Hi all, I recently started a project involving multiple CNC axes based on the C-Beam design. Today I finally managed to run the first tests and I noticed that especially one of the leadscrews (1000 mm) starts oscillating violently once the carriage reaches approximately half of its stroke. Some specifics: - OB Nema 23 Stepper motor 2.8A@24V - Wantai DQ542MA, 6400 pulse/rev - speed 2000 mm/min (250 RPM) Did anyone encounter a similar issue? Is there a way to attenuate the issue (beside running everything slower)? Any hint would be appreciated. Pietro
There are two ways to deal with this. As you guessed, one of them is to run everything slower. When I run a 1m lead screw, I typically max out at 650 mm/min to ensure that it is always reliable. I found that I was able to go higher some of the time but occasionally there would be a problem and so it was better to go slower and more reliable. The oscillation you noticed is the reason. The other way is to improve the support of your lead screw. The end plates of the C-Beam act as pillow blocks. If you fashioned your own pillow blocks and placed this second set a bit closer together (>8mm away from the end plates), then each end point would be supported by two bearings. This double support on the ends increases the speed at which you can rotate the lead screw without having problems. Or at least this is what I have read. For myself, I put up with running at slower speeds than I'd like for a while, then upgraded to a rack and pinion setup I designed when I needed to increase throughput later on. -D
I can't fathom how having a second bearing will do anything more. What you should do, if you haven't already, is operate under tension. The standard way of building an OpenBuilds lead screw driven function is to have the lock collars between the support bearings. If you move them to the outside, you can "stretch" the lead screw. The only solution for whipping is to get a larger diameter lead screw or provide physical support along the length. The physical support is out of the question for us. Working with what we have, building it under tension is the best way to go.
I'm not an expert. I've seen the idea of lead screw end 'fixity' several places and they distinguish between the single and double-bearing scenarios. For example, take a look at this link: Selecting a Leadscrew Part 1: The “Must Knows”. It makes some sense to me in terms of the geometry. Any play between the rod and the bearing in a single-bearing scenario can have magnified effects as the lead screw length increases. But if there are two points at either end where it is attached, then there is much less potential freedom of motion which means less oscillation. But, as I said, I didn't try this out on my own machine. I just lived with slower movement. -D
Ok, after a few tries here's the result: changing the bearing-collar configuration slightly improved the situation, but just slightly (and the screw is actually to short to do it properly). I than added one bearing 100mm away from each end, mounted on a PMMA support (easy to laser cut). With the intermediate supports I could spin one of the two axes of the gantry up to 6000 mm/min without significant issues (some vibrations but, it's still a 1 meter long 8mm screw..). The other leadscrew turned out to be just too much deformed to spin properly: didn't have a dial gauge at the hand, so it's hard to quantify, but is deflected enough to clearly see it. Now my question. As I see it, I could order a new leadscrew (maybe a couple) and replace the deformed one, would just take a week for delivery and a cost of <40 EUR, or I could try to straighten it back: did anyone try to go the latter way and got positive results? If so, how?
If you are considering ordering, then take a look at the 3/8", 10mm, 12mm, or 1/2" options. The 1/2"is the most common.
Yes, try straightening it first. Not sure how perfect it you can get it but that would always be the first place to start. Which gantry plate are you using with the C-Beams? With the XL gantry plates you also have another option for stabilizing the screw using, using 2 threaded nuts.
Hi Rick, thanks for the advice. The carriages I'm using are built on the XL gantry plates, already mounting to nuts. I will try to straighten the leadscrew next week, probably starting at both ends and making minor corrections every 100 mm by hand or with a drill press.
Hate to resurrect a month-old thread, but I had the same problem today with 1000mm actuator (1040mm lead screw, and one anti-backlash nut block). Following OpenBuilds recommended method of putting the lead screw under slight compression, always caused a perceptible wobble in the long lead screw when jogging the gantry. I got it as close as possible, then mounted the stepper. I still was not completely happy. Then it struck me: Once the stepper is installed, there will be no fore/aft movement of the lead screw even if the opposite-end lock collar is loose! [I do like the idea of putting the bearing and lock collar on the outside to put the lead screw under tension, but I don't want to cut down my C-Beam and lose travel capacity, and 1040mm is the longest 8mm lead screw the OB carries, so.....] Here is the solution that worked for me. Build as usual, putting the lead screw under slight compression to lock the bearing against the stepper-end end plate. Then mount the stepper as usual, tightening the screws all the way before tightening the set screw and clamp on the coupling. Then go back to the other end and loosen the set screw in the lock collar to relieve the compression. (Don't loosen that end cap.) Then, slide the lock collar gently against the end cap and tighten the set screw. No fore-aft movement of the lead screw and no wobble in the lead screw when gantry is moving. If the anti-backlash nut block is doing its job, this should cause no problems. Once I get this build completed I will test it under speed.
