Z height varies wildly between prints (pic included)  

Re: Z height varies wildly between prints (pic included)

There seems to be some confusion about the Live-Z on the forums; it doesn't mean the height difference between the nozzle and the probe. If you'll look at my photo on previous page, you can clearly see that I have enough clearance between the nozzle and the PINDA probe (~1mm), yet my Live-Z is set to -0.110mm.

Yes, you do adjust the Live-Z when the printer is running, but it shouldn't have to be a repeat process at the start of every print. I am of course adjusting the Live-Z when the bed is hot. I don't think there's any other way, because the print doesn't actually start until it reaches a set temperature.

My test square is always in the centre, 50x50mm.

I've been reading a lot on similar issues and cooling down the probe seems to help in some cases.

sorry , didnt see the pic of your pinda height , it looks ok .
but then again , if the height set correctly you should have your live z around -600um with the new firmware .
the only thing i can think of that might cause this is
- you didnt bring your z axis all the way up during xyz calibration .
- your pinda sensitivity is bad .
you didnt answer my previous question , when you home , is the pinda located exactly at the center of the marked circle ?

Re: Z height varies wildly between prints (pic included)

but then again , if the height set correctly you should have your live z around -600um with the new firmware .

Sorry, this is simply not the case. Live Z will depend on both probe sensitivity and position. The probe should be positioned suvch that the printer passes XYZ calibration with the probe as high as possible.

Peter

Re: Z height varies wildly between prints (pic included)

There seems to be some confusion about the Live-Z on the forums; it doesn't mean the height difference between the nozzle and the probe. If you'll look at my photo on previous page, you can clearly see that I have enough clearance between the nozzle and the PINDA probe (~1mm), yet my Live-Z is set to -0.110mm.

Yes, you do adjust the Live-Z when the printer is running, but it shouldn't have to be a repeat process at the start of every print. I am of course adjusting the Live-Z when the bed is hot. I don't think there's any other way, because the print doesn't actually start until it reaches a set temperature.

My test square is always in the centre, 50x50mm.

I've been reading a lot on similar issues and cooling down the probe seems to help in some cases.

sorry , didnt see the pic of your pinda height , it looks ok .
but then again , if the height set correctly you should have your live z around -600um with the new firmware .
the only thing i can think of that might cause this is
- you didnt bring your z axis all the way up during xyz calibration .
- your pinda sensitivity is bad .
you didnt answer my previous question , when you home , is the pinda located exactly at the center of the marked circle ?

Hey Sergey, thanks for getting back.

As Peter has mentioned (and has been discussed on the previous pages), Live-Z height doesn't reflect the distance between the probe and the nozzle, but compensates for PINDA sensitivity and tolerances.

I'm 100% sure that both the Z axis were all the way up. I always push them a few turns more after they start to rattle. The XYZ calibration passes without problems and I get perpendicular XY. The PINDA centers perfectly (as much as I can tell by looking) over the circle.

As I've said, letting the PINDA cool down a bit helps. Maybe I just have a slightly odd sensor that's affected by the temperature more than others.

Re: Z height varies wildly between prints (pic included)

I recall a discussion of having the heat break cooling fan pull rather than push air being a factor as it blow hot air (that flows past the heat break) against the PINDA probe.

Might be worth a try. Easy to test - just mount the fan backwards. I had mine mounted backwards for months with no issues.

Thanks for the suggestion. I went and double checked and the fan seems to be facing the right way. I went back to the manual as well, it says to make sure that the fan sticker is facing the nozzle. As I can't see the sticker on the outside, it must be facing the right way. Is it possible to verify this even further by trying to feel the airflow with my hand (i.e. air flowing outwards away from the nozzle)? I've never had a blockage as long as I've been using the printer.

Re: Z height varies wildly between prints (pic included)

So, I have been thinking about your problem.

Your PINDA probe is much more sensitive than the majority on the forum.
1) ours would never trigger for an XYZ Cal set as high as yours.
2) the electronics sees yours as lower thus your live-z relative to our numbers.

Your PINDA varyies by large amounts , where the average experience is that it is stable.

Maybe the extra sensitivity comes with variability?

Replacing the PINDA might be the solution.

Re: Z height varies wildly between prints (pic included)

Your PINDA probe is much more sensitive than the majority on the forum.
1) ours would never trigger for an XYZ Cal set as high as yours.
2) the electronics sees yours as lower thus your live-z relative to our numbers.

i think that his sensor is actually much less sensitive then standard , thats why he has his live z closer to zero .
normal pinda probe detects the cal point at x level above the bed . if you need to lower your nozzle (live z) to -600um , and he needs to lower it just to -100um , that means that your sensor detected the bed at much higher height compared to his (yours detected 500um above to be exact, if nozzle to pinda height the same)

for example if your pinda was super sensitive and was detecting the bed 1cm above (just an example), then you would need live z value of around -10000

Re: Z height varies wildly between prints (pic included)

So, I have been thinking about your problem.

Your PINDA probe is much more sensitive than the majority on the forum.
1) ours would never trigger for an XYZ Cal set as high as yours.
2) the electronics sees yours as lower thus your live-z relative to our numbers.

Your PINDA varyies by large amounts , where the average experience is that it is stable.

Maybe the extra sensitivity comes with variability?

Replacing the PINDA might be the solution.

Thank you for your help.

I'm getting stable prints by letting it cool down a little, so that's something. I don't mind replacing the PINDA though, it can't be that expensive. I've looked in the Prusa shop and it's not there. Searching the internet produced a couple of knock-offs with blue bottom part for a few dollars from China. Is there a known PINDA replacement that should work with the Prusa?

Re: Z height varies wildly between prints (pic included)

I'm getting stable prints by letting it cool down a little, so that's something. I don't mind replacing the PINDA though, it can't be that expensive. I've looked in the Prusa shop and it's not there. Searching the internet produced a couple of knock-offs with blue bottom part for a few dollars from China. Is there a known PINDA replacement that should work with the Prusa?

No, you need specifically 5V sensor (although a 12V version can be made to work); they are available in the shop: http://shop.prusa3d.com/en/mk2-spare-parts/127-pinda-probe.html , but you may need to use Live Chat to get access to that section of the shop.

But you don't need one; it will not help. You are going to have to get the probe to a constant temp so In this instance I would suggest that you do probe temperature calibration.

Peter