Every time I print, I need to increase baby steps

[TimsWorkshop]

@fcwilt very clean setup mate, very clean.

Thank you, I'm going to digest and reverse engineer some of this and do some level testing on my printer...

And with that, I've done two test already using the file @Phaedrux mentioned and both times I still had to adjust the baby steps... First print was set to -0.25, second print was -0.38. I decided to re-run the manual bed levelling and mesh, at which point I think I will re-do my probe z-offset calibration and then the MBL and ABL again.

[TimsWorkshop]

@fcwilt So I'm doing the Z-Probe config again and running into the same problem I always have been, and wondering if this is why I'm always having to adjust my baby steps, for every print.

According to the article, I should run G30 S-1 2-3 times until I get a consistent reading, and then use that as my offset in the G31 setting. BUT my problem is, I never get a consistent reading and in fact, it increases every time I jog the Z by 5mm in rapid succession.... but if I wait a minute or 90 seconds, it decreased or sometimes has a larger/smaller increment.
Is this indicative of a different problem?

Here's what I've gotten to far (read bottom to top):

01/09/2021, 09:08:39	G30 S-1 Stopped at height 1.339 mm
01/09/2021, 09:08:22	G30 S-1 Stopped at height 1.327 mm
01/09/2021, 09:08:02	G30 S-1 Stopped at height 1.326 mm
01/09/2021, 09:07:42	G30 S-1 Stopped at height 1.325 mm
01/09/2021, 09:06:51	G30 S-1 Stopped at height 1.364 mm
01/09/2021, 09:06:42	G30 S-1 Stopped at height 1.383 mm
01/09/2021, 09:06:21	G30 S-1 Stopped at height 1.381 mm
01/09/2021, 09:05:52	G30 S-1 Stopped at height 1.370 mm
01/09/2021, 09:05:32	G30 S-1 Stopped at height 1.369 mm
01/09/2021, 09:05:05	G30 S-1 Stopped at height 1.367 mm
01/09/2021, 09:04:54	G30 S-1 Stopped at height 1.356 mm
01/09/2021, 09:04:29	G30 S-1 Stopped at height 1.355 mm
01/09/2021, 09:03:29	G30 S-1 Stopped at height 1.364 mm
01/09/2021, 09:02:34	G30 S-1 Stopped at height 1.383 mm
01/09/2021, 09:01:07	G30 S-1 Stopped at height 1.441 mm
01/09/2021, 08:58:29	G30 S-1 Stopped at height 1.410 mm
01/09/2021, 08:57:51	G30 S-1 Stopped at height 1.409 mm
01/09/2021, 08:57:16	G30 S-1 Stopped at height 1.398 mm
01/09/2021, 08:56:44	G30 S-1 Stopped at height 1.396 mm
01/09/2021, 08:56:10	G30 S-1 Stopped at height 1.375 mm
01/09/2021, 08:55:30	G30 S-1 Stopped at height 1.374 mm
01/09/2021, 08:55:19	G30 S-1 Stopped at height 1.352 mm
01/09/2021, 08:55:05	G30 S-1 Stopped at height 1.341 mm

[Phaedrux]

It looks like your probe isn't very consistent. Could try reducing dive speed and adding A10 S0.03 to the M558 command.

If you need to baby step down by approximately 0.3 each time, just increase your G31 Z by that amount and see how that goes.

[TimsWorkshop]

@phaedrux thanks. I've updated M558 to M558 P5 C"!zprobe.in" H2 A10 S0.03 F650 T12000 and getting smaller changes when I run the commands in rapid succession, but still inconsistent. But when I wait 60 - 90 seconds, the trigger height has a large change...

01/09/2021, 09:21:48	G30 S-1 Stopped at height 1.305 mm
01/09/2021, 09:19:58	G30 S-1 Stopped at height 1.344 mm
01/09/2021, 09:19:48	G30 S-1 Stopped at height 1.342 mm
01/09/2021, 09:19:41	G30 S-1 Stopped at height 1.351 mm
01/09/2021, 09:19:34	G30 S-1 Stopped at height 1.350 mm
01/09/2021, 09:19:25	G30 S-1 Stopped at height 1.349 mm
01/09/2021, 09:19:16	G30 S-1 Stopped at height 1.347 mm
01/09/2021, 09:19:09	G30 S-1 Stopped at height 1.336 mm
01/09/2021, 09:18:58	G30 S-1 Stopped at height 1.335 mm

[Phaedrux]

@infidelprops said in Every time I print, I need to increase baby steps:

F650

Try half that speed and see if it's better.

[TimsWorkshop]

@phaedrux yeah, I've been playing with that and have tried numbers between 325 and 585 and when done in rapid succession, it increments by 0.001 - when I wait 30+ seconds, it decreases by roughly 0.008

Just waited 7 minutes and it was a decrease of .03

[TimsWorkshop]

So been reading other threads and kind of sounds like people with sensors like the ABL have had problems with heat/magnetic beds throwing off the calibration and have said going back to a BLT solved their problems.

[zapta]

@infidelprops said in Every time I print, I need to increase baby steps:

people with sensors like the ABL have had problems

@infidelprops, Voron uses the inductive probe for relative leveling between various bed points and and fixed switch for absolute Z homing of nozzle. It's not perfect (e.g. it ignores variations in the PEI thickness) but is overall stable despite the high chamber temperature.

