ABL Troubles

Printable View

06-30-2015, 07:32 PM
Renfro

ABL Troubles

So I got ABL installed.

Downloaded the stable firmware from github (https://github.com/MarlinFirmware/Marlin) and modified the heck out of it following some guides here and there.

I've also done the offsets calibration and checked the hardware. Still I have two problems which I believe are related.

1. The probe switch is hitting the bed because the arm is retracting before the Z axis raises.

2. I get inconsistent Z readings when running multiple G29

Code:

echo:SD card ok>>> g28 SENDING:G28 >>> g29 SENDING:G29 Bed x: 20.00 y: 20.00 z: 6.54 Bed x: 220.00 y: 20.00 z: 6.15 Bed x: 220.00 y: 220.00 z: 6.31 Bed x: 20.00 y: 220.00 z: 6.73 Eqn coefficients: a: -0.00 b: 0.00 d: 6.57 planeNormal x: 0.00 y: -0.00 z: 1.00 echo:endstops hit: Z:6.73 >>> g29 SENDING:G29 Bed x: 20.00 y: 20.00 z: 6.49 Bed x: 220.00 y: 20.00 z: 6.36 Bed x: 220.00 y: 220.00 z: 6.35 Bed x: 20.00 y: 220.00 z: 6.69 Eqn coefficients: a: -0.00 b: 0.00 d: 6.56 planeNormal x: 0.00 y: -0.00 z: 1.00 echo:endstops hit: Z:6.69 >>> g29 SENDING:G29 Bed x: 20.00 y: 20.00 z: 6.65 Bed x: 220.00 y: 20.00 z: 6.27 Bed x: 220.00 y: 220.00 z: 6.27 Bed x: 20.00 y: 220.00 z: 6.34 Eqn coefficients: a: -0.00 b: -0.00 d: 6.61 planeNormal x: 0.00 y: 0.00 z: 1.00 echo:endstops hit: Z:6.34 >>> g29 SENDING:G29 Bed x: 20.00 y: 20.00 z: 6.20 Bed x: 220.00 y: 20.00 z: 6.25 Bed x: 220.00 y: 220.00 z: 5.98 Bed x: 20.00 y: 220.00 z: 6.30 Eqn coefficients: a: -0.00 b: -0.00 d: 6.31 planeNormal x: 0.00 y: 0.00 z: 1.00 echo:endstops hit: Z:6.30 SENDING:G28 >>> g29 SENDING:G29 Bed x: 20.00 y: 20.00 z: 6.54 Bed x: 220.00 y: 20.00 z: 6.15 Bed x: 220.00 y: 220.00 z: 6.31 Bed x: 20.00 y: 220.00 z: 6.73 Eqn coefficients: a: -0.00 b: 0.00 d: 6.57 planeNormal x: 0.00 y: -0.00 z: 1.00 echo:endstops hit: Z:6.73 >>> g29 SENDING:G29 Bed x: 20.00 y: 20.00 z: 6.49 Bed x: 220.00 y: 20.00 z: 6.36 Bed x: 220.00 y: 220.00 z: 6.35 Bed x: 20.00 y: 220.00 z: 6.69 Eqn coefficients: a: -0.00 b: 0.00 d: 6.56 planeNormal x: 0.00 y: -0.00 z: 1.00 echo:endstops hit: Z:6.69 >>> g29 SENDING:G29 Bed x: 20.00 y: 20.00 z: 6.65 Bed x: 220.00 y: 20.00 z: 6.27 Bed x: 220.00 y: 220.00 z: 6.27 Bed x: 20.00 y: 220.00 z: 6.34 Eqn coefficients: a: -0.00 b: -0.00 d: 6.61 planeNormal x: 0.00 y: 0.00 z: 1.00 echo:endstops hit: Z:6.34 >>> g29 SENDING:G29 Bed x: 20.00 y: 20.00 z: 6.20 Bed x: 220.00 y: 20.00 z: 6.25 Bed x: 220.00 y: 220.00 z: 5.98 Bed x: 20.00 y: 220.00 z: 6.30 Eqn coefficients: a: -0.00 b: -0.00 d: 6.31 planeNormal x: 0.00 y: 0.00 z: 1.00 echo:endstops hit: Z:6.30

