Question 2nd Motor to mirror extruder motor motions?

[Laserbea4k43]

I'm able to print a cube but larger prints will make the filament spew out of the back.
back to the drawing board.

 

[EvilleRock]

Test cube incoming
View attachment 2312

Nice looking filament feed! Thanks for the update. I’m afraid that tight loop in reverse Bowden is going to create drag. Sorry about the unsolicited criticism, old men do that sometimes.

 

[Laserbea4k43]

Nice looking filament feed! Thanks for the update. I’m afraid that tight loop in reverse Bowden is going to create drag. Sorry about the unsolicited criticism, old men do that sometimes.

SCIENCE!

 

[Mike Eitel]

Sorry, but it will never be solvable without some kind of measurement and regulation.
Fillament Transport between two stations will never be raly equal. So you need a way to measure the actual difference. Often this is done by measuring some kind of loop. That position has to be given as differencial speed for the first, feeding, motor...
Etc..

 

[Laserbea4k43]

Sorry, but it will never be solvable without some kind of measurement and regulation.
Fillament Transport between two stations will never be raly equal. So you need a way to measure the actual difference. Often this is done by measuring some kind of loop. That position has to be given as differencial speed for the first, feeding, motor...
Etc..

Yeah! i've been noticing that. On my last print (a print that took 3 times longer for the problem to happen) I actually increased my measured rotational distance by 20mm to keep the filament from feeding in excess (I'm sure you know how that worked out).
I have an idea that might work for measuring both. I'll try it later when I'm not going to bed.

 

[EvilleRock]

I've been thinking about just using an Arduino, stepper driver, and some limit switch with a very light touch. It would just maintain a slack loop. Once the loop gets to some minimum length, it would feed out 10 or 20 cm of filament. The feed rate wouldn't matter much as long as it was faster than maximum extruder feed rate. I just haven't worked out the mechanics of the limit switch arrangement. This would be kind of crude and not integrated into the printer's control system. Maybe Klipper could do it instead of an Arduino, but I wouldn't know where to begin doing that.

I've been getting pretty good results by sitting the spool on a stand with bearings. This is placed just outside the enclosure where the PTFE tubing enters. I'm just using a 15cm length of tubing. Its a straight shot into the enclosure. All in all, it takes very little force to pull filament. If I start doing more TPU projects, I'd like to something using a 2nd motor.

 

[Mike Eitel]

In industrial foil (paper) applications there is a device called dancer. Basically this light rolling device is in the middle of a kind of "valley". If there is too much material loop is deep, to little bow gets flatter. So it's a kind of storage that can (if part of a regulation loop) equalizes the small differences in material flow. If you have your filament in two tubes, with an opening of maybe 10cm and you drag one tube a bit to one side by a low strength force, you can use the "bowing" to measure the force.

 

[EvilleRock]

In industrial foil (paper) applications there is a device called dancer. Basically this light rolling device is in the middle of a kind of "valley". If there is too much material loop is deep, to little bow gets flatter. So it's a kind of storage that can (if part of a regulation loop) equalizes the small differences in material flow. If you have your filament in two tubes, with an opening of maybe 10cm and you drag one tube a bit to one side by a low strength force, you can use the "bowing" to measure the force.

Thanks for that information. I think I can draw some inspiration from a dancer arrangement.

I removed all but a stub of tubing to minimize friction of TPU in the tube. The problem with that is the Y axis does a big part of pulling filament. Worse case would be pulling the entire travel distance of the Y axis (350mm) at 80mm/sec. This would require a lot of slack to be available or pulling off the spool pretty fast. I’m thinking that putting the long tube back in would be best for a dancer.

 

[Laserbea4k43]

Thanks for that information. I think I can draw some inspiration from a dancer arrangement.

I removed all but a stub of tubing to minimize friction of TPU in the tube. The problem with that is the Y axis does a big part of pulling filament. Worse case would be pulling the entire travel distance of the Y axis (350mm) at 80mm/sec. This would require a lot of slack to be available or pulling off the spool pretty fast. I’m thinking that putting the long tube back in would be best for a dancer.

I like that Idea, but I have space constraints. I'll use it, but i'm going to keep working on mine also.

 

[Laserbea4k43]

My new problem is that it snags(more like, jumbles up before the feed tube into the hotend) within the CW2. Not sure what to do at this point...