RE: Printer is trash !
A layer shift is often due to assembly mistakes. Inserting the rods into a bearing can damage them if rushed. Same for the delrin T-nuts. damaged bearings are often the cause of layer shifts. As is over-tightening the U-bolt clamps on the Y-Axis, and not aligning the X-Axis properly with the Z-Axis. Things need to move freely and smoothly before you even apply power.
And the SD card is the ONLY reliable test/diagnostic source for discovering what is happening with the printer. It is far more reliable than any USB connection.
RE: Printer is trash !
Hmm, I have not had these same problems. Contact support.
--------------------
Chuck H
3D Printer Review Blog
RE: Printer is trash !
So, printed once from SD and have layer shift. I would repeat the sd card benchy print a number of times and see :
- layer shift repeated in the same place?
- layer shift, but at different points ?
- no layer shift at all for a number of prints ?
This will aid in diagnosys.
Texy
RE: Printer is trash !
I would add that I am very frustrated with the latest firmware and software releases. I printed more than 4800 hours on my MK3S with fabulous results and no fiddling with settings. I upgraded to 3.7.1 and immediately had issues. I reinstalled 2.6 and things went back to beautiful. I decided to upgrade to an MMU last month. BAD decision on my part. As part of that you must upgrade the firmware and use the new Prusa Slicer. IMHO this makes the printer almost unworkable. After several weeks I am still unable to produce the quality of prints that I was able to achieve on the MK3S on v2.6 firmware with no MMU. I am thinking of reverting to non-MMU (yet more time). I use the printer to print, not as a time suck to make it work right.
I heartily support JPrusa and his products. I have helped eight people buy and build their MK3. I have more than 6 years experience on multiple printers. I believe that the connector design on the bed needs redesigned to prevent sensor wire failure and to keep the wiring from the pinch between the back rail and the bed.
I also think that the product needs a way to calibrate E-steps as the "magic" extruder setup consistently under-extrudes on my device. I have to set the extrusion to 110% to get a somewhat reasonable first-layer. Today's time suck is working through all of the filament path to see if there is a binding anywhere. I need to be printing, not chasing problems...
RE: Printer is trash !
I also think that the product needs a way to calibrate E-steps as the "magic" extruder setup consistently under-extrudes on my device. I have to set the extrusion to 110% to get a somewhat reasonable first-layer. Today's time suck is working through all of the filament path to see if there is a binding anywhere. I need to be printing, not chasing problems...
what about this little part of the start gcode in prusaslicer:
M221 S{if layer_height<0.075}100{else}95{endif}
i think this will set flow to 95% most of the time and you are probably compensating for that with over extrusion.
i never quite understood why that code was needed, if it was a problem with the slicer then surely that should be fixed instead of working around it in the start gcode.
and does code that still apply? i bet it has been there for a while.
RE: Printer is trash !
The 95% flow is set to compensate all the inaccuracies in the extrusion. I remember Prusa mentioned it somewhere that they started with 100% but after many empiric tests they ended up at 95% and since then it is being used.
Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram
RE: Printer is trash !
[...] i never quite understood why that code was needed, if it was a problem with the slicer then surely that should be fixed instead of working around it in the start gcode.
and does code that still apply? i bet it has been there for a while.
There's a long-lost post on in in which JP states that it was added to compensate for the tendency of the Slic3r code to over-extrude at larger layer heights. I find it easiest to simply comment that line out and do my own calibration. I am finding that nozzles larger than 0.60mm tend to over-extrude independent of layer height, for example. Remember that the Prusa defaults are meant to be safe for the majority of users in the majority of cases. There's no reason not to tweak and optimize them. You can certainly get better surface finish and save time by doing so.
RE: Printer is trash !
I would add that I am very frustrated with the latest firmware and software releases. I printed more than 4800 hours on my MK3S with fabulous results and no fiddling with settings. I upgraded to 3.7.1 and immediately had issues. I reinstalled 2.6 and things went back to beautiful.
Did you mean 3.6" 2.6 never ran on the S.
Plus, nice to meet a time traveler. There have only been about 3500 hours since the MK3S was released. Do you have multiple printers?
