ColorCNC Colorlight 5A-75E/5A-75B as FPGA controller board

[ 25.078490] usb 2-4: Manufacturer: FTDI
[ 25.118608] usbserial: USB Serial support registered for FTDI USB Serial Device
[ 25.118664] ftdi_sio 2-4:1.0: FTDI USB Serial Device converter detected
[ 25.118920] usb 2-4: FTDI USB Serial Device converter now attached to ttyUSB0
[ 25.118972] ftdi_sio 2-4:1.1: FTDI USB Serial Device converter detected
[ 25.119186] usb 2-4: FTDI USB Serial Device converter now attached to ttyUSB1
[ 25.119231] ftdi_sio 2-4:1.2: FTDI USB Serial Device converter detected
[ 25.119427] usb 2-4: FTDI USB Serial Device converter now attached to ttyUSB2
[ 25.119475] ftdi_sio 2-4:1.3: FTDI USB Serial Device converter detected
[ 25.119636] usb 2-4: FTDI USB Serial Device converter now attached to ttyUSB3


JTAG connection:
ADBUS0 -> J27
ADBUS1 -> J32
ADBUS2 -> J30
ADBUS3 -> J31
GND -> J34

I use external +5V power supply connect to Colorlight board.

Problem solved. I was careless when connecting the FTDI module.

Dear all,

I have some old analog servo drive width +-10V analog input.

I would like to implement the following:

I should convert the PWM output of the 5A-75E card to +-10V analog output.

I have no problem with the hardware, I can do that.


I would need another digital DIR output for the litexcnc PWM output, which would change the sign.

I don't know if the current firmware knows this or not.
If yes, how to configure this? For example: output j1:7 should be the PWM output and j1:8 the DIR output.

At this moment this is not supported, as the PWM only has a single output. What do you want to control with this servo? Is it a rotational output or a linear motion?

In case it is a rotational output, one can consider that 0V is no rotation and the negative and positive are clockwise and anti-clockwise motion. One can use the ABS component (linuxcnc.org/docs/html/man/man9/abs.9.html) to split the speed and direction. Feed the speed into the PWM and the is-positive (or is-negative) to the GPIO.

Please note:for Litex-CNC yo can also use the discussions on GitHub (github.com/Peter-van-Tol/LiteX-CNC/discussions). Questions are picked up faster over there and other users benefit from the knowledge base.

I would like to implement something similar to the Mesa 7i77 card.

One option is to use offset mode PWM so 50% = 0V
so you don't need a DIR signal.

PCW wrote:

One option is to use offset mode PWM so 50% = 0V
so you don't need a DIR signal.

Wouldn't that basically reduce the resolution with 50% as well?

Just released version 1.2.0 to PyPi with support for Raspberry Pi 5!

This release brings furthermore:

• An updated watchdog module with an enable pin. This enable pin is HIGH as long the watchdog is happy. It will go LOW when the watchdog bites, for example when the communication gets disrupted. This enable pin can be used to control / disable buffers to bring the machine to a safe state. When recompiling a card, modification of the configuration file is required.
• Several bug-fixes:
  • in the CLI (correct detection of 64-bit RPi);
  • PWM does not generate errors / warning at compilation;
  • pin numbering of HUB75HAT has been corrected (connectors were mirrored);
  • resolved error with GPIO when there are only inputs or outputs defined..

TOLP2 wrote:

PCW wrote:

One option is to use offset mode PWM so 50% = 0V
so you don't need a DIR signal.
Wouldn't that basically reduce the resolution with 50% as well?
 

Yes is does, but a direction pin is hard to use without introducing glitches in the analog output.

What we do with our firmware is add 4 bits of dither  so say at a 80 KHz 11 bit PWM we get
15 effective bits of resolution with the 3 or so KHz bandpass of the PWM filter.

You can also use PDM but it tends to generate non-linearities because of the finite
and asymmetrical switching times of the output devices, Dithered PWM is a compromise
between PWM and PDM