PCB Effector with embedded sensor

Quote
Mutley3D
...........................

@leadinglights, im sure im reading your comment incorrectly. Your not saying brass is magnetic are you. Am I correct interpreting that your iron cups were disrupting a magnetic field reducing its strength?

That's what makes it so odd. Iron is magnetic but a magnetic field finds it easier to pass through iron so it goes up the sides of the cup. This really needs a drawing or even better, a video but the effect is that the ball clings to one side for the first maybe 1/10 of a mm then the next 2 or so mm riding up the side - the pull gradually weakens until it disconnects completely. The iron cups were not reducing the magnets strength, just it's effectiveness - like a strong army pointing it's weapons in the wrong direction.

Mike

Damn! I will have to do some drawings.

Quote
leadinglights

Quote
Mutley3D
...........................

@leadinglights, im sure im reading your comment incorrectly. Your not saying brass is magnetic are you. Am I correct interpreting that your iron cups were disrupting a magnetic field reducing its strength?

That's what makes it so odd. Iron is magnetic but a magnetic field finds it easier to pass through iron so it goes up the sides of the cup. This really needs a drawing or even better, a video but the effect is that the ball clings to one side for the first maybe 1/10 of a mm then the next 2 or so mm riding up the side - the pull gradually weakens until it disconnects completely. The iron cups were not reducing the magnets strength, just it's effectiveness - like a strong army pointing it's weapons in the wrong direction.

Mike

Damn! I will have to do some drawings.

OK I think I understand what you are saying. Iron cup was distorting/reducing magnet effective strength, not reducing actual magnet strength, more just its effectiveness by disrupting/soaking up the magnetic energy..

---

Flex3Drive.com

I believe leadinglights made , what I called a "closed loop" magnetic path, but the steel ball was more attracted to the return path.

BTW, the air gap Mutley3D mentioned is more than a minor strenght reduction. I wonder how such system would improve with a magnetic fluid filling the gap between steel ball and cylindric magnet? ( eg. oil& nano iron particles )

Edited 1 time(s). Last edit at 08/08/2017 01:47AM by o_lampe.

I've thought about that the Haydn system does leave a tiny amount of delrin on the cup in the arm between the steel ball and magnet. But I'm not sure I'd want to fill it with anything abrasive. Would oil and graphite powder be magnetic? How about graphene? I haven't come up with a use for the new wonder -
substance in 3d printing yet.

Edited 2 time(s). Last edit at 08/08/2017 02:03AM by DjDemonD.

---

Simon Khoury

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

@o_lampe - glad someone understood and acknowledged the point I was trying to gently present. The air gap is indeed significant.

What is also significant, is the hyptothetical (minimal) contact area between ball and flat surface on end of cylinder magnet, where one will essentially have an increasing airgap around spherical surface of the ball joint, compared to cup magnet.

A good grade cup magnet of correct size to fit the ball, can have advantages over the alternative cylinder magnet (in rod end), given the size constraints in the application. Inferior parts will always result in poor performance.

---

Flex3Drive.com

Quote
Mutley3D
A good grade cup magnet of correct size to fit the ball, can have advantages over the alternative cylinder magnet (in rod end), given the size constraints in the application. Inferior parts will always result in poor performance.

That's true but I think this discussion is rather a mix of theoretical discussion on magnetic joint systems, and practical experience with the various systems currently on offer. I'm talking about my practical experience of k800 versus Haydn's.

Can you put forward a cup magnet system that's currently available which performs as well as the Haydn magnet system? Can you make one? If so you can probably sell it to delta printer owners if it works better. I think there is a reason why the most popular magnet system currently in use amongst the serious delta printer owners is the Haydn system, and its not that necessarily it is magnetically the strongest theoretical configuration, its that it works, its very well made (rods guaranteed to 0.05mm accuracy on length), its quite reasonably priced, and the delrin ends mean there is no issue of wear on the balls/cup shaped ends of the rods/or magnets that is present when cup magnets are directly fixed to the balls.

