cold vs. hot mesh bed leveling
It's been said that hot mesh bed leveling (both heated bed and hot end are not) is superior to cold bed leveling, because the heated bed and even the nozzle assembly may change in size due to the heat, and those are the conditions under which you are printing.
However, the downside of that is possibly oozing little bits of extrusion all over the heated bed during the 9 point bed leveling, and needing to clean those off in a hurry before the print begins.
Bottom line: how much, if any, accuracy is lost with cold bed-leveling instead? Is it negligible? Because of the above downside, I'd prefer to do cold bed leveling, and so I'm wondering what the cost of that the trade-off is.
Anyone happen to know? If not, how could I measure it? e.g. Would printing two calibration cubes, one with hot bed leveling and the other with cold bed leveling, and then measuring them afterward with calipers be sufficient to detect any important differences? I mean, presumably only the z-axis would show differences, if they exist, right?
Re: cold vs. hot mesh bed leveling
You can always use my method, a partial heat. A full heat of the bed and the nozzle to around 160. That eliminates the oozing during the levelling. However while it will be more accurate than cold levelling it will not be quite as accurate as hot levelling. I think however that it is a good compromise.
See No Ooze Mesh Bed Levelling In Slic3r PE 1.39.2 for details of this method.
Regards,
Martin
Martin Wolfe
Re: cold vs. hot mesh bed leveling
Bed leveling depends on the probe, not the nozzle. The nozzle's heating has minimal effect on the probe. At least, in terms of leveling. Only the changes in the bed's shape is account for with the leveling.
That said, one downside to probe vs nozzle leveling is that you can't take both thermal expansions into account. However, as you state, the oozing of the plastic would affect the nozzle leveling anyway.
Given the method implemented, I don't see a reason to heat the nozzle for leveling. It makes sense to heat the bed, mesh level, then finish heating the nozzle.
Re: cold vs. hot mesh bed leveling
I don't see a reason to heat the nozzle for leveling.
Any ooze remaining in the nozzle needs to be soft (otherwise calibrations can fail), which is why I heat the nozzle to 170 degrees before levelling.
At 170 degrees, the plastic is not too soft to ooze during levelling.
Peter
Please note: I do not have any affiliation with Prusa Research. Any advices given are offered in good faith. It is your responsibility to ensure that by following my advice you do not suffer or cause injury, damage…
Re: cold vs. hot mesh bed leveling
I don't see a reason to heat the nozzle for leveling.
Any ooze remaining in the nozzle needs to be soft (otherwise calibrations can fail), which is why I heat the nozzle to 170 degrees before levelling.
At 170 degrees, the plastic is not too soft to ooze during levelling.
Peter
I was having the oozing filament issue with my new MK3 (first printer BTW), and tried setting to 170 at bed leveling time, but still got a (smaller) bit of oozing. Possibly from the temp overshoot on heating, when it ramps up to 170 and peaks around ~175 before settling back to 170. I changed to 160 and haven't had issues since.
Prusa i3 MK3 kit
MK3/S Upgrade Kit
FW 3.7.0
Slic3r PE 1.41.3
Re: cold vs. hot mesh bed leveling
I solved the oozing problem in another way with start Gcode.
I let bed and extruder heat at full temperature, then I lift Z axis of 10mm (in case your nozzle is close from bed), I extrude about 15mm of plastic and then make a quick retractation (about 6 mm).
I remove the plastic from the noozle and the bed leveling start.
After homing I push back 6 mm of plastic quickly before priming.
No more oozing problem and mesh bed leveling in the same conditions than printing.
Re: cold vs. hot mesh bed leveling
The really strange thing is this: if having a Pinda of >35C is so important, why doesn't the XYZ Calibration ensure that it is? The calibration is done with Pinda at ambient temperature, but in our g-code we do our Pinda measurements with Pinda >35C.
That makes me question the accuracy of the calibration results if they are used in this way. Or am I wrong about that?
Re: cold vs. hot mesh bed leveling
The 35C pinda restriction is entirely optional and activated via gcode in the slicer. Some folks seem to have very non-linear pinda probes, and this 35C technique seems to help with that. My pinda is nearly perfectly linear at PLA and PETG temps, thus I don't have to worry about the Pinda temp. I only make sure my pinda is not cold (like not less than 20 C) and print all the time. No z-adjustment, no temperature concerns. I guess I'm one of the lucky ones.
Re: cold vs. hot mesh bed leveling
Well, assuming I'm unlucky and have one of the non-linear pinda probes, then would it work better if I made a point of doing the XYZ calibration after first ensuring that the PINDA temperature is ~35C, so that it would match the in-use conditions? I'm just thinking out loud here. I hadn't been doing that, and I hadn't read of anyone else who is using the g-code enhancement doing it either (not that I've read every thread ever written on the topic, but no one has mentioned it lately).
Re: cold vs. hot mesh bed leveling
It's a known issue, and luckily, is being worked on:
Automatic mesh bed leveling causes oozed filament deposits to be placed on the bed
https://github.com/prusa3d/Slic3r/issues/913#issuecomment-395204872
Re: cold vs. hot mesh bed leveling
I changed the starting script a bit and made my mesh bed leveling sequence as below:
1. Heat up and wait for the desired bed temperature
2. Home all
3. Start Mesh Bed Leveling (9 points)
4. After Mesh Bed Leveling, it should be returned to 0,0 position automaticallly
5. Heat up and wait for the desired nozzle temperature
6. Once the desired nozzle temperature is reached, start initial wipe line and proceed the printing
So that the mesh bed leveling is done under heated bed, while the nozzle is still cold and will not leak any filament.
But this indeed lengthens the total heat up time.
RE: cold vs. hot mesh bed leveling
A follow up to this old thread: I got a Prusa MINI (so I don't know how this works on the other models) and on MINI, the leveling process is not very efficient. It does roughly these steps before each print:
1. Heats/Cools the nozzle temp to 170C
2. Brings the bed temp to target temp for the current filament type
3. Waits for both temps to stabilize at their target temps
4. Starts the bed leveling procedure.
5. Once the leveling is done, continues to printing
OK, this all looks reasonable, right? Except if you just did a print right before (or calibration, or manually heating the nozzle, etc) and the nozzle temp is already hotter than 170C target. In that case it simply waits for the nozzle to cool off by itself which may take some time.
I propose one of the following improvements:
- it should either kick on the print fan at 100% to help cool off the head if the nozzle temp is > 170C, OR
- leave the nozzle temp as is if it's higher than 170C (I never seen any oozing in between prints, although my nozzle was hotter than 170C, probably around 200C)
What do you think about this idea? If it makes sense, I can create a PR for the PrusaSlicer with either of these two implementations so it becomes the default behavior (I believe that this behavior is controlled by the "before print" custom G-Code instructions in PrusaSlicer, not in the firmware).
Cheers,
Florin
RE: cold vs. hot mesh bed leveling
Another way is to get the bed levelled (physically) at your target temp and bypass MBL step.
Google "Nylock Mod". You level it physically whilst at your desired temp so you know every time that it's a that temp it has the same mean deviation you adjusted it to, in theory you can bypass MBL step if you have a low enough deviation since you leveled your bed ahead of time.