Does anyone know if I should migrate to the development version of the firmware. By what I was able to gather the "enhanced G29 command" was merged to the development branch around March so it's not part of what I have today.
06-30-2015, 10:31 PM
Roxy

OK... I'm sorry to tell you this, but right now the main branches at https://github.com/MarlinFirmware/Marlin are not ready for prime time. We can get into the reasons tomorrow if you care. But the right answer is to grab this branch of Marlin and put your settings into the Configuration.h file. If you have any problems, make another post with what is going wrong in response to this post.

This branch is optimized for MakerFarm printers. And it does not work with Delta printers. You have to be using a Cartesian coordinate machine. But with that said, this branch has *ALL* of the good stuff in it and it is stable. It works. It is not flakey. I'm running it. And on top of that, a lot of people here are familiar with it and running it. There will be lots of people to help you get it going.

https://github.com/beckdac/Marlin

Assuming you don't have a MakerFarm printer, just start putting the right stuff into the Configuration.h file and see how well you can get it configured. If you do have a MakerFarm printer, you just load and go and we will never hear from you again! :) (Actually... If you have a MakerFarm printer, you will load and go... And most likely we hear back from you once to say "It works great!")
07-01-2015, 04:30 PM
Renfro

Thanks for replying Roxy!

I got a Makerfarm i3v 10" and the stable marlin from github.

Currently G29 is scraping the bed with the probe so I wonder if there is any way I could merge your enhanced G29 to my files instead of deploying the development branch.
07-02-2015, 01:36 AM
Roxy

Quote:

Originally Posted by Renfro

Thanks for replying Roxy!

I got a Makerfarm i3v 10" and the stable marlin from github.

Currently G29 is scraping the bed with the probe so I wonder if there is any way I could merge your enhanced G29 to my files instead of deploying the development branch.

The answer is yes. Except the work has already been done. Download, compile and load the BeckDac fork listed up above. It's going to run the first time. The only thing you will have to exactly fine tune is the Z probe offset from the extruder. Really... It is set up for MakerFarm printers and it has all of the good stuff merged in.
07-02-2015, 02:35 PM
Renfro

Sorry Roxy for not realizing this sooner. I'm usually not that stupid!

BeckDac downloaded and flashed and as you said MUCH BETTER NOW!

Code:

Bed Level Correction Matrix:1.000000 0.000000 -0.000967 0.000001 1.000000 0.000885 0.000967 -0.000885 0.999999 echo:endstops hit: Z:1.96 Bed Level Correction Matrix: 1.000000 0.000000 -0.001052 0.000000 1.000000 0.000359 0.001052 -0.000359 0.999999 echo:endstops hit: Z:1.97

G29 probes all 9 points without retracting the probe between each reading and when it does retract (after it's finished), it behaves nicely raising the Z axis before engaging the servo. So no hits.

Unfortunately it does go back to the same bad manners when running verbose mode. It retracts the servo at the reading Z height, thus hitting the bed each and every time. This results in lots of fluctuations in the following readings.

Is there a way to stop it from doing that?

Thanks Again!
07-02-2015, 07:17 PM
Roxy

Quote:

Originally Posted by Renfro

Unfortunately it does go back to the same bad manners when running verbose mode. It retracts the servo at the reading Z height, thus hitting the bed each and every time. This results in lots of fluctuations in the following readings.

Is there a way to stop it from doing that?

It should not be behaving differently just because you tell it to be Verbose. I'll try to take a look at it this weekend. But in the mean time, you shouldn't need Verbose mode if everything is working.
07-03-2015, 05:08 PM
Carrot_or_Stick

Hey Roxy,

Just wanted to say I tried out the BeckDac fork you linked to and it works wonderfully. ABL is working great with my non-capacitive sensor and all my troubles have disappeared. Thanks for such helpful info!