RE: Belt tensioning issues - firmware tensioning guide and website tensioning "app"
The release candidate of the most recent firmware update was causing alot of collision detection errors and it took forever for a print to start; although the print worked out successfully.
This version of the firmware is out of RC so I'm hoping that the stable version address these issues.
RE:
Hey Butters,
I am on the stable (out of RC) version, 6.4.0. I hope everything works out ok for you now.
My additional problem with the tensioning screw turned out to be less severe than I thought and I managed to put things back together.
I started again from loose belts. The quick gantry adjustment just doesn't work for me and I didn't want to "push" too much so I just assume either my difference between end stops is too small or I have a crooked part somewhere ; so I skipped that step and went straight to step 3 and 4 here: https://help.prusa3d.com/article/adjusting-belt-tension-core-one-l-core-one_845048
So I just did that sort of combined with all the advice here; when my belts were very loose I:
- tightened both screws almost equally, tightening a bit more on the side with more play. For example 3 quarters (or a bit more) of a rotation on one side, 1 full rotation on the side with play
This didn't get me in the perfect spot so when my belts were too tight I:
- loosened both screws almost equally, loosening a bit less on the side with more play. For example one full rotation on one side, 3 quarters of a rotation (or a bit more) on the side with more play.
I randomly stopped at some point, re tightened the belts (in a similar fashion as above).
Somehow after this I got both the web app tuner happy and the manual belt tuner in the firmware (where no adjustments were needed). I re-ran input shaping calibration and phase stepping callibration, the values changed a bit and I AM NEVER TOUCHING IT AGAIN (joking, will surely do). Hope to get printing soon.
Thank you again to everyone in this thread
RE: Belt tensioning issues - firmware tensioning guide and website tensioning "app"
- tightened both screws almost equally, tightening a bit more on the side with more play. For example 3 quarters (or a bit more) of a rotation on one side, 1 full rotation on the side with play
Hi @cupu, if possible it's best not to do this. First you square the gantry (if needed) when the belts are loose by manually bending it with your hands. Then when you are tightening the belts, as long as you are tightening both screws by the exact same amount alternating between them, then the gantry should not bend or skew. If you try to re-square it by unevenly tightening the screws you can could possibly cause more issues later.
RE: Belt tensioning issues - firmware tensioning guide and website tensioning "app"
Hey hyiger,
Thanks for drawing attention to possible problems later. I will muster up some courage and try again in a few days.
RE:
Hey hyiger,
Thanks for drawing attention to possible problems later. I will muster up some courage and try again in a few days.
Prusa had instructions (at some point) to remove skew in the gantry by changing the belt tension (turning the screws) but the current state is to ensure the gantry is square before tensioning the belts. When you've done it once or twice it becomes pretty routine. Also, it's a good idea to recheck the tension every once in while. Like checking the air in your car tires.
RE: Belt tensioning issues - firmware tensioning guide and website tensioning "app"
That's good advice about checking tension now and then, I've set a reminder in my calendar to check in 3 months.
I do not believe my gantry can be squared without any tension in the belts. I've tried many times over the course of this forum post. I am afraid that the only thing left to do is put more force into "fixing" the loose-belt gantry and I will break something. Maybe I have a part that's not exactly at a right angle or something ; I don't feel like taking the printer apart. We'll see if that is a mistake or not.
RE: Belt tensioning issues - firmware tensioning guide and website tensioning "app"
That's good advice about checking tension now and then, I've set a reminder in my calendar to check in 3 months.
I do not believe my gantry can be squared without any tension in the belts. I've tried many times over the course of this forum post. I am afraid that the only thing left to do is put more force into "fixing" the loose-belt gantry and I will break something. Maybe I have a part that's not exactly at a right angle or something ; I don't feel like taking the printer apart. We'll see if that is a mistake or not.
The two brackets that attach the gantry to the Y axis bearing are not always at 90°. They are also made from mild steel and are not difficult to bend. With the belts completely loose, if the skew is to one side then grab both sides and bend the other side by hand until it is square. It doesn't take much pressure. However, unless you go caveman on it you won't break anything.
