Posts made by sverreb

I got a message on heater channel X201 that I had a short. After some debugging I isolated the fault to the daughterboard itself:
I repeated the fault with a brand new PT100 and new wiring
The fault cleared when using the X200 channel.

How precisely is the fault identified as a short by the system. Are there other problems that can give a false short indication?

Using a multimeter over the pins of the daughterboard of the failing channel measures ~100Mohm without the heater (around 100 with) which is what I would expect.

Is there anything I can do to further isolate and possibly fix the problem or will I need a new daughterboard?

Maybe this is a bit unfair as I just tested cura right now for the first time (because of this thread) coming from s3d.

Parts layout is not very nice. I could not find any way to drop the part to the bed and had to resort to rotating and translating manually. This is way to clumsy. A basic drop face to bed needs to be implemented. Auto arrange placed parts outside the build volume

In the gcode preview I accidentally moved a model which wiped out the gcode I was reviewing.

No manual supports (yes I knew that, this is why I really didn't bother with it until today), and I do not se that i have any additional control over supports to make up for it. I use manual supports in 90% of my models (Most of the rest need no supports). I had to go hunting in the menus to actually get any semblance of controls to be visible. The usual problem I use manual support to solve is that I have parts with gaps that are too cramped to be able to extract supports from but can be bridged or can tolerate a little drooping, while I have other areas that needs the support. Hence the lack of control of the support is pretty much a dealbreaker.

I could not find any way to save the print settings changes I did as a profile. I guess there is a way, but it is not obvious. Clicking the create button in manage materials (which is where I thought it would be) does not appear to do anything.

You might want to list some of the settings you have tried.

From the config.g you posted my immediate reaction is that it seems to be a low amount of pressure advance for the length of bowden cable you use.

Did you try a little negative extra restart distance?

As a starting point I would likely begin with ~4mm retract, 80mm/s retract speed 0.2 pressure advance, no wipe (unless it is PETG), no coast, no extra restart.

@dc42:

To those who have experienced this issue:

1. Which Duet are you using?

2. Do you also have a DueX2 or DueX5 connected?

3. Which firmware version(s) were you using when the problem occurred?

1: Duet Wifi
2: Yes, duex5
3: 1.19

From the reply I gather you suspect some interrupt storm/high traffic density from the wifi module. I do have access to a MSO scope so if there any test point relevant to the issue I might try to monitor them (If it works out logistically)

I can confirm 1 and 2
I'll do 3-5 once I can reproduce

I have a reliability issue with my duet wifi/duex setup. Occationally and in the middle of a print the printer slows down dramatically executing a couple of cm of moves then pauses for several seconds. Even when I pause the print movement to homne is really slow. Other functions are also slow, web interface is barely responsive. Jog controls on the panel due are unresponsive or extremely slow.

Obviously this slowdown causes a failed print.

The only remedy is to reboot, cold or warm reset does not matter.
This is on 1.19

VIdeos and picture: https://photos.app.goo.gl/6nvOED4R3x1qW2hi2

In my printer profile for PETG I am adding some z offset by using baby stepping:

[[gcode]]
G28 ; 
G29 S1 ; Load and activate detailed bed map
G1 Y1.0 Z0.2 F1000.0 ; prepare to prime
G92 E0 ; reset extrusion distance
G1 X60.0 E9.0  F1000.0 ; priming
G1 X100.0 E12.5  F1000.0 ; priming
M290 S0.20 ; set baby stepping

This works perfectly the first time, however if I make a second print (which uses the same starting script) I see the printer using 0.4mm baby stepping offset.
The documentation states that homing should reset baby stepping but in my case it does not appear to do so.

I see a very uneven motion on high speed XY moves (travel moves) At first I thought I was getting lost steps but a test print with three objects at far ends of the platform prove otherwise.

What I observe is that long travel moves appear to be split into 2-4 segments where the carriage significantly slow down then speed up again between the segments.

It is also observable when using the panel due movement controls so it would not be segments declared in gcode

These are my speeds settings:

[[language]]
M350 X16 Y16 Z16 E16 I1 ; Configure microstepping
M92 X80 Y80 Z1600 E2755 ; Set steps per mm
M566 X1200 Y1200 Z12 E40 ; Set maximum instantaneous speed changes (mm/min)
M203 X18000 Y18000 Z400 E2400 ; Set maximum speeds (mm/min)
M201 X1500 Y1500 Z250 E120 ; Set accelerations (mm/s^2)
M906 X1500 Y1500 Z800 E650 I30 ; Set motor currents (mA) and motor idle factor in per cent
M84 S30 ; Set idle timeout

@DjDemonD:

Both the flex3drive and nimble require to use low motor current, the motor can't make a fast direction change at high current as quickly. I use 500mA with nimble and a 34mm motor. A 20mm pancake motor or nema 14 would also work. Acceleration needs to be low 120 to facilitate the direction changes. Jerk at around 40. More motor current is a red herring.

In my experiments with a nimble and a pancake stepper I had to go to 600mA. 500mA and less the stepper stalled on high extrusion speeds. (~12-16mm^3 on a 0.6mm volcano nozzle).

