Duet Wifi config on a Flashforge Creator Pro
-
I got my Duet wifi last night and connected everything up and everything tested out ok. I set up homing but didn't test more yet. I did notice that my Z-stepper was hotter than normal tho and it was making a sound I didn't hear from it before. I set the idle current to 840 for XY, 800 for extruder and 700 for the Z. I suspect that this might be high but I can't find any type info on this at all. Does anyone here run this board on a FFCP printer and would be willing to share some config info in general? Z stepper current would help a lot but would love to see the rest of the config to if possible to check if I did anything wrong.
I also did mount a fan blowing directly over the stepper drivers. Is this needed for the low current stepper motors I'm using? I'd love to share some photos but can't figure out a way to upload them to the forum.
-
You don't need a cooling fan for the electronics with the motor currents that low.
Is there a type number label on the Z stepper?
-
Is there any way to control it so I run it if the temp in the stepper drivers rise or are there no sensor in them?
No, there is no label on it like for the other steppers. It's one with a ballscrew rod mounted directly on it. I tried searching everywhere for it but can't seem to find any info even on flashforges own page. I
-
This is the Z stepper.
http://www.flashforge-usa.com/shop/parts-accessories/z-axis-motor.html
Edit; found this info on a makerbot 1, sounds a lot better now. will measure heat after a while
The MCP4018T-502E/LT digipots value is set to 118 for X, Y, A and B which gives Vref 1.75V (810mA max.), and 40 for Z which gives 0.60V (278mA max.).
Edit2; this solved my heat problem just to reply incase anyone else comes across this.
-
Then I've done my first test print of a cube. Came out pretty good at 0.1mm or so to small in dimensions but I guess that's possible to solve. I did however come across another problem that I couldn't find any solution to on this forum. I use the stock thermocouples type-k from the flashforge and they work as they should but they are occasionally showing 2000degrees for half a second. This is enough to have the fans constant powering on/off when it's idling. This happens to both probes and I can't really see a bad connection. Is there any software adjustments to this or have I done anything wrong?
-
It's possible that you have a bad connection between the thermocouple and the daughter board, but a more likely explanation is that the thermocouple wires are picking up noise. The thermocouple voltage is only about 40 microvolts per degC which is why they can easily pick up noise through induction even though there is a noise filtering capacitor on the daughter board and the firmware ignores short-term errors in the reading. To avoid noise pickup, the thermocouple wires should be twisted pair all the way back to the daughter board - most thermocouples are supplied with twisted pair wires anyway.
If the thermocouple wires are twisted pair all the way back and you still have a problem, try to identify the source of the noise. Most likely it is a stepper motor or a heater. If you reset the Duet and don't move any motors or heat any heaters, do you still get temperature spikes?
You can turn stepper motors off individually using the M18 command.
When you have identified the source of the noise, either route the thermocouple wires away from the wires to the other device, or use twisted pair wiring for the other device too.
-
Would a measurement with a oscilloscope tell you anything about the noise? I got a 100Mhz DSO so I can do a little sample if that helps.
-
Would a measurement with a oscilloscope tell you anything about the noise? I got a 100Mhz DSO so I can do a little sample if that helps.
Quite possibly, but it's probably simpler to start with everything turned off (you can even start with USB power only), check there are no temperature spikes, and then see what you need to activate to make the temperature spikes appear.
-
Will do
Thanks for the suggestions. By the way, the thing about silent steppers was something I thought was a sales gimmick but they are so quiet now that I almost only hear the fans. -
Will do :)By the way, the thing about silent steppers was something I thought was a sales gimmick but they are so quiet now that I almost only hear the fans.
A lot of us are looking for quieter fans! Another option would be to add a thermistor to monitor the hot end heatsink temperature, so that we don't always need to run the hot end fan at full speed. I already plan to add firmware support for a thermostatic electronics cooling fan (although in many cases, the Duet WiFi doesn't need one anyway).
-
I cleaned up some wiring now and separated all stepper wires from the thermocouple wires and it seemed to have solved some of the issue. Before I could see it as long as the others steppers where active but now it seems to only show when I'm printing. That's kinda logic since the thermocouple wires go side by side on the stepper motor wires. Do you think it would help to use a ferrite on the stepper motor wires? If so should I do it at the driver end or the stepper end or both?
This picture shows that the noise before is there even before I start heating/extruding.
Here is a video showing the problem after I separated the thermocouples wires from the rest of the stepper motor wires. All except the extruder stepper motor that is in the same tube towards the hotend. I suspect that to be the noise source.
https://www.youtube.com/watch?v=6hnf7UEhhngEdit: you where spot on. I separated the thermocouple wires from the extruder wires and now I don't get this problem anymore. Adding ferrites to the stepper only made the problem worse. I managed to ruin one of the thermocouple cables when I rerouted them tho so now I only have one measurement. I'm tempted to replace these for some PT100 sensors and new E3DV6 hotends with bowden anyways so this might be a good time to do it. I guess PT100 is less prone to measurements error like this? atleast from what I am experiencing from my work with PT1000 sensors they are pretty stable.
I will post this even if it's not quite on topic here but I guess it can be nice info for a Flashforge owner looking to upgrade to duet so I will post. Delete if it doesn't fit
This is the stock Mightyboard on the FFCP before I started tearing it down.