As for checking the belts I do it irregularly at about once a week or so.
RE:
Simple... Fix the gantry by bending it, it's even in the instructions (do read them, and don't apply force to the gantry without loosening the belts or you may rip out an idler pulley).
I hate to break the bad news but you do have a part that is not exactly at a right angle. That "part" is sheet metal (a pair of symmetrical parts, actually, where the error cancels out, ideally. Not necessarily in reality though).
Bend the gantry in shape with belts off (doesn't require "violence" just controlled use of moderate force), then double-check it's still orthogonal under tension and fine-tune screw balance until aligned again.
RE: Belt tensioning issues - firmware tensioning guide and website tensioning "app"
Hi again,
Sorry to resurrect this topic but I said I would reply if i try again.
I had some time to work on this again (so many months later), I did and experienced the following:
- I tried to twist the gantry like in the prusa instructions but I couldn't get it to touch both sides of the front bar at the same time when bringing it forward.
- I completely unscrewed the belt tensioners and then screwed them back in for a very small length, so that I can start tightening them from the same reference point
- To my surprise, at this point, the gantry looked to have no isses at this point, it was touching both ends of the front stops at the same time. All right, I have a gantry that is "square" enough!
- I tightened the belts mostly the same amount in turn
- I ran the belt tensioner, good results, tests passed
Okey , I have never been in such a good position, thank you everyone!
Wait let's run the input shaping and phase stepping processes with the accelerometer, all good as well.
I ran the homing calibration step. Endless head banging and 2/3 times it fails.
But since it passes once, I accept it.
Now I have much more "head banging" before every print. Sometimes it does it 15 times, sometimes 4 times. Eventually it starts.
I will accept that life is not perfect and so are my processes and attempts. I take the belt tuning as being successful, I take the homing as "good enough to still print"
Thank you again to everyone who posted here
RE: Belt tensioning issues - firmware tensioning guide and website tensioning "app"
Wow, I forgot how long this thread was! Anyway, I think the latest firmware does bang the print head a few more times than in the past, but I'd get nervous if it exceeds six diagonal cycles at the start.
That said, I still don't like your first statement about not getting the gantry square. That's just fundamental to ever having the system work correctly. With the belts loose, and that means with the tensioners either relaxed or undone entirely (not necessary), the gantry has to hit left and right stops at the same time. Not just close, you need to be really fussy about this. It should be achievable with slight bending, basically just moving the gantry to the front-back center and forcing it in the direction you want it to bend. The whole process should take 1-2 minutes at most.
Report back as to whether you can achieve that alignment!
RE: Belt tensioning issues - firmware tensioning guide and website tensioning "app"
Hey @conrad,
Thank you for the reply.
To be clear: with the belt tensioners fully loose, entirely undone, the whole setup is square. Maybe when I was loosening the belts to test for a square frame I was not loosening enough?
I have also twisted on the gantry a lot of times, sometimes with quite some force, before completely undoing the tensioners (but with the belts loose) so maybe that also helped.
RE: Belt tensioning issues - firmware tensioning guide and website tensioning "app"
OK- great! With the old firmware I could reliably count 4 diagonal bangs. That's rare now and I get 5 or 6. Maybe 8 on a bad day! So how does it print?
RE: Belt tensioning issues - firmware tensioning guide and website tensioning "app"
I am sure I get more than 8 bangs now, but I will take some notes and see what my average is.
I am only slightly annoyed by it since I always print at night but my family has not complained yet! It's a minor issue, really.
I've never had problems with it actually, it has already printed good enough for what I was trying out (I print various functional parts around the house, nothing taxing). Maybe it prints perfect (that's how I feel about it), more likely I am not skilled enough to notice if it is lacking.
I never did benchmark prints. I do have the califlower benchmark tool on my to print list for a while.
This whole thing only started because I couldn't get the firmware belt tuner to give me an OK result originally. Then the discussion went on here because of all the good people trying to help.
As soon as the heat wave relaxes a bit here, I will print a copy of my engineer's square and see if they line up 🙂 .