Enclosed VC3.1 300mm Belt Tensioning
I've been trying to make sense of the belt tension and resonance graphs for an enclosed VC3.1 300mm I just made. It has the FMMM, Motor Decouple-Damper, an EBB42 and it's sitting on the floor (on floor tiles, so maybe that could contribute to my problems, though it does not have any noticeable rocking). I've watched the youtube video that is recommended everywhere for
how to read and analyze input shaper graphs..., and took a look at a fair amount of posts/messages here, but in trying to follow guidelines and recommendations from various posts, I'm still at a loss of how to tension my belts. I can't tell if they are over-tensioned, under-tensioned, or if something else is causing what I believe may be a non-ideal belt tension graph. The only thing I understand is that they should both have the same tension...which I can't really tell that either...maybe the upper one might need to be tightened a little more to match.
5 Replies
you can't really tell from the graphs when they are a mess. i'd go in the following order.
1. verify that ADXL sensor works correctly:
- call
ACCELEROMETER_QUERY it should respond with something along the lines of adxl345 values (x, y, z): 70.719200, 41.438400, 9728.196800 one value close to 10 000 other 2 close to 0
2. if you are using the toolboard built in accelerometer, make sure the toolboard is rigidly mounted to the toolhead, if it's not the graphs won't show anything useful
3. measure diagonals https://yko.im/Z5Dv.png to make sure printer is square.
3. follow this https://canary.discord.com/channels/582187371529764864/791048619997659176/1196002549606666260 for initial belt tension
4. generate graphs again.
if you have any questions feel free to @ me :)
Alright, so I've done a small number of things that have improved my graphs somewhat
I'm using the EBB42 toolhead, so the ACCELEROMETER_QUERY doesn't work for me, but the toolboard is mounted, and as far as I can tell, isn't loose.
For the diagonals, I'm about 1mm off between the two. Hopefully that doesn't completely ruin things (nothing seems to be binding).
I did take off my belts, re-measured them and cut them to have the exact same number of teeth and length
Re-belted, re-aligned the X-gantry, and probably one of the more important issues: I removed the nozzle from the hotend (forgot I just screwed it in for safe keeping, and only remembered when I looked at the print head and saw it spinning around), put some thermal grease and heat tightened it
The belt tension graph...looks better? Though I still don't see anything like an ideal twin peaks, they are still at least roughly about the same relative tension. They could still be loose too maybe
The X and Y resonances look what I'm assuming to be alright and the loose nozzle I had explains away the double peaked X resonances I was getting before.
@Prav Are these graphs still too messy to determine anything?
not ideal but much better than before. Belt tension wise you are pretty much equal. Your X and Y graphs should have higher frequency and higher peaks.
Y around 50Hz and X 75Hz if it's a stock 300mm v-core