This started as Chris said above, he was going to do away with magnets and go back to mechanical joints, which I think would be a retrograde step, there will always be backlash and therefore effector tilt with mechanical joints compared to magnets. Maybe he might be better off trying a better magnet system, if I recall he was using K800 at one point like I was, they don't work as well Haydn's do.

Edited 1 time(s). Last edit at 08/08/2017 03:56AM by DjDemonD.

---

Simon Khoury

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

@djdemond - All based on practical application after testing theory. I supply an end effector with option for ball or cup magnets. Reason being, the balls can be screwed flush into an accurately turned down arm. The ball studs have a flat machined surface so will be accurate on the rods. This gives options for different setups.

The cup magnets I use have quite a lot of force on Haydns balls i'd say they easily hold at least 0.5kg individually ball to cup. It almost feels closer to 1kg. I can do a short vid of an experiment to demonstrate when i get a few moments to determine what force they do hold with. so just to clarify, I offer two options, both using Haydns ball studs.

note these are not the same cups that Chris was using.

My comments are not in response to Chris' decision to move to traxxas joints, the discussion of mag joints, and pitfalls/strengths etc started much earlier in the thread.

Edited 3 time(s). Last edit at 08/08/2017 05:33AM by Mutley3D.

---

Flex3Drive.com

I don't know if cup magnets are different from a magnet with an attached iron pole piece but I don't think they would be much different.
I did a quick drawing of the sort of thing I tried. You can see thatthe magnet is still attached to the ball in the second set, it has just moved sideways and up



The difference between the force on the first one and the second is minimal.

I then checked out the magnetic fluid idea - the fluid was a bit old but I think these photos give the idea. With a dry magnet the ball stuck with lots of force but with the magnetic fluid it was much less sticky.



In both cases the magnetic field has an alternative place to go so the disconnect from first position to second is much less

As Wishbone said in an episode of Rawhide, "It ain't what we don't know that's the trouble, it's what we do know that jest ain't so"


Mike

you get rid of movement with traxxas ball ends with springs and i switched 1 not because of the k800 magnet cup system it would print and stay put but wasn't confident it would stay on going faster at far points of x or y axis 2. i wanted at 1st a mini kossel i could change the head to different tool, like i said the only reason why at 1st i chose magnet system. but im not having it as a multi purpose machine no more. 3 i could of changed the cup magnets to stronger ones loosing a couple of mm of height or i could of put another magnet behind the ball opposite pole so it would of given more attraction, like i said i like the magnet idea but in my eyes would use only if i wanted to change heads for a different tool

---

Check my rubbish blog for my prusa i3

up and running
[3dimetech.blogspot.co.uk]

That was my experience with the k800 magnets they worked as long as you didn't push it too hard or go to close to the edges. Then they disconnected. Switched to Haydn's and never looked back.

I have one delta with RC joints and elastic and it performs okay but its a micro sized machine, so you can get away with way more inaccuracy than in an XL sized machine.

This thread about the tolerances of the arms and balls on the Haydn system, brings up an interesting idea, that since there appear to be a subtle range of sizes of balls, it might be possible with some careful measurement of the individual balls, knowing the arm lengths as Haydn marks them when they ship, to truly create equal length arm assemblies to a very high degree of precision by choosing balls to match arms. I suspect it will have little impact on actual printing (unless you are printing micro-sized objects with 0.15mm nozzles) but in the quest for ever lower calibration deviation, it might yield some results. I'll try it, if I ever get through the queue of objects that need printing before I start tinkering with the printer.

Edited 1 time(s). Last edit at 08/08/2017 08:32AM by DjDemonD.

---

Simon Khoury

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

Quote
chris33
you get rid of movement with traxxas ball ends with springs

A problem with using springs is that they bend the arms a little, which shortens the effective arm length. Even if all the springs have the same strength, the sideways force they exert (and hence the amount of bending) depends on the position of the arms. Hence the effective arm length depends on arm position. There is a thread in the forum at duet3d.com where a user was getting apparent high spots between the towers, and changing to weaker springs partly solved the problem.

Maybe the problem can be reduced enough to be to insignificant by using sufficiently thick and rigid arms.

Edited 1 time(s). Last edit at 08/08/2017 08:55AM by dc42.

