6XD I/O >> Issue solved

tecno

@droftarts
Hi Ian,
Re multimeter = no as you can see it is not so accessible in the video. Later on today.
https://photos.app.goo.gl/L56Zn79kaww26Q4U6

out2 is OK

2024-09-25 09:26:27 m586
CORS disabled
HTTP is enabled on port 80
FTP is disabled
TELNET is disabled
Multicast Discovery is disabled
MQTT is disabled

Cheers
Bengt

tecno

Pictures of 6XD board https://photos.app.goo.gl/6ME2iY7CesXBjeKPA

droftarts

@tecno Were you able to do any tests with a multimeter? I can't see anything obvious from the pictures.

Ian

tecno

@droftarts
Pls repeate exactly what you want me to measure now that the board is in the open.

droftarts

@tecno From the video I can see that you're using 3 of the pins of each driver connector: 5V_EXT, D#_STEP- and D#_DIR-. I can't clearly see Driver 5, but I assume it is the same as the others.

With the multimeter, and the Duet powered, and being careful not to touch probe to the wrong pins (especially not to touch across 5V pins to 3.3V pins):

Test for 5V between 5V_EXT and GND on the driver 5 connector. This should read 5V.
Test for 5V between 5V_EXT and D5_DIR-. Send a command that moves the axis one way, then the other. You should get 5V in one direction, and 0V in the other direction.
It may be difficult to see a signal if you put the multimeter between 5V_EXT and D5_STEP-, because the step pulses happen faster than the multimeter can register them. You could try doing a fast move, which might register.
You could also do the tests on Driver 4, so you can see what the expected response is.

Let me know how that goes.

Ian

tecno

@droftarts

Hi Ian,
Looks like the 3 wires original cable is not to par, brittle so crimps go bad. Will make new wireing to all motors.
This el cheapo chinese cable, found also twisted wire splices with only shrink tubing loosely fitted.

DMM test shows driver to be OK

I hope I have enough crimpconnectors

Cheers
BEngt

droftarts

@tecno Okay, that's good news, I guess! But more work for you. Let me know how you get on.

Ian

tecno

@droftarts

Yes a lot of extra work to make new cables. Will report!

tecno

New set of cables ready.

tecno

@droftarts
Hi Ian,

All axis are now moving

Happy camper here
Bengt

droftarts

@tecno Phew! Glad it's all working. Hopefully your faith in the 6XD is restored.

Edit: You weren't tempted to wire in the enable signal as well?

Ian

tecno

@droftarts

No need for enable as I see it. 6XD is a very nice controller.

Now I have to sort out further I/O

dc42

@tecno you only need the Enable connections to the drivers if you want the Duet to be able to turn off the motor current. If you leave them disconnected, the motors will be energized whenever power is applied to the drivers.

tecno

@dc42
@droftarts

Just finished cable from 8 LED circular strip from
Ali DC5V WS2812B Led Ring 5050 RGB Individul AddressabIe Round Lights Module 3Pin Black White PCB 8 16 24 35 45 Pixels BuiIt-in IC

Only one led lights up so my question is my settings OK?

M950 E0 C"led" Q3000000
M150 E0 R255 P128 S20 F1

Cheers
Bengt

droftarts

@tecno In M150, the F parameter tells the firmware what to do next:

Fn Following command action. F0 (default) means this is the last command for the LED strip, so the next M150 command starts at the beginning of the strip. F1 means further M150 commands for the remainder of the strip follow this one.

So you need F0 to finish and send the command, eg:
M150 E0 R255 P128 S20 F0
This should set LED string 0 (E0) to 255 Red (R255) at brightness 128 (P128), 20 LEDs (S20 you only have 8 but doesn't matter) and no more commands (F0).

If the first LED is getting stuck on red or another colour, or you can't control any of the LEDs, what firmware version are you using?

Ian

tecno

@droftarts
Duet Web Control 3.5.1

Yes it is stuck on one LED no color change

droftarts

@tecno RepRapFirmware 3.5.1 as well? Please update to 3.5.3 (RRF and DWC) and test, there have been some fixes for LEDs. https://github.com/Duet3D/RepRapFirmware/releases/tag/3.5.3

This was a fix in 3.5.2:

[Duet 3 MB6HC] [Duet 3 MB6XD] When a Neopixel LED strip was configured on a part other than the LED port, the first LED in the strip sometimes had the wrong colour (issue 996). This was a new bug in RRF 3.5.0.

It could also be the M950 Q parameter. We say:

The Qnn parameter sets the LED clock frequency. This is 4x the bit rate. Most datasheets for LEDs (at least SK6812 and WS2812B based LEDs) suggest a maximum data rate of 800Kbps, so 800000 * 4 = 3200000. The default of 3000000 appears to work well with most LEDs.

So try M950 E0 C"led" Q3200000, and M950 E0 C"led" Q2800000, but after you have updated the firmware.

Also, check your wiring!

Ian

tecno

@droftarts

Wires OK

Same lonly LED

Sanity check

droftarts

@tecno I've just tested it on a 6HC, on RRF 3.5.1, and it's working for a 7-segment LED strip. All light up correctly. The fix in 3.5.2 was if you were using IO ports for LEDs. I don't have a 6XD to test, but it should be exactly the same as 6HC.

Can you send M115 and post the response?
Can you post a picture of the other side of the neopixel ring?
Also, I'm not sure using the uninsulated wire as the high frequency data line is a good idea! Most likely it has higher resistance than a 'normal' wire, and may not be transmitting the signal effectively. Use a different wire.

Ian

tecno

@droftarts

m115
FIRMWARE_NAME: RepRapFirmware for Duet 3 MB6XD FIRMWARE_VERSION: 3.5.3 ELECTRONICS: Duet 3 MB6XD v1.01 FIRMWARE_DATE: 2024-09-18 11:27:56

This cable is a HQ Audio twisted pair with foil shield. Change White to shield.