Belt Calibration Nightmare
Hello,
a week into assembly, after just a few prints, I suddenly encountered a homing issue. It started with the home function: it started bumping and bumping and bumping and then failed. No way to home the printer.
I performed the Home calibration test, and the printer wouldn't pass the homing test 😭 and so I contacted PRUSA.
I'll spare you the six-hour test story, but basically, the belt tension was uneven and the gantry was warping. 😣
It was perfect a few days earlier, but now it wasn't. So I spent the entire time with PRUSA support figuring out how to calibrate the belt tension and make the gantry squared at the same time, but with the optimal belt tension. This is tricky because if you change the tension, the gantry become uneven, and it is very difficult make it square with the correct level of tension. You may have it squared with the wrong level of tension, or you may have the correct level of belt tension but the gantry is not squared any more 😫 !
Moreover this, you have the issue of getting the tension value!
The problem is that the strobe lighting system is too subjective when it comes to adjusting the belts. I, on the other hand, had used a dynamometer in the past, which always gave me the exact measurement, and I had always calibrated my printers with that. The problem was that we couldn't agree on the dynamometer value to use, since they don't use it.
Then, support sent me a video of one of their printers that was (almost) correctly calibrated, to show me what it looked like under the strobe light.
From there, I tried it myself, but the tensions were still approximate. The oscillations were quite subjective and imprecise, and even with the video, because it didn't provide precise measurable data. Eventually, he suggested a different method that he used to tune his printers, again with strobe lights, which somehow allowed me to calibrate it, and the printer passed the tests 😊.
Then came the turning point, when he also sent me a video of the sound I heard when plucking the straps 🤩 . Luckily, having perfect pitch, I immediately reproduced his frequency adjusting the belts to the same frequency. Then, however, I measured the PRUSA value with a dynamometer and discovered that it was exactly the initial value I had set during my assembly 😆 .
I did not had any issue since then, and almost all head bums in homing went away.
Have some of you had a similar experience with belts and gantry ?
Best Regards
RE: Belt Calibration Nightmare
just curious, what values did you measure with your dynamometer? information like that should be spread far and wide!
RE: Belt Calibration Nightmare
So I spent the entire time with PRUSA support figuring out how to calibrate the belt tension and make the gantry squared at the same time, but with the optimal belt tension. This is tricky because if you change the tension, the gantry become uneven, and it is very difficult make it square with the correct level of tension. You may have it squared with the wrong level of tension, or you may have the correct level of belt tension but the gantry is not squared any more 😫 !
Frankly, I can't relate to that problem. The most recent belt tensioning instructions are pretty clear in my view, and break the process down into a sequence of steps:
If the gantry is skewed, relax the belts fully and straighten the gantry by bending it. Then gradually tighten the belts, always turning the same amount on both sides. Once you get them into the range where the stroboscope wizard can take over (> 60 Hz if I recall correctly), use that. I found the resolution of the stroboscope tuner easily good enough.
The only poorly defined step is "getting into the range where the stroboscope wizard can take over". Depending on one's experience with musical or technical sounds, one may not have the right gut feel for that, and may need a few attempts until the stroboscope finds a resonance.
RE: Belt Calibration Nightmare
Depending on one's experience with musical or technical sounds, one may not have the right gut feel for that, and may need a few attempts until the stroboscope finds a resonance.
I've gotten to the point where I can pluck it with my finger and hearing the tone get within +/- 5hz
RE: Belt Calibration Nightmare
Yes, that's what I meant by gut feel. If, based on experience, you have an idea what properly tensioned belts sound like or what 60 Hz sound like, it's very quick to do the coarse approach. But for those who do not have that experience, Prusa does not provide guidance how to get into the ballpark.
RE: Belt Calibration Nightmare
I just finished my build and had the same problems with as Antimix with the directions and built in tuner. The idea of relaxing the bells completely and then tightening an equal number of turns is ludicrous. There’s no way to know precisely when a belt is fully relaxed. Therefore the starting point could easily be several turns off between the two belts.
Here’s the procedure I used:
1. I got the gantry squared with no belt tension.
2. Added tension alternating between belts while maintaining squareness, disregarding of difference in turns that it took.
3. Once the belts had reasonable tension the screws were further tightened by the same amount each.
4. With the carriage moved to the right, Used the app on an iPhone to get the tensions to recommended frequencies strumming the belts in the front.
5. Slight adjustments still needed between screws to maintain squareness.
6. Switched to the built-in test and found the values to match those obtained by the phone app.
Prusa’s procedure needs to be revamped.
RE: Belt Calibration Nightmare
I'd not recommend trying to use the phone app at all.
The directions at https://help.prusa3d.com/article/adjusting-belt-tension-core-one-l-core-one_845048 are easy to follow and quite complete.
RE: Belt Calibration Nightmare
There’s no way to know precisely when a belt is fully relaxed. Therefore the starting point could easily be several turns off between the two belts.
Just loosen all the way until the tensioner pulley detaches from the screw then reattach both until the thread just catches. Both belts should be in the same state. Then tighten both starting with 3 turns each. Once the belts get tight, move the head to front right, pluck them and start reducing the number of turns until it "sounds" in range. Then switch over to the strobe tuner.
Anyway, this kind of belt tuning is a bit too manual for me. My "other" printer auto-tunes the belts before every print. Hopefully Prusa will figure this out by the time the "Core Two" is released.
RE: Belt Calibration Nightmare
That assumes that you got both belts fastened identically. What did it call for - 4 or 5 teeth visible?
RE: Belt Calibration Nightmare
That assumes that you got both belts fastened identically. What did it call for - 4 or 5 teeth visible?
No I didn't. In the video the sound from the two belts had different frequencies, so I tuned mine according to the video.