Best posts made by zemlin

There are a lot of virtual knobs to turn when getting a machine dialed in. Is there consensus or an expert opinion on what to tackle first? There are so many interactions between speed, accel, jerk, PA, DA, X, Y. Z, & E. Also, when starting to adjust the first variable, how do you have the others set?

I'm not having any significant issues, but it seems there could be an endless cycle of test prints and tweaks as one sorts out the interactions. I just feel like I'm probably not taking the most efficient approach.

Opinions/Advice/Links to existing threads? TIA

I have replaced the control boards in my 3 printers with Duet boards, and it's the best thing I ever did to these machines. When I've had issues I've been able to post specific questions here on the forum and have always received prompt help that got me going.

Take a deep breath and give us enough information to start getting you going.

The Duet Wifi board has been replaced with an Ethernet board, and the machine is now on a separate network that will not be accessible from the public wifi or publicly accessible computers.

Thanks.

Achieving dimensional accuracy isn’t as simple as a single scale measurement. I use S3D for slicing and it has X, Y, Z part scaling plus a horizontal offset. I don’t know IdeaMaker well, but I’m guessing X and Y-axis compensations in the Printer Settings dialog work for XY scale. Part Scaling can also work, although I don’t know why they have chosen Width Depth and Height over X, Y, and Z. Under the OTHER tab, XY size compensation may be the same as S3D’s Horizontal Offset.

Attached you will find a part I'm calling a Calibration Cross - STL and JPG drawing. It includes internal and external features, both large and small. I have also attached an Excel Spreadsheet (Zipped) which allows you (by trial and error) to see the impact of scaling and horizontal offset values in correcting part dimensions.

Print off a Calibration Cross with no scaling and no horizontal offset. Enter the measured values into the spreadsheet and experiment with different X & Y scale factors as well as the horizontal offset value to minimize the resulting error on all of the measured dimensions. The solver add-in can help narrow in on the best values.

Whenever I’m going to print using a new setup, filament, significantly different print speed, layer height, temperature, etc, I start by running a Calibration Cross and use the spreadsheet to calculate scaling and offset numbers. This technique has proven to be a reliable way to ensure dimensional accuracy on parts. The Z-scale on my N2+ has proven to be quite accurate, so I just use 100% on that.

I can generally achieve part accuracy better than +/- .1mm (.004") on all dimensions.

calibration-cross.xlsx
calibration-cross.stl
100mm.stl

@phaedrux
Good call - Homing has an A1. Will change that. Thanks.

update: Editing the M588 command in the homeall and homez files resolved the issue. Thanks.

@Phaedrux

M98 P"config.g"
HTTP is enabled on port 80
FTP is disabled
TELNET is disabled
Warning: Heater 1 appears to be over-powered. If left on at full power, its temperature is predicted to reach 687C

I haven't tuned heater 1.

DISCLAIMER - I haven't red the DWC documentation carefully - if the answer to my post is RTFM, I'm OK with that - and I did search first but didn't find anything. I doubt I'm the first person to raise an issue.

I use an iPad (Gen 4 retina display) at my printers - landscape mode. I also pop the printer interface up on my phone pretty often.

I just updated one of my printers to V3 today. I'm disappointed with DWC on the smaller displays.

Here's a comparison of my iPad display:
DWC V2

DWC V3

On the new DWC all the elements are a lot bigger and it just feels cramped. The machine movement buttons were the first thing that really got my attention. When setting the Z-Offset in a new nozzle, the Z movements are 5mm and .05mm. On the older layout there's more room for buttons and there's a .5mm. I wrote a macro for 1mm so I could get close enough to use the .05mm.

I have to scroll to access macros - scroll to access extruder control.

The display on my iPhone is also a lot more drawn out with less room for stuff because all the elements are larger.

Would like to have a setting to scale things down so more information can fit on the display.

Thanks.

Probing speed also impacts accuracy. I was getting good results with 2mm/s probe, but slowed it down to 1mm/s for good measure - reduce the probing height to help recover the time.

I really appreciate all the help.
I changed the can lines per @DIY-O-Sphere to

M950 F1 C"fan2" Q500 ; create fan 2 on pin fan2 and set its frequency
M106 P1 S1 H1 T40 ; set fan 2 value. Thermostatic control is turned on

Functioning properly and not showing up on the dashboard slider.
The Z-Probe on this machine is accelerometer-based and that's a little flaky, but doesn't seem to be software related.

Up and running now, so 'til next time ...

@rechrtb Nothing has blanked out in since I installed the beta on all three machines.
Thank you ... seems to be solved.

Have you tried printing with that height map? How does your first layer look?

Just an FYI to, perhaps, close this out. When I upgraded my machines to V3 firmware I tried enabling both extruder thermocouples. Both sensors came up without issue. I don't expect this to be related to firmware, but it was stable for quite some time. I've had other issues/adventures on the thermcouple front, but not related to this issue.

I've been working with these sensors on my Duet2 Wifi - thought I'd share my experience. I thought it was just going to be a simple microswitch connection with a common, NC, NO - but when I started checking continuity on the leads in different states it was clear that it's not that simple.

I saw a review on Amazon that talked about connecting to a RASPI board with three leads - ground (black), voltage(red), and signal(blue).

I tried that first on the Duet board going into E1 endstop. There are LEDs in the sensor - green when filament is present, flashing red when not. The LED on the Duet board was changing state - low (LED ON) with filament present and high (LED OFF) with no filament.

Added M591 D0 P2 C4 S1 to my config.g.