---

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
DjDemonD
I've thought about that the Haydn system does leave a tiny amount of delrin on the cup in the arm between the steel ball and magnet. But I'm not sure I'd want to fill it with anything abrasive. Would oil and graphite powder be magnetic? How about graphene? I haven't come up with a use for the new wonder -
substance in 3d printing yet.

I think graphite and graphene are based on carbon? So they are conductive but not magnetic.

Edited 1 time(s). Last edit at 08/09/2017 02:12AM by o_lampe.

I think the reason why the magnetic fluid failed, was the direction of magnetic force. The cup pulls on the ball from all different angles ( upto +/-45° )
As soon as the forces got out off balance, the ball wanders in one direction.

With a cylindric magnet the force would come from 0° only. Right?

Fluid (or ferrofluids) have no rigidity so no structural contribution unless im overlooking something.

Edited 1 time(s). Last edit at 08/10/2017 03:49AM by Mutley3D.

---

Flex3Drive.com

Quote
o_lampe
......................... The cup pulls on the ball from all different angles ( upto +/-45° )
As soon as the forces got out off balance, the ball wanders in one direction.
......................

Almost exactly my point - my earlier analogy was not quite right, the strong army is not pointing their weapons in the wrong direction, they have left the field of battle and gone elsewhere.

In another forum they are talking about using a closed field [groups.google.com] (most recent posts) I did some preliminary trials on using a closed field some time ago but found that if the outer pole came close to the ball then the ball would rather hang out on the outer pole. If the outer pole stopped short then it short circuited the magnetic field and made everything less powerfull - there may be things between these extremes that work but probably too little to gain.

Mike

For closing the loop, I had mesh wired steel hose in mind. The stuff you often see around brake lines on motorcycles. This mesh is stainless and unfortunately nonmagnetic.
But even if we could find it in mild steel, we can't change the effector so easily with the mesh enveloping the whole mag-ball system.

yes it seem the way to go for plug and play solution
but i am scared about one thing:compatibility with marlin and smoothie board
when they say 45 without vat does it outside eu union?
so for european people it s 45+shipping cost or we need to pay extra cost

Whilst I can't say for sure, compatibility is likely to be no problem. Although the PCB effector works differently to our sensor solution they have developed in parallel over the last year and I would be highly surprised if the PCB effector does not output an endstop style signal. I.e. (preferably) Normally Closed 3.3v (would still work on Marlin/ramps) or 5v output which goes to 0 on triggered.

See here [duet3d.com]

It takes either 3.3v or 5v depending on your system to power the nozzle contact sensor, which means it modulates 3.3v or 5v to show triggering. It appears to be programmable from Duet electronics to adjust the sensitivity, but its default setting will most likely be functional when working from other controllers. This makes sense as I'm sure DC and the T3P3 team would like to see the PCB effector as part of their ecosystem (which is great BTW I have two duet controlled printers now and am very pleased with them), but would also like to sell these to everyone else too.

---

Simon Khoury

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

ok so it should be compatible
do you know if vat is already inclued for all european customers?
i think yes..but i am not sure 100%

I am not affiliated with T3P3 or DC42, but via these and other forums do know them quite well. Search through this thread [www.duet3d.com] and if the answer is not already there ask there,

---

Simon Khoury

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

The £45 price does not include shipping or (where applicable) VAT. It includes the custom E3Dv6 heatsink, carriage adapters and connector pack as well as the effector. And of course support (via the Duet3D forum) and warranty.

Whether it will work well using the default sensitivity level on a printer not equipped with a Duet WiFi or Duet Ethernet will depend on the level of vibration during a Z probing moves. The drivers on the Duets support microstepping up to x256, so it is possible to get very smooth movement and low vibration. Printers that use lower microstepping may have sufficient vibration during Z probing moves to trigger the sensor at its default sensitivity.

If you need to change the sensitivity, it would be possible to send the programming command from an Arduino. I may get round to writing an Arduino sketch to do this when I am less busy.

The firmware must support inserting a delay between a travel move and the start of a Z probing move. I understand that there are builds of all the major firmwares now with this support (thanks to DjDemonD for this!).

---

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].