Controlling a Cetus3D with Duet3D 0.8.5

[A Former User]

@ted said in Controlling a Cetus3D with Duet3D 0.8.5:

As I'm fine with that temperature I want to ask you if it is possible with the Maestro to limit the current or voltage for the heatbed as well?

It is only possible to limit the average power, by means of controlling the maximum duty cycle of the PWM signal for the bed.

https://duet3d.dozuki.com/Wiki/Gcode#Section_M303_Run_heater_tuning
https://duet3d.dozuki.com/Wiki/Gcode#Section_M307_Set_or_report_heating_process_parameters

P parameter in M303 and S paramterer in M307

[Veti]

if you get an external mosfet for the hotbed its easy to use 2 power supplies. You can even use different voltages that way.

[Ted]

Thanks for the fast answers! I then will buy a new PSU. I go for a one PSU for everything solution.

Are 320 Watts of the HLG-H320-24A enough for that small heatbed and then 100°C? Next would be 480 Watts and much more expensive. I prefer the HLG series as it works with passive cooling.

[Veti]

there is the Meanwell UHP-350-24 in between.

[Ted]

I now migrated to an Duet Maestro to using a absolute silent and not even warm HLG-H320-24A PSU.

While I'm very happy with the basic setup I have problems with those original high impedance motors. The problem is the same as mentioned before: The motors are getting very loud when moving. For me that happens only for fast travels starting with > F270 movements (G1 Z50 F270 S1 at 256-microsteps and 1359,68 step/mm). Altering the relevant depending values this is regardless of microstepping mode (1-256) and current (200-1400 mA) chosen.

With command G1 Z50 F270 S1 the movent can't be heared at all.
With command G1 Z50 F273 S1 it is a random mix of loud and absolute silent movements intervals.
With command G1 Z50 F275 S1 it is a loud all the time.

I've attached my config.g to this post: 0_1556283965533_config.g

My motors have something about 30-31 Ohm per phase.

Do you have any idea?

[Veti]

M350 X256 Y256 Z256 E16 I1 do you really need this? this might put quite a high strain on the cpu. 16 with interpolation should be fine in most cases

M203 X1000 Y1000 Z1000 E1000 Thats a very high z speed.

You Probably want to adjust the Hybrid Threshold V for the motors to not go into spreadcycle at higher speeds
see https://duet3d.dozuki.com/Wiki/Gcode#Section_M569_Set_motor_driver_direction_enable_polarity_and_step_pulse_timing

[Ted]

@veti said in Controlling a Cetus3D with Duet3D 0.8.5:

Hybrid Threshold V

Was hard to find out what you meant what you meant with Hybrid Threshold V but found the relevant parameter. It is working now – In absolute silence! Thanks a lot for the hints!

I set the threshold to 100 (whatever unit this is) by extending the already in my config existing command M569 P0 S0 to M569 P0 S0 V100 for all my axes (P0 up to P4).

I not completely understood how this threshold works. Especially I'd like to know why someone would want to switch to Spread Cycle for higher speeds? Higher torque?

[Veti]

@ted said in Controlling a Cetus3D with Duet3D 0.8.5:

I not completely understood how this threshold works. Especially I'd like to know why someone would want to switch to Spread Cycle for higher speeds? Higher torque?

if you enter M569 P0 you will see the speed at which the driver switches from stealthchop to spreadcycle.
however at high speed (i think above 60mm/s) the switch can cause layer shifts.

to enable spreadcycle all the time use the D2 option

[2rad]

i did the conversion as well with a maestro.it took me a long time to figure out the trick with the V value;-) now i have a question. i set my V values below 100. can this cause a problem? right now im impressed how quiet the printer is!

M569 P0 S0 V70 D3
M569 P1 S1 V70 D3
M569 P2 S1 V120 D3
M569 P3 S1 V120 D3
M350 X64 Y64 Z64 E32 I1
M92 X320.00 Y320.00 Z320.00 E232.00
M566 X1200.00 Y1200.00 Z1200.00 E1200.00
M203 X30000.00 Y30000.00 Z1500.00 E1500.00
M201 X3000.00 Y3000.00 Z500.00 E500.00
M906 X400.00 Y400.00 Z500.00 E500.00 I60

@fotomas thanks for this entry it helped a lot!!!

[dc42]

@2rad said in Controlling a Cetus3D with Duet3D 0.8.5:

i set my V values below 100. can this cause a problem?

Yes, also you may get missed steps spurious short-to-ground errors reported. The stealthChop mode doesn't work well at high speeds.

Another cause of noisy motors is running out of motor voltage at high speeds. The motor EMF calculator at reprapfirmware.org will tell you the speed at which you start to suffer loss of torque. The motors will become noisier above that speed too because the driver can't implement microstepping properly.

[cookie]

@fotomas I've successfully tested the Cetus by heating up hot end to 240c and heated bed to 90c to 2 hours .. and it seems stable now.

Now it's time to level the bed and test printing. However, I found that the end stop for Cetus is at the top, so Z = 0 cm is at the highest level and Z = 18+ cm is at the bed level.

So, how did you config the Duet with Cetus flipping Z?

Update 1:
I have changed endstop location to "At High End" and flip z motor direction and the machine can home successfully .. then when I try to move in -Z nothing happened but when I try +Z it keep going up against the end stop! But after that stuttering I can move in -Z about 2.5cm and that's it.

Update 2:
I've found the solution by editing "homez.g" in setting:
;G92 Z0 ; set Z position to axis minimum (you may want to adjust this)
to
G92 Z180 ; set Z position to axis maximum

Thanks,
Cookie

[fotomas]

@cookie
I am using an IR sensor so I have z 0 at bed.
But what you need to do is set that the end stop is at the high end of the axis and in the homing file home to Z 250 mm.

