Solidoodle 4 print issue solved (Page 2 of 2)

I had two pulleys on the motor. The outer one had a wire on it, and a spring. The wire went from the motor to a pulley block on the bottom front,  and then over another pulley on the front tensioner, and attached to the carriage. This caused the belt to be constantly loaded.

Essential it works identically to an anti-lash nut and spring on the the z threaded rod. Instead, it's done with two belts.

As for the belts. Mxl belts aren't designed for linear motion. But rotation timing. GT2 belts are. The reason we have issues with the mxl belts is that there is a bit of give. We tighten them by stretching the belts with the tensioners. This distorts the belt teeth, altering contact angle with the pulleys. This tiny slip causes the lash problem.

This is a thought I've been having lately.  Those front idlers on the 4 introduce a lot of slop into the system because the pulley screw is not supported at the rear, only at the front by 2 screws. You can imagine what this does on load reversal. Belt tension holds them in place, so when the direction changes, they... Go backward, which changes the tension in the belt. You can jiggle the pulley by hand when it's not under load (loose belt or belts off), because at least on my machine the nut wouldn't grip tightly enough on the printed part without rubbing the pulley. The easiest way to deal with this is to print idlers that don't have a tensioning mechanism, but that means you're swapping carriages, as well.

EDIT: on my phone so I can't link, but if you chop the belts and put them under the keepers it's no different than the method pirvan came up with in hacks/mods.

Can you adjust the belt tension without loosening the nut on the screw? If you can, then it is too loose and will be slop.  I can on my machine but I think my idlers are broken from extensive use. The pulley spins on that screw so it should be as tight as possible and not move forward/back even under a lot of force. If that screw moves in any way, the tension of the belt is altered and this causes all kinds of problems, because it changes the effective size of the contact area between the belt teeth and pulley. Look at the X axis and notice how firm that screw is.

I plan to replace my idlers with something different for fishing line. I have the nut tight enough that the screw cannot rotate, but I cannot get it tight enough that the screw has absolutely NO forward/backward movement without rubbing the pulley and causing the axis to bind up.

Because the screw should have absolutely no movement relative to the idler. If you can move it with a screw, don't you think the Y motor and all the weight it's slinging around can as well? Especially consider the case where the axis is traveling toward the front of the machine, where the load is not pressing the pulley bolt against the tensioning bolts as hard.  I can see my belts glide forward a hair on my left idler, but the pulley may have also been ground out of round, I haven't had a good opportunity to take it apart yet.

Any movement of the bolt pulley changes tension. This is just by definition. To have minimal lash, the belts need to retain their tension as much as possible in all directions.

Here's the bottom pulley block.

Here's the tensioner pulley.

Dual pulleys

Here's the pics of the parts in the setup.

yaa that MXLs dont have enough lash to prevent perfect circles, unless we're talking about high speed. you've seen several printers with MXL belts make perfect circles so i don't really see why people go GT2 when fishing line is so successful, and can be run far tighter than any belt could dream of... this is not to say GT2s will not give you an improvement, but fishing line will give you more of an upgrade with far less cost.

no, i have not done it to my printer. but the library at my university has a fusion 3 which uses a similar kind of line, and I've seen it fill a 12x12 build area at 300mm/s+ travel with extreme accuracy. other parts of that printer make it far more accurate than a stock solidoodle can ever be, but they could not use belts with those speeds, all else being equal.

I found the real problem tonight. Here's my results at 100 mm/s. This is with only the bearings in the tensioner as a mod, so these results could easily be achieved by a totally stock unit.

I don't know how we missed this.

I'm guessing you missed this thread
http://www.soliforum.com/topic/9165/how … r-circles/