Magnetic push rods causing fan malfunction

I'm having problems similar to the one shown in this video (not mine, but the problem's the same):

[www.youtube.com]

In case you missed it, or just don't want to watch, the magnetic fields generated by my magnetic push rods are causing my hotend cooling fan to stop (or to not start at all). I'm using magnetic rods and matching screw-in balls from Ultibots. My effector is one I'm designing myself (not the Delta Smart Effector in the video), with 51mm spacing between the balls, similar in form to this one. (The 51mm spacing matches that on the effector I used to use, so I don't need to change the carrier on the other end of the push rods.) Unlike the effector shown in the video, mine puts the balls lower, so the fan and the magnets in the push rods are even closer than in the video. Alternating north/south magnetic field orientation, as suggested by one comment in the video, can help, but the problem still occurs sometimes, typically near the end of the effector's reach, when one or the other rod gets closer to the fan.

So are there any tips on overcoming this problem? Four thoughts occur to me:

• Buy a better fan -- This is likely to be the simplest solution, if a fan that's more resistant to magnetic fields exists. The trouble is finding it; "resistance to magnetic fields" isn't exactly a widely-advertised feature of 30mm fans! FWIW, I'm using a fan salvaged from an old data center KVM switch, but I've used a fan that came with an E3D V6 clone hotend, too, with no better results.
• Relocate the fan -- Given the effector design, or that of any effector, really, this could be tricky; but putting the fan on a tall duct that carries the air down might do the trick. This would be an ugly solution, but if it works....
• Relocate the push rods -- As I'm designing this effector myself, I can widen the inter-ball distance as much as necessary, but I'd rather not do this. (FWIW, the effector I used to use was this one, which if you look at the geometry, put the balls further from the fan. I don't recall ever having a problem with it.)
• Magnetically shield the push rods and/or fan -- I know relatively little about this, but I gather that some alloys provide some magnetic-shielding effect. Any advice of things to try?

If anybody else has solved this problem, or understands the physics and design of fans well enough to offer suggestions, I'd appreciate hearing from you. Thanks!

I see more options:
Shield the fan with a common fan guard ( steel ). Quite heavy, but you can cut off most of the concentric rings except the ones in the middle.
Connect the ball studs with a steel sheet metal. That would also "shortcut" the magnetic field. N/S orientation is important then.
Rotate the fan +/- 90 degrees around it's own axis, so you align the hall sensors of the fan controller in a different angle.
All options are untested by me.

I've had the issue of fans stopping randomly due to magnetic arms. Also fans generating interference on smart effector strain gauge.

Make a fan shroud to extend the fan out or down by 10mm or so, the magnetic force drops to the square of distance so you don't have to move it much.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

The hotend cooling fan shown in the video cannot work because it is located under two ball studs. The fan should be moved to be between the ball studs in order to operate outside of the magnetic fields generated by the rod ends. After the fan has been moved the rods on each side will need to be opposite poles as they will still slow or stop the fan if the poles match.
HTH

Make sure that when you fit the rods, they have alternating N and S poles as you go round the effector. This reduces the magnetic field away from the rod ends. If you have already done that, you will need to locate the fan at a greater distance from the rod ends as has been suggested in the previous replies.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Thanks to all for the suggestions! As noted in my original post, alternating north/south orientation of the magnets does help, but it doesn't completely eliminate the problem. What seems to have done that is lampe's suggestion of rotating the fan -- I rotated it 90 degrees counter-clockwise. I haven't tried this in combination with uniform north or south orientation of the rods, so I can't say how that would work by itself, but the combination seems to be effective.

If you imagine that the effector is triangular, with the ball studs near the vertexes, then you can minimize this problem by rotating your hotend and especially its cooling fan so that the cooling fan is at the middle of any side of the triangle -- this will place the hotend cooling fan as far as possible from the ball studs.

Quick nb question, how to check if rods are alternating N and S poles? I have moved fan out to middle/ between and see reduction of this effect. I also changed my fan so it now works much better than before, however I would like to understand more as I am new to pretty much everything in mechanical engineering.

@Haydn Thank you so much for the fantastic arms, together with the smart effector they work just great Amazing quality!

Quote
leif
Quick nb question, how to check if rods are alternating N and S poles?

If adjacent rods attract each other at the ends, they are alternating. if they repel each other, they are not.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote

Quick nb question, how to check if rods are alternating N and S poles?

I'm pretty sure, if you'd make the arms float on water, they will align N/S after a while.