Bent Z rods on Hadron

[peter]

Hi folks,

I've been putting together my Hadron in the evenings over the past week, and it's pretty much complete. I tried running it today with one of the Buildlog 3-axis CNC controllers that I'm using for a laser cutter, but there were some fairly wild oscillations on the z-axis. On closer inspection, the Z rods themselves appear to be slightly bent, which seems to be causing this.

One of the rods appears much more bent than the other -- I loosened off the bolts on the Z motor brackets and the delerin nut block in the hopes that everything would naturally find its spot (this assumes the rods are straight), and there was a good amount of twist on both sides, and certainly more on the particularly bad side. If the nuts are tightened, the rods bind fairly quickly.

To check if they were bent, I sat them atop the channel of some Misumi extrusion and rotated them. It's not huge, but they do seem to bow out some.

Any thoughts?

thanks,
Peter

Hadron picture

[IPvFletch]

Welcome to the club.

It seems this may be the ONLY fatal flaw in the ORD bot design... Some are exploring more robust "lead screws" (cut to length) while others are exploring bushings or otherwise nylon or rubber couplers for where the rods mount onto the motor shafts... I'm not far enough along to know how much my bent rods will impede my print-ability, but I'll know soon enough...

[Gijs]

I also think these rods are better replaced by some lead screws. It is difficult to install them properly. I've put some tape around the end to take out some of the play, but I am going to replace them with lead screws sooner or later

Also I found it easiest to bring the gantry down like in your photo, then screw in the rods from above until they almost touch the motor shafts. This way you can precisely line out the motors.

What I see in your photo is that you don't have the blue electronics plate installed. Are you sure you have everything perfectly square? I found http://www.buildlog.net/wiki/doku.php?id=ord:mainassy2 this way the best method.

But also after installing the Y axis it is recommended to check for squareness between Y and Z

All these won't make your rods straight but at least it will iron out other possible errors.

Cheers,
Gijs

[peter]

It's unfortunate to hear that this is a more common issue, and not just one that I'm experiencing. I'm pretty sure the Z axis is square, despite not having the electronics plate installed. I had a good go with a square on the bottom while I was building it to make sure, and just checked the rods again -- they're definitely bent.

I've tried to use regular ready rod (that you might find at the hardware store) before in machines, but usually come across the same problem -- it ends up being just a little bent, and causes the machine to bind. Sometimes you can go through the pile with a long nut and find one piece that just happens to be okay. On the second 3D printer we put together, that's the method we used, and stuck a flex coupling on the motor interface just to work with any other alignment issues from the cheap rod.

Are there any official plans to fix the issue? Either shipping out working rods, or creating an inexpensive parts kit that has definitely working lead screws?

thanks
Peter

[crispyfry]

It's unfortunate to hear that this is a more common issue, and not just one that I'm experiencing. I'm pretty sure the Z axis is square, despite not having the electronics plate installed. I had a good go with a square on the bottom while I was building it to make sure, and just checked the rods again -- they're definitely bent.

...

Are there any official plans to fix the issue? Either shipping out working rods, or creating an inexpensive parts kit that has definitely working lead screws?

thanks
Peter

This is why using couplers that have built-in angular compliance is best. I described why in more detail here: viewtopic.php?f=33&t=1261&start=20#p11027

People using Prusas and earlier reprap printers have been able to achieve remarkable results with non-straight Z rods.