High temperature enclosure printing - bed height dropping requiring Z adjustment
I've built up an enclosure for printing ABS and to get good results with larger parts and ones that have long flat parts in particular, I've found I need an enclosure temperature of about 50C. This works fine and I've been printing some ABS parts in this for months, the power supply is out of the enclosure and the rest of the printer seems to tolerate it alright - I even upgraded from a MK3 to a MK3S with MMU2S and have had good results at temperature. The issue I'm seeing happened on both configurations.
My issue is an odd one, and I'm wondering if others have observed it or they know what may be happening: I'll set Z height adjustment, print successfully for weeks (usually with a gradually increasing thickness of glue stick applied), and then suddenly, will lose most of my bed adhesion on the first layer because the Z height adjustment is off. I've observed this maybe half a dozen times in the few months I've been printing things at this temperature, and at least I don't remember it at 40C or below temperatures as I was improving my enclosure and trying gradually higher temperatures, but the most recent one I corrected today and needed almost 80um of adjustment (-1.540mm to -1.620mm live Z adjust) to get comparable first layer adhesion.
Now the number is a bit high, so if it happens again I'll probably also lower the height of the PINDA probe to get a more reasonable value, but what could be causing this? Is it a heat sensitivity of the PINDA probe thing? Is something supporting the bed or supporting the Z axis warping or shifting suddenly with thermal cycles? Has anyone else observed something similar?
Visually, I can't see anything that's doing it, and I don't see/hear anything that's loose. When I rebuilt the extruder for the MK3S upgrade I went back and remounted the heated bed to be sure and removed a little play in one Y axis bearing.
RE: High temperature enclosure printing - bed height dropping requiring Z adjustment
Pinda probe still good? or is this getting to hot, and getting higher or lower because of heat.
After printing you need to clean the bed, and put fresh glue stick.
Maintenance of the printer? clean rods put oil on rods, grease on gears, etc.
Check all bolts if they are loose.
RE: High temperature enclosure printing - bed height dropping requiring Z adjustment
No loose bolts (none when I made the MK3S upgrade either), no signs of issues with lubrication, things seem to move freely and there is no additional noise compared to the newly built printer. You say new glue stick but I know from experience this isn't true 😉 I usually go around 10 prints with recoating in a thin layer of glue stick before fully cleaning, but the issue is actually the same with built up glue layers and the fully cleaned print surface and if anything, the height of the glue would decrease the distance to the nozzle, partially compensating for this offset. Using the included glue stick and using the replacement I've found (one earlier replacement was a terrible choice), this method has been quite reliable.
This could be some longer term temperature drift thing with the PINDA probe, but it is consistent hot or cold (once the issue happens, it doesn't reset when run cold). If I switch over to PLA from ABS and leave the chamber open, the newly lowered setting is still the correct one. I seem to remember hearing that Prusa tested them for heat drift at the factory too, so while it's possible, I wouldn't have expected it.
I do think this is somehow heat related, though. I could do a couple dozen non-ABS prints at normal temperatures over a long time without observing this drop, but while it doesn't generally happen on the first ABS print, it happens after a half dozen or so. Maybe there's some loosening/tightening mechanism from thermal expansion that's making a part move gradually, though as mentioned, I'm not seeing a gradual decrease in adhesion/first layer adjustment height as much as a sudden one that then remains constant.
RE: High temperature enclosure printing - bed height dropping requiring Z adjustment
With the MK3S - it is known that printing at 250c will cause the exruder plastic to melt which causes the PINDA to droop and lose calibration.
Printing inside an enclosure just makes that defect much worse. Reading the threads on the issue - printing a new extruder housing in PC or other high temp plastic is the common solution.
RE: High temperature enclosure printing - bed height dropping requiring Z adjustment
Hmm, that certainly seems like a design oversight given the included temperature profiles and advertised temperature range. I do have a spool of PC (I think it's clear blue that's a perfect match for my orange parts 😑 ), but I haven't printed anything in PC yet, maybe I'll get some benchies done and give it a shot. While it sounds like a good change to make given the temperatures I'm working at, I did see this same issue with the base MK3, did that have a similar problem?
Is there any benefit to printing the PETG parts in ABS instead, or is that still too low? I assume most nylon options are too flexible to be good choices.
RE: High temperature enclosure printing - bed height dropping requiring Z adjustment
ABS is also perfectly fine. Only issue is that you will have tight openings if printed in ABS. It shrinks.
Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram
RE: High temperature enclosure printing - bed height dropping requiring Z adjustment
There is an additional aspect of the PINDA shifting beyond the R4 tendency to melt itself.
PINDA body is actually round and threaded. We also create a loop with its cable when dressing it into the cable guide. That loop of wire can exert a rotational tension on the PINDA. If the holder gets just a little loose, that rotational stress can thread the PINDA up/down in its holder even before you notice it has come loose. Basically, things don't need to get loose enough for the PINDA to slide up/down. It just has to get loose enough to rotate.
PINDA mount melting is a bigger problem. I mention this as a smaller, but easily concealed issue.