspecting that the problem is speed.
I say this because I followed your advise and disconnected everything else not related with the SPI lines, then ran the matrix_demo and still no activity. Then, went and lowered the speed on the spi interface at the matrix_demo.py program and I could see the display with some activity. I tried to reduce the speed further but there's an assert under the "luma.core.interface.serial" library file which doesn't allow me to specify a value of Hz lower than 500,000.
Question: Can I further reduce the speed here? how?
Suspected about the speed first because I'm using a voltage level shifter because the SPI in DB410c outputs at 1.8v while the max7219 display works at 5v. And based on this forum entry I found that we must use a lower speed when shifting from/to 1.8v.

I checked the specs on the DB410c board and the Raspberry pi and the DB410c has a higher clock speed than the Rpi2 but has the same as the Rpi3. But even so, according to this link, the DB410c still remains faster than the Rpi3 under different performance tests.

Another test I did was to use the spidev_test utility and all looked ok as after bridging the MOSI and MISO with a wire I was able to confirm the looped data.

Thanks again for your help.

[lmsb]

@rm-hull ,

Would you have a basic set of byte commands to be used with the spidev_test to drive the max7219 LED display ?

thanks,
Luis.