RGBW Neopixel on duet 2 wifi

droftarts

@Woody78 See https://docs.duet3d.com/en/User_manual/Connecting_hardware/IO_Neopixel_DotStar

M950 E0 T1 C"connlcd.5" U3

For RGBW, your M950 T parameter should be T2.

M150 E0 X3 Q3000000

This line is redundant, and should report an error. M150 doesn't take the X or Q parameters in RRF 3.5 and later, they are replaced by M950 T and Q. Q3000000 is the default.

M150 E0 R255 P255

Add an S parameter for the number of LEDs you want to set. Without it, it's only setting one. The F parameter, which tells the firmware if this is the last command setting LEDs, defaults to F0, which tells the firmware to send the command to the LEDs, so you only need an F parameter if there are subsequent pixels to set.

Try:

M950 E0 T2 C"connlcd.5" U3
M150 E0 R255 P255 S3

I'd try the LEDs without the level shifting, in case that's causing a problem. Most Neopixels I've tried recently work fine on 3.3V signalling.

Ian

Woody78

@droftarts i´d tried your config with and without level shifter. but nothing happend... i don´t know why...
it is possible to test the led´s on another way... to check?

droftarts

@Woody78 What firmware version are you on? Post your full config.g, please.

Ian

droftarts

@Woody78 I've just tested on a Duet 2 Wifi on RRF 3.5.1, works as expected. Maybe post a picture of your wiring, too.

Ian

Woody78

@droftarts okay, i checked another led strip in spare directly on the board without level shifter. it worked fine. so i have to check my wires to the stealthburner, i hope it is no problem that the wires are over 2m long

droftarts

@Woody78 I wouldn't have thought 2m would be an issue. I have a hobby servo running of a Duet 2 WiFi, on 3.3V, with a wire that long, and it's fine.

Ian

Woody78

@droftarts do you know if i can measure the 3.3v every time or only when the board gives a signal?

droftarts

@Woody78 I don't know. I would have thought that the signal would be 3.3V, so it would be 0V unless a signal was sent. But there may be some other signal, eg a clock signal. Measure at the Duet board end, either with no LEDs connected, or with the LEDs that work connected.

Be very careful not to bridge your multimeter probes to other pins if you probe directly onto the Duet pins; the 5V pin is very close to GND, and you don't want to put 5V into the connlcd.5 pin, it will destroy the MCU. Better still, connect some wires to the pins you want to test, measure remotely.

Ian

Woody78

@droftarts yes i know. i have only the connection problem in my last connector on the hotend...

DonStauffer

@Woody78 Be aware that on the Duet 2 M150 will only work when there is no movement of the printer carriage. The workaround I'm using is the Adafruit NeoDriver card, which works by I2C communication between Duet and NeoDriver, so you use M260 instead of M150. It's not fast, so you won't get fancy stuff, but it works for progress metering adequately, I think.

If you use the NeoDriver, be aware that although the NeoDriver has pull-up resistors built in, you have to add your own too. I think 4.7 k ohm, but in theory that depends on how long your I2C wiring is. The resistors go from 5V+ to each of the I2C lines (data and clock). Before I added pull-up resistors I was getting intermittent "I2C Transmission Error" problems.

Using the NeoDriver, you don't need the voltage level shifter circuit.