Printers may experience a slight tilt of their carriage for their movement. Tilts that are small enough that they do not matter for actual printing may still cause problems during bed probing as the probe is often significantly offset from the pivot center (where the hot-end tend to be) This causes errors when probing the bed in which probes done when travel approaches from one side along the sensive axis vs. the other creating a bed height map with ripples that do not correspond to an actual imperfection in the bed.

To minimize this the carriage should ideally approach each probe point from the same direction on both axis during probing.

As a simpler but mostly good enough solution it is suggested to allow the user to select if the g29 command traverses X first or Y. While not a perfect solution it does solve any situation there tilting occurs on only one axis

Understood.
Now assuming that most printers will usually only be vulnerable to this on one axis just providing the option of y first or x first is likely good enough for almost all cases.

I can observe a slight tilt on my x carriage during g29 probing by that the probe consistently measures 0.07mm less when moving to the probe point from the right than when it moves to the probe point from the left.

Now ideally this should be removed from the mechanics, but in reallity it will be hard to ensure that there is no such tilt at all and for printing it does not really matter as the nozzle unlike the probe is on the carriage centerline so the nozzle does not change height it only get a very very slight angle shift, so redesigning mechanics for this is a bit overkill.

The actual problem is really only during g29 probing of the bed as the carriage approaches each probe point from alternating directions so automatic bed compensation thinks my bed is wavy with a period that matches 2x the probe pitch.

For me this is only a problem on the x axis, but for a general soulution I guess it can also occur on the y axis.

I would like to see some way to adjust how the bed is probed instead of the default zigzag pattern with x low to high, y, then x high to low. Possible additional patterns could be
x low to high, y, then x low to high again (making sure to approach the first x point from the low side by overshooting the return then coming back)
y low to high, x, then y high to low
y low to high x then y low to high

Is there a way to achieve this now? If not could this be interesting to add in the future?

Your steps/mm should be analytically known. This is given by tooth pitch or leadscrew pitch. Scaling this linearly to hit a single point error is not the right thing to do.

In all likelyhood your calibration cube deviation is not due to a linear deviation from the analytical steps/mm. (Hard to imagine any mechanism that would give rise to such an error)

More likely sources for the error is overextrusion or artifacts from slicing. If you are satisfied you are extruding correctly try to look for horizontal compensation settings in your slicer. This will take off some fixed amount from all dimensions, and not scaled to the dimensions size. (I.e. a -0.2mm adjustment scales a 20mm cylynder to 19.6mm and a 40m cylynder to 39.6mm)

If you want to verify to yourself that you do/do not have a linear error try to slice your cube at 100% and 200% scale and print with the original 80 steps/mm setting. The likely result is that both cubes will be off by the same amount (or close to it) which indicates that a linear compensation is wrong. If however you find a 40mm cube is 2mm off while the 20mm cube is 1mm off, then you should consider scaling, but as I stated, this is unlikely.

I am reasonably certain I will not have a fire hazard regardless. The heating element have a 185C thermal fuse built in.
The 3m 200MP adhesive (which seems to be the ubiquitous choice) is rated to 149C long term.

I do intend to add insulation. I am planning to tape some superwool HT (refractory insulation, non flammable) to the backside using high-temperature aluminum tape.

The steel plate is quite thin, so it does not add that much thermal impedance. The magbase is likely the worst offender. The topside including magnets are good to 125C (limited by the polymer coating on top). The bottom side is good to 150C (When the glue starts to break down).
I tend to see 10-20C difference between the topside and bottom during heating so this is a bit too close for comfort if I want to reach 120C bed temperature. I also haven't insulated the bottom yet which may momentarily increase this difference. This is why I'd ideally would want to have the bottom sensor in the regulating loop as well to I could ensure that the heater never exceeds 150C at the price of slowing down the heating a little.

The plate plus various things on top (magbase, spring-steel plate, and build plate coating) is approximately 8mm

I have a relatively thick print bed which means there is a high time constant between heat input at the bottom of the bed until the top responds. I want to regulate the heat to the topside temperature using a probe inserted into the bed near the top for better accuracy, but then I risk having the heater overshoot the temperature by so much it becomes a safety risk and/or a regulating problem.

Is there any way I can regulate a heater to two temperature sensors. Where one is the setpoint and the other is for safety? (I.e. state that the setpoint is to get the top sensor to 120C, but never exceed 150C on the heater element)

I guess as a primitive solution I could set a thermostatic fan to the safety sensor then wire the output from the fanheader to a spare endstop and use that as an e-stop or emergency heater-off event, but this seems a bit clunky and limited, so I am wondering if I am missing some better way?

I was considering wiring both X and Y limit switches with both Vcc and GND. I.e the switch drives the output to GND when inactive (Not pressed) and Vcc when active. This is to improve noise immunity as the wiring is close to stepper and heater wires.

The switches are simple 3-pin SPDT switches (ZW50) so there are no series resistance to speak of.

Are there any issue in the duet wifi endstop wiring that would preclude this wiring?

I think you are right. I tried moving the leadscrew by hand and only got 20% of the movement on the bed. Likely the bed resists torque around the x axis enough that the motor mounts (which are printed) deflects more than the bed rotates.