Still having problems with bed compensation with the dc42 probe

darookee

I just realigned Z=0, did another G29 and printed the same thing as usual:

Still too high on the right and too low on the left

The heightmap can be found here: http://imgur.com/AQMXnUt, it looks similar to the one before…

darookee

I tried another test to confirm it should be working. I homed, loaded the compensation map and moved the head to Z0.05 X0 Y100, then as fast as possible to Z0.05 X200 Y100 and it stopped along the way to adjust the height.
Then I tried the same thing with the sheet of paper I use to set Z=0, it should have the same drag on X0 and X200, but it does not. The paper is stuck on X0 and more loose than in the center on X200.

InSanity

Did you try what I said about M208 ? See above.

Jeff

dc42

It could be that you have a difference in the sensor trigger height at the two ends of the bed. Try measuring it. If you do find there is a difference, it could be caused by head tilt. So measure the height of the bottom edge of the sensor above the bed using feeler gauges, when the nozzle is touching the bed, and check that height is the same with the print head at both ends of the bed.

darookee

@dc42:

It could be that you have a difference in the sensor trigger height at the two ends of the bed. Try measuring it. If you do find there is a difference, it could be caused by head tilt. So measure the height of the bottom edge of the sensor above the bed using feeler gauges, when the nozzle is touching the bed, and check that height is the same with the print head at both ends of the bed.

I checked this using the macro below. First I just did G30 S-1 and measured the distance from the probe mount to the bed with a caliper and got these values for the points:

39,09mm
39,24mm
39,16mm
39,11mm
39,22mm

Then I found out what a feeler gauge is and that I do have one, but I couldn't get the single feelers out from the case-thing so I just guestimated that, indeed, there is a larger gap on the left than on the right. Also, at X200, it looks like the probe is slightly tilted upwards on the right side. So I will look into getting that right and try again.

Thanks for the support so far!

G28
M561
G1 Z10
G1 X10 Y50 F3500
G30 S-1
G1 Z0
G4 S15
G1 Z10
G1 X100 Y50 F3500
G30 S-1
G1 Z0
G4 S15
G1 Z10
G1 X200 Y150 F3500
G30 S-1
G1 Z0
G4 S15
G1 Z10
G1 X100 Y150 F3500
G30 S-1
G1 Z0
G4 S15
G1 Z10
G1 X10 Y150 F3500
G30 S-1
G1 Z0
G4 S15
G1 Z10

dc42

Could it be that your two parallel X rods are not quite parallel but have a slight twist between them? Try holding a spirit level vertically against them both at the left hand end, and then again at the right hand end.

InSanity

@dc42:

Could it be that your two parallel X rods are not quite parallel but have a slight twist between them? Try holding a spirit level vertically against them both at the left hand end, and then again at the right hand end.

I have this exact problem along with bowing and G29 bed comp is working really well for me. I have in the past had issues exactly like are being described, they have gone away after moving to the BLTouch from an inductive probe. To be fair I also start with a really well aligned bed. It's as if the bed comp works great so long as you start with something that has a really good chance of working with it off. IDK, it feels like gremlins at work in the code however the evidence during testing says otherwise. I really would like to see this bed comp map zero based.

Jeff

darookee

I just found a way to check that, my spirit level did not fit in there, so I had to improvise a bit. But it looks like this could be at least part of the problem, if not the problem.

It makes sense, when the angle of the probe is different for different points the results are different, wich gives a different trigger height resulting in incorrect compensation.

This would not be a problem when using the bltouch, because, AFAIK it touches the bed resulting in a consistent trigger height (at least with a small change of angle).

Now I'm trying to fix the twist, or get a BLtouch…

Thanks for all the help!

InSanity

@darookee:

Now I'm trying to fix the twist, or get a BLtouch…

The BLTouch more or less mimics a nozzle even down to the size of the probe tip. So yah, it doesn't suffer from these problems.

Jeff

Rommie

I'm tempted to try the bltouch as I have the same sort of issues, prints fine 1 side of the bed but no the other, bed is ali tool p[late so I know its flat with printbite on top which again I know is flat by testing it. when I do my comp the height map says its all over the place and cant get a flat bed at all so might try and see if the new probe will help, either that or try one of the new dc probes as mine is an old version

Dougal1957

Sean

Have a look in DJD's piezo sensor thread?

deckingman

@Rommie:

I'm tempted to try the bltouch as I have the same sort of issues, prints fine 1 side of the bed but no the other, bed is ali tool p[late so I know its flat with printbite on top which again I know is flat by testing it. when I do my comp the height map says its all over the place and cant get a flat bed at all so might try and see if the new probe will help, either that or try one of the new dc probes as mine is an old version[/quote]

If you know the bed is flat, then why bother with compensation? I have a similar issue which I suspect is due to me using 3DLac which changes the reflectivity but as I know the bed is flat and level, I just use the probe for homing.

Dougal1957

Ian Rommies issue is that the Effector tilt causes the probe to trigger at differing height around the bed and unfortunately this negates the use of the Bed Comp anyway and the prob will be just the same with a BLTouch.

I think DJ's Piezo idea will do away with that error as it uses the Nozzle tip itself as the probe point.

I wonder how difficult it would be to use Strain gauges on a

CB material or even better a Carbon Fibre effector with a set of Hayden's rod's?

Doug