RE: Printer is trash !
My major beef with the MK3 is the fact it has bed mesh leveling, and to this day is still isn't producing a level layer one. Temperature changes have way too much effect on how the PINDA responds, and the bed warp over long temperature soaks (multiple serial prints) for whatever reason affects the mesh level. So print one is okay, but print four has a layer one issue. It wouldn't be so bad except the mesh uniformity is what changes with time - not the absolute layer height. And uniformity is the entire reason for mesh level.
What's curious, is the custom code I was using for 7x7 mesh level (before 3.7.1 was released) seemed to NOT have this problem. So something Prusa has done differently is making it not work as well as it is supposed to work.
RE: Printer is trash !
@tim-m30,
I am down to two printers, The MK3S and a Lulzbot Taz5. I've had the MK3 running 24x7 since I received it early in the year. The hours are approximate since flashing the firmware seems to have ether zero'd or removed the lifetime hours. I know I had 1890 hours in May when 3.7 was released. Given there has only been ~1700 since then, it might be closer to 3000. You are correct in that it was 3.6, not 2.6. Rereading my post makes me aware my frustration was too evident. I am one of those people who want/need to get good prints without too much hassle. I don't mind tweaking the setup to make it happen, but if I have to start rebuilding and/or modding to make it happen, then I start thinking about a different printer. I am getting down to crunch time in my primary hobby (Extreme Christmas Lighting) and need to be focused on that to be ready for the season.
I've noticed the bed leveling is off as well. They seem to have moved to a much faster PINDA probing in 3.7. Maybe they are having a challenge with "settle" time.
RE: Printer is trash !
... The hours are approximate since flashing the firmware seems to have ether zero'd or removed the lifetime hours. ...
Zero'd and formatted differently in statistics.
RE: Printer is trash !
Heading a bit extreme don't you think? Nothing to do with you and poor assembly.?Have you tried the very helpful people in chat at Prusa.
Nigel
Life is keeping interested and excited by knowledge and new things.
RE: Printer is trash !
[...] I've noticed the bed leveling is off as well. They seem to have moved to a much faster PINDA probing in 3.7. Maybe they are having a challenge with "settle" time.
I'm a little baffled that you guys are having these issues. I'm still running 3.7.1, mostly because I've been in the middle of printing drawer and cabinet organizers since doing an office makeover and have my printer going as close to 24x7 as I can. I enabled the 7x7 5 probe mesh bed leveling when it was 1st released. I do use the PINDA warmup routine to ensure leveling occurs at a consistent 35C, but other than that, don't put much thought into it. My current mesh bed level numbers seem decent with my currently-mounted PC sheet:
I spent the last week printing on a PEI sheet that showed a .37mm variance on the right. I've been printing with 1.00 and 0.80mm nozzles all week, and the 1st layers have been very consistent even on large prints, from 135X200 bins to to 210x162 inserts at 1.2 and 0.96mm layer widths. Bed levels have been excellent. The only "problem" I'm finding is that the big nozzles tend to over-extrude, so I'm knocking 10% off my extrusion multiplier consistently.
Are you guys saying you're no longer getting level beds with the 7x7 mesh bed leveling?
I do lose my "lifetime" stats regularly, usually when I do a factory reset after a major firmware update.
RE: Printer is trash !
The issue I've been fighting is when printing 6 copies of something, same gcode ran consecutively, layer one changes print to print. Noticeably: One over extruded, the next not so much, and the last with visible strings.
I did not have this issue with mionut's 7x7 firmware from GitHub.
But I found something today as I was answering another thread. PINDA Calibration is now called Temp Calibration or some other silly name with no association to what it actually does. Seems I had never run the cal after resetting my 3.7.1 install. I ran the cal today and printed some 4x6 inch flats. The mesh problem seemed to be gone. Knock on wood.
I really need to write down all of the cal's I have done so the next time I do a factory reset I have some chance of getting back to where I started. It'd be really nice for the printer to dump all the cal constants via Pronterface -- or does it and I just haven't found it yet? I should probably look it up.
RE: Printer is trash !