[TimsWorkshop]

So I'm still at a loss here... After all this work and getting MBL and ABL working and nice even surface, the problem still exists.

Printing the same file, the same filament, the same printer, etc...
First print, adjusted baby steps by -0.015
Second print, had to adjust baby steps by -0.027
Third print, had to adjust baby steps by -0.02

I've also noticed, that consistently, when facing the printer.. the right side of my pint seems flatter and adheres better where the left side appears a bit higher and adheres worse.

I've confirmed that the gcode is setting the mesh on line 18:

;FLAVOR:RepRap
;TIME:340
;Filament used: 0.205003m
;Layer height: 0.2
;MINX:59.1
;MINY:65.7
;MINZ:0.2
;MAXX:280.9
;MAXY:288.3
;MAXZ:0.2
;Generated with Cura_SteamEngine 4.10.0
T0
M190 S60
M104 S200
M109 S200
M82 ;absolute extrusion mode
G28 ;Home
G29 S1

And here's what I have in Cura for my start code:

G28 ;Home
G29 S1; Load Mesh

G92 E0 ;Reset Extruder
G1 Z2.0 F3000 ;Move Z Axis up
G1 X1.1 Y1 Z0.28 F5000.0 ;Move to start position
G1 X1.1 Y200.0 Z0.28 F1500.0 E15 ;Draw the first line
G1 X1.4 Y200.0 Z0.28 F5000.0 ;Move to side a little
G1 X1.4 Y20 Z0.28 F1500.0 E30 ;Draw the second line
G92 E0 ;Reset Extruder
G1 Z2.0 F3000 ;Move Z Axis up

[fcwilt]

@infidelprops

What is material is your bed made out of?

How thick?

How long do you heat it before trying to print?

Do you create your height map with the extruder and bed at temp?

Thanks.

Frederick

[TimsWorkshop]

@fcwilt

It's a TH3D flex plate, magnetic - Not sure of how thick it is
I let it get up to temp and then the print starts shortly after
I created the height map with both bed/hotend at temp.

Interesting enough.... I just did a 4th print and did not need to adjust the baby steps....

Going to wait 20 minutes and run another test print and see how it goes

[fcwilt]

@infidelprops said in Every time I print, I need to increase baby steps:

@fcwilt

It's a TH3D flex plate, magnetic - Not sure of how thick it is
I let it get up to temp and then the print starts shortly after
I created the height map with both bed/hotend at temp.

Interesting enough.... I just did a 4th print and did not need to adjust the baby steps....

Going to wait 20 minutes and run another test print and see how it goes

Well you must have something supporting that plate.

For example I use 6mm thick cast aluminum tool plate.

Frederick

[zapta]

@infidelprops said in Every time I print, I need to increase baby steps:

First print, adjusted baby steps by -0.015
Second print, had to adjust baby steps by -0.027
Third print, had to adjust baby steps by -0.02

I wonder if

@infidelprops said in Every time I print, I need to increase baby steps:

First print, adjusted baby steps by -0.015
Second print, had to adjust baby steps by -0.027
Third print, had to adjust baby steps by -0.02

I wonder if this widget, which can count Z stepper steps, would help diagnosing this problem. The expectation is that two consecutive prints should have the same step count at the same stage of the print.

https://github.com/zapta/simple_stepper_motor_analyzer

[Phaedrux]

@infidelprops said in Every time I print, I need to increase baby steps:

First print, adjusted baby steps by -0.015
Second print, had to adjust baby steps by -0.027
Third print, had to adjust baby steps by -0.02

Seems to match the variance of your probe results.

Do a couple point tests around the surface of your bed with the mesh enabled, and carefully move the nozzle to Z0 and see if it's accurate or not.

Just as a test to eliminate other factors, change to manual probing by using M558 P0 This will pop up a jogging prompt whenever a G30 is done allowing you to manually move the nozzle to just touch the bed surface. You only really need to do it for setting Z0 before the print.

This removes the probe from the equation. Though if the heightmap is inaccurate from the first test, then you might need to create a new heightmap with maybe only a 5x5 grid to keep the number of points to manually probe reasonable. But if it works and gives you a good first layer, then you know the probe is the problem.

[dc42]

@infidelprops said in Every time I print, I need to increase baby steps:

I've also noticed, that consistently, when facing the printer.. the right side of my pint seems flatter and adheres better where the left side appears a bit higher and adheres worse.

That usually means that the probe trigger height relative to the nozzle tip is varying across the bed. You can check by using feeler gauges to measure the nozzle height when the probe triggers. The usual cause of that using an offset Z probe and the print head tilting, for example because the axis is sagging or the bed plate is bowed.

[TimsWorkshop]

So... I've ran multiple layer test over the last few hours and I haven't had to adjust the babysteps at all and all layers are sticking well... Not sure what changed as I've done so much.

I also just kicked off a new print for a prototype and it was LH 0.28, and here as well I did not need to change the babystep, first layer went down beautiful.

I'm going to keep an eye on it, but I'm bookmarking this thread as a lot of good troubleshooting steps in here for later.

@fcwilt & @Phaedrux - thank you both for all the help! If I can buy you a coffee/beer, just let me know where to donate to.
@jens55 & @zapta - also thank you for your input, very useful as well.

[TimsWorkshop]

@dc42 Yeah, that sounds like a probable cause indeed. Once this current print finishes, I'll take a look at that as well. Thanks!

[fcwilt]

@infidelprops

Glad to hear things are looking good.

Frederick