Everything removed. This is when I first realized that I migth be in over my head on this project with so little documentation on my printer..
The standoffs doesn't fit so I used some longer M3 plastic standoffs to mount the new duet wifi approx in the same place as the stock board went in. Notice how the fan firts perfect to blow at the stepper drivers. I have now disconnected this fan after dc42 said it was not needed at this low stepper currents.
I deserved a beer
Everything hooked up except the hot end heaters. Ready for a first test.
And this is the final result after cleaning some cable spaghetti.
Then I did some z-probe testing today with something I had lying around after a CNC router. It triggered at Z 4.3mm and went off at 7.2mm exactly for 10 tests in a row. Didn't try stepping below 0.1mm so not sure how accurate it is that way. It was only ment as a rough test anyways since I cannot mount it like that on the hotend.
Using some powerfull magnets below the bed as a trigger for it.
Here is a video of the short Z probe testing for those interested.
https://www.youtube.com/watch?v=SezJwvR4vREAnd here is finally a video of the sound reduction. I found a old video where I filmed a random print. It's not exactly the same parameters prolly but it is filmed with the same camera/mic so it gives you an idea of how much lower the sound level is. Impressing!
https://www.youtube.com/watch?v=S_qfKTQFUfU -
If you connect the fans on the extruder heatsink to fan1 in thermostatic mode, the printer will be quieter when the hotend isn't hot.
-
If you connect the fans on the extruder heatsink to fan1 in thermostatic mode, the printer will be quieter when the hotend isn't hot.
I have those fan connected on fan 0 and fan 1 set to start at 45degrees. The part cooling fan is set to start at third level but that one isn't working automatically since I can't tell slic3r that P2 is the part fan. Starting it manual with M106 P2 S200 works perfect tho.
-
If you connect the fans on the extruder heatsink to fan1 in thermostatic mode, the printer will be quieter when the hotend isn't hot.
I have those fan connected on fan 0 and fan 1 set to start at 45degrees. The part cooling fan is set to start at third level but that one isn't working automatically since I can't tell slic3r that P2 is the part fan. Starting it manual with M106 P2 S200 works perfect tho.
I see three solutions, if I understand the setup correctly:
1. Connect both fans on the extruder heatsink in parallel to the Fan1 connector, then the Fan0 connector will be free for the part cooling fan.
2. Swap the Fan0 and Fan2 connections.
3. Use the fan mapping feature of the M563 command.
-
Thanks dc42
I haven't had time to check what command slic3r uses in the gcode yet to start the fan but my first thought was to do that and then remap the extruder fans so it works through slic3r. Will have a look at your methods to but first it's Christmas time so will be a few days -
Sli3r sends M106 commands with no P parameter to control the print cooling fan. So it needs to be connected to Fan 0 unless you use fan mapping in the tool definition.
-
Then I've fitted the IR probe to my flashforge. Just test mounted it for now since I have a E3DV6 in the mail. I have set it up with G29 mesh probing and got a good result when I did the first run. -0.078/0.037 was the first test. When I did one after I put glue stick on I had offsets from -0.230/-0.030. Isn't that much deviation? I do a G30 when the prints starts (in my G28 routine) can that disturb anything?
And I couldn't figure this out but do I need to do something for the printer to remember my G29 height map or is that done automatically until I do a new G29 S0. Even after power cycling the printer?
-
I had the same exact noise issues with my Creator Pro , I dumped the thermocouples, the hot ends, the fans, the carriage, the stepper motors etc. I had good luck with a normal thermistor and now I'm running PT100's with an E3D Cyclips / Chimera and external Bowden extruders.
In any event the only thing I found that helps with the thermocouple issue is to get the wires far far away from the stepper motor wires. They induce current into the thermocouple wires and you get all that noise.
Jeff
-
The thermocouple problem is solved with routing the wires away from the stepper wires. Haven't had a problem since that. I will change them as soon as my E3DV6 extruder comes with a normal thermistor.
Got any pictures of your setup? The carriage with those springs that wobble all around is a nightmare..
-
I'll get some pics of my setup. My goal was to reduce weight as much as possible to achieve better accuracy and/or far more speed. I replaced the stepper motors a few times tying out some combinations. I'm currently running a short pancake motor on the x-axis to cut down even more on the weight the Y axis motor has to move around with great success.
The stock Z motor on my FFCP is only rated for 400 ma max and the X and Y at 850ma. They are high inductance motors and really are kind of designed for higher torque and not so much speed. I switched to 1.2 A on the Y side with a different stepper and 1 A with that pancake stepper on the X side. I can do 150mm/s prints without issue and normally print at 90mm/s. I also went with 0.9 deg large steppers for the bowden extruders on the back with mk7 drive gears. Don't fear the bowden, my prints are of far better quality then I ever got with the MK10 setup.
I've got a custom E3D V6 carriage that I snagged off TV and then modded to work a bit better including mounting the stock active cooling fan, inductive sensor or BLTouch. I also of course have a custom carriage I designed for the E3D legends pack. Spring loaded bed clips for holding glass that I printed in Polycarbonate that make switching out glass a 1 second job. Also my own (mostly) design.
Jeff
