Accuracy Problems with e3d PZ Probe

[GeorgFoc]

Good evening,
I need some help with my e3d pz pobe. I am building a voron 2.4 right now (it is my first 2.4, I have already an 0.2 running).

I decided to use the e3d revo pz probe. I habe an BTT SB2209 Toolhead Board (CAN MODE).
I have a BTT M8P Board with a raspberry Pi cm5.

I connected the Probe to the Probe Port, and it looked good at first trys. The accuracy was below the voron requirements. Sensorless Homeing of X and Y worked, and Z worked as well with the Probe. Also QGL worked.

So as a next step I tried to test the probe, when the printer is heated up to ABS Temperatures. I preheated the Hotend to 240 Degree Celsius and the Bed to 100 Celsius. And then give it 30 Minutes to heatsoak. Than I tried the Accuracy test again. And it was not good.

probe accuracy results: maximum 6.607500, minimum -0.527500, range 7.135000, average 0.641000, median 0.142500, standard deviation 2.008853

Than I start testing the other Preseds from here:

PZ Probe Support: PZ Probe

Table of Contents Warning Features Setup Guide Troubleshooting Warning Make sure to read this guide in its entirety before using PZ Probe. This product is rated to a maximum of 5V. Exceeding this voltage can lead to serious complications! This product is not plug and play. Damage to the printing...

e3d-online.com

But this made it worse. False triggers, and I assumer sometimes to much force.

My biggest Problem is, that I am now not sure which preset is active. I removed UART and I2C Cables from the pz cable, because I assume I will not need them. Maybe a mistake?

I assume that preset 3 is the way to tell which preset is selected? This seems the only preset where Z Endstop and Probe are both always triggered. So I assume I am at Preset 0 right now. But this is only a assumption.

I disasembled and recheckt the SB 3 Times, to make sure it is rigid.

And when the printer is cold, the accuracy looks fine for me.

I also reduced the motor current for probing. Here is my Probe Config.

I will be really thankfull for tips, to get the pz probe working at ABS Temp.
```
[probe] ## Other parameters in config.cfg
pin: ^!EBBCan:gpio22
x_offset: 0.0
y_offset: 0.0
z_offset: 0.0
speed: 3.0 # Speed (in mm/s) of the Z axis when probing. The default is 5mm/s.
samples: 5
sample_retract_dist: 3.0
samples_tolerance_retries: 1
lift_speed: 10
activate_gcode:
G4 P200 ;Wait 200ms
SET_TMC_CURRENT STEPPER=stepper_z CURRENT=0.25
SET_TMC_CURRENT STEPPER=stepper_z1 CURRENT=0.25
SET_TMC_CURRENT STEPPER=stepper_z2 CURRENT=0.25
SET_TMC_CURRENT STEPPER=stepper_z3 CURRENT=0.25
deactivate_gcode:
{% set run_current = printer.configfile.config['tmc2209 stepper_z'].run_current | float %}
SET_TMC_CURRENT STEPPER=stepper_z CURRENT={run_current}
SET_TMC_CURRENT STEPPER=stepper_z1 CURRENT={run_current}
SET_TMC_CURRENT STEPPER=stepper_z2 CURRENT={run_current}
SET_TMC_CURRENT STEPPER=stepper_z3 CURRENT={run_current}
```
The default config had a current of 0.4

I am running out of ideas. So I would be really Gracefull for Tips.

If the PZ Probe is garbage I have to rethink my setup. I liked the idea, because I want to swap nozzles a lot and thought with the pz probe, I don't have to reset the z offset every time I switch the nozzle.

But I also like the idea to have an eddy current probe, but I think they cannot be used for this nozzle swapping feature? And did I understand it correct, that currently I cannot use 2 Probes in Klipper. So If I want to use somethin like an eddy probe for bed mesh, and pz probe for Z Offset, I cannot do this right now.

 

[GeorgFoc]

So a short update.
I replaced the X CARRIAGE with a newly printed one. My current assumption is that the heated inserts were not correct inserted, or some other stability issues.
After replacing it, the accuracy with cold bed and cold nozzle is: 0.001

So i have to go to work now, but tonight I will test the heated situation. Maybe the ABS i used is not termal resistant enough

 

[GeorgFoc]

Looks like I fixed it. For now. QGL on 100 Degree Celisius Board and Nozzle at 240 is running now.

What I did was as mentioned reprinting th e carriage and tripple checked all screws and the belt tension.

I hope this was it.

Maybe this helps someone.

I would be still interresting how I can check with preset is active on the pz probe.

 

[GeorgFoc]

Looks lke a was a little bit to confortable.

It looked quite good. But now again, i got the error while QGL

"
Probe samples exceed samples_tolerance
"

I retried QGL and it worked.

Than I got the Error while doing the Bed Mesh.

I tested the probe accuracy immediatly after the Bed Mesh and got:
"
probe accuracy results: maximum -0.086717, minimum -0.091717, range 0.005000, average -0.089217, median -0.089217, standard deviation 0.001118
"

So I am still not sure, what is the reason. Will check the Z Belt Tension again.

 

[GeorgFoc]

Some other theories I have.
- Maybe my ABS softer?
- The PZ Probe is really not for me, and I should by an carthographer probe? Does the carthographer works with Nozzle Swapping?

No I got these Values:

probe accuracy results: maximum 1.500000, minimum -0.222500, range 1.722500, average 0.230000, median -0.070000, standard deviation 0.627490


I am quite depressed. So for all recommendations or tips, I would be really gratefull. I will now go for walk, to get my Brain working again.

 

[GeorgFoc]

This will be kind of a logbook for myself, or maybe for someone who search for similar challenges. I am learning a lot. I have to say, even when this situation is a little annoying, I really love the voron project. It improved a lot of my skills so far, and still a lot to go.

After tweaking an playing around, it looked much better. I reassembled the heater bed. And it looked quite promising. Until it startet again. After some time in the heated chamber the accuracy collaps.

So i again disassembled the stealthburner and noticed that one head insert is lose in the X Carriage. So I will reprint it again, and tripple check correct heat insert insertion.

But something in my brain ask myself, if ABS is the right material for a pz probe?

Maybe I will rethink this.

Enough for today. I will start the print of the x carriage, and let the printer cool down.

 

[GeorgFoc]

Ok I give up for now on the pz probe. I have, from other projects the parts for tap and for an eddy sensor. I will start with tap for now. Maybe I will use the pz probe on my voron 0.2

But for now, I am not happy with pz probe. Maybe a CNC Carriage improve the situation.

 

[claudermilk]

This has been a useful thread to read to keep my curiosity on the pz probe under control. Sticking with the dead-reliable Tap for now.

 

[GeorgFoc]

Besides looking for help, this was also my idea. To share my strugles with this tool. With Eddy everythings went well until now.
I also contacted the e3d support, but it was not really helpful. I really assume, that you need a really ridgig system (CNC?) for the pz probe. Or maybe when you go deeper into the filter and programming things. But with Eddy, I am quite happy.