are woefully undersized. After having to retighten the tiny M3 set screws on my pulleys (3GT (GT2-3M) Timing Pulley - 20 Tooth attached to NEMA 23s) after noticing machine position was inconsistent, I needed to do something. I had used non ball end Allen wrenches to get as much torque as possible on the tiny set screws. Some options: - Loctite set screws - Epoxy pulleys onto motor shafts - Pin pulleys onto motor shafts - Drill and tap pulleys to accept larger screws - A combination of the above I ended up drilling and tapping the pulleys to accept the largest feasible screws (M5). The best option I had on hand were M5 x 8mm ultra low profile cap screws, which I had to turn the heads down to clear the pulley shoulder. I've run all the same tests as before and so far the pulleys haven't moved. Wolfgang
Bigger screws are definitely a good idea. I stuck with the M3 size, but I remove the set screws and use socket cap screws and loctite. I also put one in place then using the other hole as a guide, I drilled into the lead screw the tapped that hole too and added loctite. It has been five years and nothing has come loose. On my most recent build, I marked where the set screws would hit the lead screw, the went to the bench grinder and ground a flat in those locations for both the coupler and the lock collar. That was way easier than tapping the lead screw like I had done previously. If I do another, I think I would take your advice and move up to the m5.
I had mentioned it before on this forum, I too, like Giarc, kept the m3 but use socket screws. They don't require any machining, unless they are way too long, and can be tightened really well. For what the pullies are designed to do the m3 are way adequate. If I really need it I would add a second m3 screw at 90° offset. Just MNSHO BTW, when I had mentioned that I dump the m3 set screws for socket screws other people said the do it too, like Peter and so on.
Another vote for regular M3 Socket heads... On the other end of the coin though: M3 vs M5 - keep in mind the smaller pitch, means I have more (numerically) thread engagement with M3s, than an M5 would give me (; - as there isn't a lot of meat in the shoulder sometimes, i think having more "threads" engaged feels more secure to me. And a dab of locktite (non-permanent) between pulley and shaft of course
A comparison of shear stress in the tapped pulley hole... Using the formula: tau = 2*F/(pi*d*h) M3: tau = 2*F/(pi*3*2.825) = 0.0751*F M5: tau = 2*F/(pi*5*2.825) = 0.0451*F A comparison of bearing stress... Using the formula: sigma = -4*p*F/(pi*h*(d^2-dr^2)) M3: sigma = -4*0.5*F/(pi*2.825*(3^2-2.459^2)) = -0.0763*F M5: sigma = -4*0.8*F/(pi*2.825*(5^2-4.019^2)) = -0.0408*F A comparison of shear area... Using the formula: As = pi*dm*Le/2 M3: As = pi*2.675*2.825/2 = 11.9 mm^2 M5: As = pi*4.480*2.825/2 = 19.9 mm^2 A comparison of Allen wrench hex size: M3 set screw: 1.5mm M3 cap screw: 2.5mm M5 cap screw: 4mm The M5 screw will have a larger foot print on the flat of the motor shaft, and in my opinion less bending strain on the screw which I assume leads to loosening of the screw. I think this is significant. The stresses above are approximate, but the comparison is interesting and in my opinion shows that M5 is better than M3. The down sides to the M5 are the reduced wall thickness (from the tapped hole to the end of the pulley) which could crack, and time spent modifying the pulley. But good to know that M3 socket head cap screws are good enough.
