Do you guys think that upgrading the i3v threaded rods for precision lead screws would make any noticeable improvement?
Printable View
Do you guys think that upgrading the i3v threaded rods for precision lead screws would make any noticeable improvement?
I looked at doing this myself, and ultimately concluded it could be a mixed bag. In the stock i3v, the M5 rods have a pretty fine pitch. Issues with thread quality on that fine of a pitch should be proportionally smaller than with larger threaded rods like M8. The X motor mount and X idler mount should ride solidly in the vertical rails, so lateral thread issues or minor rod bends will likely just cause the M5 threaded rods to flex. IMO, heftier rods could lead to binding or actually end up leading to more print quality issues than with the M5 rods, especially if there is any misalignment between the lead screws and the motor shafts. Likewise, I think it is best to NOT add bearing guides to the top ends of the threaded rods - I want to leave the M5 rods flex if there's a reason for them to.
The lead screws are, AFIK, all multi-start, so you'd definitely pick up more travel per rotation and a faster Z movement. Although the issues haven't been with i3v printers, I've read how some people found their motors didn't have the torque required to drive the multi-start lead screws. Especially with four-start threads, backlash can become a new problem. Ball screws might be better in this regard. The i3v design has the rods spaced fairly close to the vertical rails - I was also concerned with how hard it would be to fit a lead-type nut onto the X motor mount and X idler mount while maintaining that close spacing.
I personally haven't attributed any print quality issues to the M5 threaded rods. I was looking at the lead screws as a way to obtain a faster Z movement. Rather than possibly get into new issues with the lead screws, I'm improving on the Z travel by doing things like increasing the acceleration factor on the Z motors.
Looking forward to a response from someone who has migrated to lead screws, with information on how they were implemented.
you could use geared steppers on the Z-axis with lead screws to regain your Z accuracy, that you would lose by moving to the lead screws, that wouldnt really solve any issues on either end but it "should" increase Z accuracy depending on how fine your lead screws are...
You may not need to regain the Z accuracy. At 4000 steps per mm in DEFAULT_AXIS_STEPS_PER_UNIT, the stock config has a Z resolution of 0.00025mm. I've never heard anyone say what resolution is required, but that seems crazy small. Geared motors would probably solve the torque issue though.Quote:
Originally Posted by sniffle
EDIT: For example - let's assume a lead screw with a 4mm lead (travel per revolution). Assuming microstepping is left at 1/16, that would have a DEFAULT_AXIS_STEPS_PER_UNIT value of 3200 steps per revolution * 1 revolution per 4mm lead, or 800 steps per mm. That's still 0.00125mm per step resolution.
I was considering the same thing until I read Printbus's rationale for not doing it. Thank you for saving me a considerable sum of money. Heck, for the price of one lead screw I can buy several M5 rods to have at hand in case one of them becomes suspect.
Im confused about the steps per rotation. I thought the steppers were 200 steps per one rotation. so each step was 1.8degrees (i hope im wrong cause thats not very good)
Yes, the stepper motors are 200 steps per rotation, but the stepper drivers on the RAMPS can be set to do micro-stepping, up to 1/16, which lets us achieve 3200 steps per rotation instead.
Well, I'm still interested in seeing if a different opinion comes up from someone who has actually done the retrofit. I could still reconsider doing it if the rationale is there and the mechanical aspects are reasonable. Before I understood all the details, I had picked out some lead screws and was very close to ordering them. It turns out they were four-start screws (four separate threads are cut into the screws) with a lead of 8mm per revolution. Between the torque issue and the backlash issue, my guess is I would have regretted purchasing them.Quote:
Originally Posted by kd7eir
------------------
adamflip, AbuMaia is correct on the microstepping detail. On RAMPS, jumpers set the level of microstepping. I believe RUMBA also uses jumpers. With RAMBO this is set from firmware. A related aspect to be aware of is that the torque available from the motors drops off as you increase the microstepping. In other words, a motor driven with full steps (200 steps per rev) will have more torque than one driven with 1/16 microsteps (3200 steps per rev), but at a cost in Z axis resolution. Just another detail in the potential can of worms that could be opened here.