The exact commands maybe someone else can help with?

[Veti]

@dc42 said in Controlling a Cetus3D with Duet3D 0.8.5:

Yes, also you may get missed steps spurious short-to-ground errors reported. The stealthChop mode doesn't work well at high speeds.

what i have found is that with 24v stealthchop2 works ok at high speed.
what does not work at high speed is hybrid mode. Switching from Stealhchop2 to Spreadcycle at higher speeds will cause layer shifts.

so i have set my hybrid mode to 200 so that it does not do the switchover and it seems to work.

[dc42]

@cookie said in Controlling a Cetus3D with Duet3D 0.8.5:

Update 2:
I've found the solution by editing "homez.g" in setting:
;G92 Z0 ; set Z position to axis minimum (you may want to adjust this)
to
G92 Z180 ; set Z position to axis maximum

It's better to remove the G92 command completely. Set the M208 upper Z limit to the height at which the endstop switch triggers.

[cookie]

@dc42 said in Controlling a Cetus3D with Duet3D 0.8.5:

It's better to remove the G92 command completely. Set the M208 upper Z limit to the height at which the endstop switch triggers.

Thanks, I've replaced G92 with M208 in both homeall.g and homez.g already.

I've tried a bit with manual mesh bed compensation using G29.
However, the height is about 18 cm. so I'm not sure I should set the machine Z maximum to 17.5cm or 18.5 cm prior doing mesh bed compensation .. the hot end try to dig the bed if I do 18.5 cm .. So should I set the height a bit shorter to 17.5cm?

Would there be any effect if I set the height quite short and compensate like 4-5mm for all point? @dc42 any suggestion about how to do it correctly?

[dc42]

@cookie said in Controlling a Cetus3D with Duet3D 0.8.5:

However, the height is about 18 cm. so I'm not sure I should set the machine Z maximum to 17.5cm or 18.5 cm prior doing mesh bed compensation .. the hot end try to dig the bed if I do 18.5 cm .. So should I set the height a bit shorter to 17.5cm?

I suggest the following:

1. Home the printer.
2. Position the nozzle over bed centre.
3. Jog the nozzle down until it is just touching the bed. If it won't go low enough, send M564 S0 to remove limits temporarily.
4. Send G92 Z0 to define that as Z=0.
5. Send G91 G1 H3 Z999 Fxxx where Fxxx is the same feed rate that you use for homing Z. This will stop when the switch is triggered and set the M208 axis limit to the measured position.
6. Read off the Z height.
7. Use that Z height in your M208 command.

[cookie]

Hello,
@dc42 Thanks for very clear instruction, I've got it working

Now the next step, I want the heated bed to cool down to 32 Celcius before beginning auto eject macro which I've written for small PLA prints. However, I noticed that M190 is not working under 40 Celcius .. bummer .. not sure why. So how should it be done?

Cookie

[dc42]

Ambient temperature around a printer can be up to 40C under some conditions, so RRF treats temperatures below 40C as cold.

You could wait for temperature to fall to 40.1 then use G4 to wait an additional fixed period .

[Peter120]

I designed some parts for my Cetus Duet2 WIFI conversion
https://www.thingiverse.com/thing:4322092
Don’t judge me but I started this about a year ago in the middle of some online configurator tool issues and had to modify the config files the hard way, much easier now.
I choose inductive sensors for homing all the axes. All three of my inductive sensors are stationery requiring Z be homed before Y, however, this keeps the accelerating mass down and the sensor wiring stationary. To interface the inductive sensors output to the Duet I used Schottky diodes. Looking at the numbers I believe the lower forward diode voltage drop meets the worst case Duet input voltage thresholds, regardless it works fine. The diodes are wired so they can only pull down the Duet input pin while the inductive sensor output signal swings from 0.6V to 24V. The homing is done in 2 passes, fast and an extremely slow second pass (F0.1). The resulting trigger position is very repeatable and any faster results in the inductive trigger position having 2 or more nearby but stable values. Could be sensor jitter or CPU capture jitter, anyway it is very repeatable if done extremely slowly.
Cheers
Peter

[Peter120]

After a bunch of tuning I converged on a few variables that might be useful to others.
I choose 256 micro stepping without interpolation. Extrusion steps for the original MK2 extruder mechanism that has a filament release lever is 1650 steps/mm. I also have the warranty replacement MK3 style but the prints were just not as good so I eventually repaired the original.
The steps/mm is the iterative result is from a printing nuts and bolts sliced with prusa. It’s like a micrometer in that it measures the clearance between the printed nut and bolt threads. Measuring thread geometry is a good extrusion test and it quickly reveals temperature, layer height, filament brand, colour etc. all change the steps/mm. https://www.thingiverse.com/thing:3983705
The axis motor currents can be 450ma any lower does not improve the print finish and at 475ma they start making noise.
The extruder motor rotor is shorter and scaling the axis motor current you get a drive current of about 333ma. Any lower does not improve the print finish and higher adds an annoying minor washboard texture to the print surface.
I find that with 0.1mm layers the washboard surface texture increases while it is very hard to see when printing with 0.15 and thicker layers.
My 200W 24V silicone bed heater needed a lower duty cycle and higher PWM frequency to not trip the 120W Cetus AC adapter’s internal short detection circuit. M307 H0 B0 S0.45 F200 it heats to 55C at the same time as the nozzle get to 215C. The upper bed temperature is limited to the 80s and after a long warm up using the stock Cetus power supply.

All in all a very satisfying conversion.
-Peter