Another thought; in reading these forums for months and months, I've heard enough anectdotal evidence that the USB port on the EINSY has "just enough" cpu power to download new firmware. That using it for anything else is prone to failure. The Prusa firmware is not written for real time applications, and is too busy handling print functions to deal with USB overhead while printing.
This might be an incorrect opinion based on misinformation, but it's the general sense of the issue. Any time people have problems printing, and Octoprint is involved, the first recommendation is to unplug the USB cable and try printing from the SD card.
nah the USB connection is fine. its handled by a different CPU remember. its serial whether it gets it from the raspi or the usb. it doesn't even saturate the usb link. i agree its a good thing to eliminate sources of problems until he works out whats wrong but i wouldn't blame usb. usb is all i ever use. sd cards are bad news
RE: Printer is trash !
reading this thread has made me terrified to update my firmware. first they remove usb, then the world-map of people's prusa printers without their knowledge, then dodgy firmware updates, - is prusa turning into an evil company? surely not.
RE: Printer is trash !
reading this thread has made me terrified to update my firmware. first they remove usb, then the world-map of people's prusa printers without their knowledge, then dodgy firmware updates, - is prusa turning into an evil company? surely not.
Well, if you're going to buy in to the thought process used by the guy whose first instinct was to take a sharpie and use his new $700+ printer as a coloring book, you can look at life that way.
- I'm not sure what you mean by "remove USB" unless you mean from PrusaSlicer. PS is under heavy development, and they took on an existing project that wasn't without problems to start with. There are many things they still need to fix (undo), but I think they're having to prioritize and triage efforts. Octoprint is so well done that it would make sense (to me at least) to simply officially hook into Octo for printing and not try to do it (poorly) in PS.
- The map thing was a goof, but I think Prusa was led down a primrose path by whoever is responsible for this overall site design. They did respond to the complaints about geolocation. I'm not sure they've made it "right" but at least "better". They're still trying to make these forums work as well as the old software did, with only mixed success.
- Nothing forces you to install a firmware update early on. Just wait a few weeks if your printer is otherwise working fine. That's a good idea with any software you depend on, not just Prusa firmware.
RE: Printer is trash !
[...] But I found something today as I was answering another thread. PINDA Calibration is now called Temp Calibration or some other silly name with no association to what it actually does. Seems I had never run the cal after resetting my 3.7.1 install. I ran the cal today and printed some 4x6 inch flats. The mesh problem seemed to be gone. Knock on wood.
I have inquired whether they are indeed the same thing, but never found an official answer. I've added the temp calibration as part of my printer calibration routine as part of a factory reset procedure.
I'm still sticking with the PINDA warmup which accomplishes the same thing on a per-print basis, albeit at the cost of slow startup. Perhaps that's why I'm not seeing the issue?
I recall reading that completing the on-board Live-Z calibration set a flag or variable in non-volatile memory and was required for consistency. Perhaps the temp calibration operates the same way?
I really need to write down all of the cal's I have done so the next time I do a factory reset I have some chance of getting back to where I started. It'd be really nice for the printer to dump all the cal constants via Pronterface -- or does it and I just haven't found it yet? I should probably look it up.
RE: Printer is trash !
Another thought; in reading these forums for months and months, I've heard enough anectdotal evidence that the USB port on the EINSY has "just enough" cpu power to download new firmware. That using it for anything else is prone to failure.
nah the USB connection is fine. its handled by a different CPU remember. its serial whether it gets it from the raspi or the usb. it doesn't even saturate the usb link. i agree its a good thing to eliminate sources of problems until he works out whats wrong but i wouldn't blame usb. usb is all i ever use. sd cards are bad news
That actually isn't entirely accurate. The secondary cpu handles the USB to an extent, but all buffering happens in only 2k of ram, unless the serial link to the primary cpu can absorb bits. So a multi-packet transaction can overwhelm the design.
RE: Printer is trash !
As for the cals, PINDA doesn't need a PID cal (Proportional Integral Derivative, it's been too long, but this adjusts the proportional control of heat applied to improve set temperature control accuracy and prevent under shoot and overshoot during ramps), so the old name was inaccurate. A change was probably in order. But why leave PINDA out?