Voron 2.4 R2 - From Ontario Canada HEY!

[PFarm]

Try to do a QGL, it does the first two and after probing the 4th corner does the adjustment on both bed probes, when it probes the left front corner on the third try it drags the nozzle across the bed going to the left rear corner!!! I'm I missing something here?

 

Edited January 24, 2023 by PFarm

[mvdveer]

What is the output on klipper? Does it give an error in the console e.g sample tolerance exceeded / Measuremnet increasing/ probe triggered before movement /or something like that?

Where on your bed is 0,0 Left front (normaly here)

You can also change horisontal_move_z to a larger value as an interim measure to see if that is an issue. (May be that by adjusting the gantry, 10mm movement is not enough for the adjustments)

[PFarm]

When in drag across the bed it will show triggered before movement.  Was "normally here" supposed to be a link? I'll give the Z hop a go and see. Z offset is here

Edited January 24, 2023 by PFarm

[mvdveer]

40 minutes ago, PFarm said:

"normally here"

No sorry, Normally on the bed X0y0 will be left front. (On a Voron X-endstop normall at X300/350+ Right side of bed, and Y-Endstop normally at Y300/350+ Back of the bed)

Z-Offset in that section you posted "Do not edit....." is after probing is done and is the valued stored from probe_calibrate. https://www.klipper3d.org/Probe_Calibrate.html

Home the printer, then probe_calibrate as an initial setup.  Then run QGL. Once the gantry is level, re-run probe calibrate again. You will see that value may change

It may also help in setting z_offset in your [probe] section to 0 initially

[PFarm]

Here what is happening: https://drive.google.com/file/d/1lKidYbnO7zz7BeoN_1q2nfGOc5ekOCwU/view?usp=sharing

The last two movements of the printer head are dragging on the bed and stop with the probe triggered before movement. I've installed Tap by the way. What is the proper position of the magnets on Tap?

Edited January 24, 2023 by PFarm

[mvdveer]

36 minutes ago, PFarm said:

Here what is happening: https://drive.google.com/file/d/1lKidYbnO7zz7BeoN_1q2nfGOc5ekOCwU/view?usp=sharing

The last two movements of the printer head are dragging on the bed and stop with the probe triggered before movement. I've installed Tap by the way. What is the proper position of the magnets on Tap?

It seems the toolhead is not lifting high enough off the bed after probe of the point on the bed

Try this:

In the probe section set samples to 3 (currently = 1)

Also if not already, set horizontal_move_z to 15 (currently = 10) We can change that back later

The magnets on the TAP needs to physically touch the two screws on the backplate of the TAP.

 

[PFarm]

When I do QGL with samples =3 I get Probe samples exceed, samples tolerance.  I do a probe accuracy I get a 0.474833 deviation 

 

It trends upwards.

 

Edited January 24, 2023 by PFarm

[mvdveer]

Change sample_tolerance to 0.01 or even 0.1 (currently 0.006) - just as an experiment

[PFarm]

Slowed speed to 5 and tolerance to .01 

 

 

[PFarm]

At .1 

[mvdveer]

I am at a loss - big variations which lead me to believe there may be an issue with the probe. You using TAP with PCB and optical, wired optical or UNTAP ?

 

[PFarm]

PCB optical. Could get the Klicky to pass the accuracy test either no matter what I did. I did the accuracy test heat the bed and nozzle and test is an hour long and probe looked good in that test. So I'm at a lost why I can't get it to work properly. Does it matter if the diode is still in the circuit it shouldn't it only used to prevent voltage feedback.

Edited January 24, 2023 by PFarm

[mvdveer]

Just had a look at your config again. For the Voron TAP, in the [probe] section the X and Y offset should be 0, as the nozzle is the probe.

Still searching and reading to find a solution

[PFarm]

ok thanks, I'll change that

[mvdveer]

Clutching at straws, but did you "break in the probe" as per the Voron Tap post install instructions?

 

Post-Install Setup

1. Update your printer.cfg as recommended in Tap Klipper Instructions  (IMPORTANT)
2. Home Z and test virtual Z endstop by lifting tool-head
3. Heat soak your machine and run a couple probe_accuracy samples=100 to "break-in" your probe
4. Run a few more probe_accuracy checks (default of 10 probes)

For well-built machines you can expect to see between 0.0000 and 0.0008 standard deviation.

 

[PFarm]

Yes I did all of that and like I said numbers where good. 

[PFarm]

@mvdveer I really appreciate the help you given me with trying to solve this issue. I'm at a lost as well gone thru all of the settings can't figure out why the probes won't pass the test, optical, Klicky and proximity all failed to have the proper tolerance and no info in my research on how to fix an unstable probe all three can't be defective! When I run the automated probe accuracy test for Tap all the numbers are good (seen some 0.0000) but can't understand why it won't pass doing the "Probe Accuracy Test". Do I attempt to do a clean installation on both Pi and Octopus?

 

Edited January 25, 2023 by PFarm

[mvdveer]

18 hours ago, PFarm said:

. Do I attempt to do a clean installation on both Pi and Octopus?

I don’t think that will make a difference. Have not given up yet. Still searching for an answer. We can go to the extreme and set the sample_tolrance in the [probe] section to 1, to see if you can finish a QGL

[PFarm]

I'm at work 2,500 km from home here till Feb 28th followed by retirement. Took advantage of having a few parts shipped to me to save a bit on shipping. Order some extruder bearings since the supplier had pressed them onto the gear and I damaged them by removing them. Had spares so sanded the shaft so they would slip on properly, Also ordered Hartk Stealthburner 2-part PCB kit.

 

 

 

Edited February 8, 2023 by PFarm

[smirk]

20 hours ago, PFarm said:

Feb 28th followed by retirement.

Party at yours!

[PFarm]

2 hours ago, smirk said:

Party at yours!

You bet you coming over! 

[PFarm]

Well, I'm home for good! Back to figuring out why I can't get the PROBE_CALIBRATE to pass an acceptable probe tolerance. When using TAP does the diode need to be removed and jumpered on the hot-end PCB?

Edited March 3, 2023 by PFarm

[mvdveer]

Don't know but have seen a lot of debate around that. To see if it is the PCB, make up an extension to wire TAP directly to the board. See if that makes a difference. I initially had problems with an Euclid probe and the PCB, ended up wiring this directly to the octopus and all was well. Euclid was replaced with TAP and there has been no issues. For me it was easier to run the wires directly with an extension (Plug a the toolhead so the toolhead can still be easily removed)

[claudermilk]

16 hours ago, PFarm said:

Well, I'm home for good! Back to figuring out why I can't get the PROBE_CALIBRATE to pass an acceptable probe tolerance. When using TAP does the diode need to be removed and jumpered on the hot-end PCB?

The BAT85? It should be ok. Keep in mind I couldn't get the PCB (v2.1 24V) to work at all, so I went to the wired sensor per the assembly manual. That is working fine with the BAT85 still in the wire harness.

[PFarm]

@mvdveer@claudermilk Thanks for the reply guys, my tap PCB is the 5V version, I've removed the BAT85 and jumpered the terminals from the hot-end PCB but still no improvement.

Tap setup indicates the should see 0.0000 to 0.0008

But this is what I get doing the probe-accuracy. Same deviation I got with the inductive probe and the clicky probe. Gone thru the printer.cfg setting and for the life of me can't see why I'm getting this result.

This is my Probe setting:

This is the result of the probe_calibrate

Edited March 3, 2023 by PFarm