Initially it looked great. Printer started printing and paused with filament out notification when I pulled the filament. At that point, however, I was unable to restart the print. I put filament back in the switch and pressed RESUME - the print head moved over the part and then reported NO FILAMENT again - pausing the print. The end stop LED on the board was still turning on and off without filament, but M591 D0 reported no filament regardless of the state of the switch or LED on the board.

The Red and Blue leads on the switch act like a direct NC connection, so I wired those to the E1 end stop ground and signal - just the outer connections. Changed the P2 to P1 in the config file and all is working fine. I lost the blinky lights in the sensor, but that's minor. A little mystified as to why the LED on the board was switching states with the original setup, but M591 would not report the changing state after the initial filament out warning.

With some help from the forum here I got one of my machines running 3.2 Beta4?

No change in the stuttering. Tried 566 P1, that doesn't have an impact either.

Update - found the 3.2.beta4.1, but since it was mentioned that anything after 3.2 would have the improved processing, not going to duplicate the tests unless you think it's worthwhile.

Yesterday I made up new leads for the BL-Touch using shielded cable, and I routed the lead along the filament feed tube rather than through the print head harness.

All issues have cleared up on both machines. Gone a whole day without bending a pin!!!

Thanks for your time and assistance @droftarts & @Phaedrux

Have a couple more issues to clear up on these machines, but those will be different threads.

karl

I know this is an old thread, but thought I'd drop my experiences in here.

I converted a Raise3D N2 and N2+ to Duet2 Wifi. I used a PT100 board for the bed heater, and the Raise printers use type K thermocouples in the print head.

I did not know there was an issue mounting the thermocouple board remote. I used 26ga ribbon cable and 10 pin ribbon cable connectors to make an extension cord. The PT100 board is on the Duet board. I mounted the thermocouple board up on the print head.

On the smaller N2 the ribbon cable is roughly 1.5M long. It's been running fine since June. At one point I had an issue with a thermocouple reporting 2000 degrees - I replaced the thermocouple and the error went away. The thermocouple did not appear to be bad when it was off the machine.

The larger N2+ has almost 2M of ribbon cable to the thermocouple board. I had persistent issues on that machine with the thermocouples faulting on power up and reporting 2000 degrees. I swapped some thermocouples around and got it so E0 was reporting OK, but not E1. Poking around the forum at that point was when I learned I had done everything wrong.

On the N2+ I moved the thermocouple board down to the control board and ran shielded, twisted, type K thermocouple wire up to the print head. That did not resolve the startup fault problems - but the machine was working and I had work to do.

I purchased more thermocouple wire so I could rewire the N2 if it started giving trouble. I haven't had the need.

After 6 months or so of heavy use, the N2+ started getting temperature spikes on E0 and faulting. I tried a different thermocouple, but that did not resolve the spikes. The thermocouple extension wire is foil shielded, so it may be that the foil split in the flex to the print head. Just speculation.

Since the smaller N2 has worked great for 9 months or so, I decided to try reverting the N2+ to the ribbon cable extension, knowing that it's not supposed to work.

When I got the ribbon cable re-routed and powered up the printer last night, the thermocouple that had been faulting reported 2000 degrees. I replaced it and that seemed to make things happy.

I've now been printing straight for about 24 hours and haven't seen a temperature spike.

I don't know enough about this sort of thing to make any claim about right or wrong, but it's working for me. I know I might just be lucky. (I don't usually consider myself lucky)

I recently upgraded my Raise3D N2 to a Duet 2 Wifi. I'm using both PT100 and thermocouple daughterboards. The PT100 board is for the bed sensor - could not use the stock sensor. The hot ends use thermocouples. I mounted the thermocouple board on the print head so I didn't need to run thermocouple wire down to the control board.
I added a BL Touch.

I kept the factory daughterboard on the print head, but I'm only using it for the heater terminal blocks. I didn't have a better option in the parts bin.

The machine has also been modified to have a heated enclosure on an independent temperature controller. The bed has been modified to to get the clips out of the build area and make it easier to use alternative bed materials. It has the bondtech dual extruder upgrade.

The print head has additional fans for improved cold-end cooling and controllable print cooling.

All the 3D printed parts in the photos have been replaced with new prints since the upgrade since I'm getting much better results from the machine with the Duet board.

Heaters and fans for the heated enclosure.

I also have the 12x12x24" N2+ - that will get the Duet treatment soon.

@dc42 said in Struggling with PT1000 calibration:

If you are sure of the accuracy of those resistors, then RRF is working as it should. Either your PT1000 sensor is inaccurate, or your thermocouple sensor is inaccurate, or you have not succeeded in getting those two sensors at exactly the same temperature.

I have far more faith in the accuracy of the resistors than the accuracy of the thermocouple readout. The 2000 ohm is a pair of 1K .01% resistors in series. The 1096 is a 1K and 100 ohm resistor - but I sorted through a number of 5% resistors to find a pair that added up to 1096. Checking these with a good Fluke digital meter. I have a couple more Slice PT1000 sensors on the way. If they are in line with the other, I won't sweat it. Having them consistent across several machines is the goal.

@Phaedrux No ... That was not the version of the plug I've been messing with. That one went in fine.

It's always the little things :-p. Thanks for the assist.

Going to throw out one more idea - I didn't read all the prior posts as the discussion was all over the place.

If the vertical frame members are resonating, it might be a pretty simple matter to kill the resonance by making them asymmetrical.

What if you were to take three 1 lb (or whatever) pieces of steel and some tee nuts - secure the steel weights to the legs at different heights. That will change the resonant frequencies, and if they are at different heights on each leg, that will kill the symmetry of the vibrations.

Can't promise that it will work out, but it seems like it might be worth trying.