How precise is Haydn diagonal rods when assembled
-
Hi Frederick,
I know they work and I know Haydn puts in a lot of effort to make the arms very precise.
But the balls are not that precise.
Personally I consider this an opportunity to match balls to arms to get even more precision.
I know he bought 1000 or so of the ball studs in a job lot. All he has to do if he can supply 6 rods exactly the same length is supply 12 balls of the same diameter. But since the rods will vary by up to 0.05mm maybe he would be willing to tape two balls to each rod to equalise the total length.
Simon.
Hi,
Haydn had these parts made - given his attention to detail I would not think he would be selling parts that did not perform as intended.
Ask him if you have doubts.
Frederick
-
Okay so I checked my balls (sorry)
Of the 6 spare ones I have not attached to a running printer
9.38mm x1
9.40mm x2
9.48mm x1
9.50mm x2I will check the other 12 that are currently deployed whenever I get chance, and then get the arms measured accurately, and see if a "perfect" combination can be found.
I'll raise this with Haydn - looks like he may be making the arms to significantly closer tolerances than his supplier makes the ball studs.
Sure as I said earlier it's very marginal, and I suspect it will have almost no practical consequences, but then we go to extreme lengths for precision so measure your balls people!
-
Come to think of it, is it diameter we need to be measuring alone, or distance from the mounting face of the ball stud to the end of the ball?
-
Well IMHO what we should really be measuring is probably not possible without Laboratory equipment and that is the distance from the rotational centre of the ball and the mating face to the rod end spherically God knows how you would do that tho.
Doug
-
Come to think of it, is it diameter we need to be measuring alone, or distance from the mounting face of the ball stud to the end of the ball?
It depends on how Haydn measures the length specified. If he makes it in a special jig with etalon ball studs then we can be sure this length accounts for an imperfection in end cup mating surface. If the length of rod is measured itself and specified length is derived from it then we forced to rely on fabrication precision. For first case we can get get the length difference for some deviation of the ball diameter dD as
dL = dD/2 * 1.85 for each ball. ( 0.85 is approximation for the derivative of the length from mating surface to the center of ball vs ball diameter).
For second case if a fabrication tolerance is precise we get the same result. If not it would be another story. Anyway it relates only to the absolute length. If we are interested only in relative match of lengths I think that thing would be simpler: the same ball diameters give approximately the same lengths.
-
Well IMHO what we should really be measuring is probably not possible without Laboratory equipment and that is the distance from the rotational centre of the ball and the mating face to the rod end spherically God knows how you would do that tho.
Doug
And the distance from the ball centre to the carriage or effector it is attached to. Since this seems to vary also.
This is getting more like a physics problem now.
-
When you order just the rod ends from Haydn, to make your own custom length rods, he does include a couple of his ball studs to make a jig.
-
When you order just the rod ends from Haydn, to make your own custom length rods, he does include a couple of his ball studs to make a jig.
You can use the ball studs you have and match the length for final assembly that would be a perfect case for you in some sense.
-
Well IMHO what we should really be measuring is probably not possible without Laboratory equipment and that is the distance from the rotational centre of the ball and the mating face to the rod end spherically God knows how you would do that tho.
Doug
And the distance from the ball centre to the carriage or effector it is attached to. Since this seems to vary also.
This is getting more like a physics problem now.
So many imperfections in the world! Excuse me for initiating such a lengthy discussion that may turn out as a full trash from practical point of view.
-
Thats the most interesting sort of discussion.
-
Thanks for your support! BTW how can i include smiles in posts?
-
Manually make them. Colon Capital D
-
-
I just picked 12 ball studs at random and measured them, and I'm afraid that my conclusion matches David Crocker's: I build the sets of arms so that they match within 0.05mm, and typically match within 0.03mm, however the ball studs have not been manufactured as accurately as the arms. The ball studs are manufactured in an machine shop in China, and I have been very thankful that this particular machine shop has given me very good service – I've dealt with others who were obviously not as good.
When I'm building the arms, I use G25 rated chrome plated ball bearings in my jigs, not the ball studs themselves.
This makes it easy to get the arms into and out of the jigs without varying the tension.The 12 I just measured had the following diameters:
9.43
9.48
9.51
9.53
9.54 (three)
9.55 (two)
9.57
9.59 (two)
Nine were in the 0.07mm range from 9.53 to 9.59.I also measured the distance from next to the M3 threaded stud to the tip of the ball on the same 12 ball studs. This isn't as precise as measuring the diameters, but I suspect this is more important. The results were more uniform:
12.46
12.59 (nine)
12.62 (two)
Nine were very close to 12.59, and two more were in the 0.04mm range from 12.59 to 12.62. There was only one outlier.I expect that it is more important to have this distance consistent than the diameter.
So far, I've built almost a dozen printers using these and my experience has been that all of the printers have calibrated and printed well, so I'd guess that even though this isn't ideal, practically it doesn't cause problems. That said, for $12 extra, one can get two sets of ball studs and select the ones which match the best.
If I remember correctly from what other people have reported, the microswitches we use for the endstops have 0.05mm or so of variation from one triggering to the next, and many other parts probably have comparable variations.
Next time I'm placing an order for more ball studs, I'll ask them what would be the cost/tradeoffs to make these more uniform.
I'm sorry these aren't perfect and I feel a bit embarrassed! -
Wise words sir, I wouldn't feel embarrassed in the slightest, your arm systems improve any delta printer they are attached to so the real world effect on printing is negligible. If this had a big effect I wouldn't be calibrating at 0.004mm deviation.
Maybe its just a case of as you said, selecting the best matching balls, to make up sets. It would be worth asking here of those who have a deep knowledge of the geometry involved, what is the most important factor the ball's diameter or the distance from the end of the ball to the mating surface where the thread emerges? Or a combination of both?
-
I would say the important measurement would be from the mating surface to the centre of rotation of the Ball itself.
This would then reduce the effect of any inaccuracy of the diameter of the ball by a factor of 2 so we are now into the 35-40 micron error region which in real world terms obviously doesn't make that much difference.
-
But as you said before Doug, thats not so easy to determine since the balls may not be round. Granted if they are round and a set of 12 is selected of all the same size, and therefore distance from ball's centre of rotation (which is known if they are round) and the mating surface, then bingo were golden. But are they round?
I think its probably best to supply a pair of balls matched to each rod (or arrange this yourself if you already have them, this enables the slight differences in rods to be evened out). I would also propose we organise a swap-shop so that existing users can make up a set and swap ones they don't need. Of course it doesn't matter if the 6 similar balls on the carriages are a different "size" than the 6 on the effector as long as they all create equal length rod/ball sets.
-
I would say the important measurement would be from the mating surface to the centre of rotation of the Ball itself.
This would then reduce the effect of any inaccuracy of the diameter of the ball by a factor of 2 so we are now into the 35-40 micron error region which in real world terms obviously doesn't make that much difference.
It seems to me that it is not the case. Except the distance to the center of the ball matching surface and the distance from the surface to the center change also. I gave approximating formula above, it is a factor of 1.85 so total result would be of order ball diameter not radius. After all said I should confess that now I think that the difference in the ball diameter is not very detrimental to quality of print as there are many other imperfections in effector carriadges and so on of comparable intolerance and as many people used the Haydn's rod with very positive results it say for itself. Thank you all for very fruitfull discussion!
-
I am wondering if getting hold of 12 precisely made ball bearings (3/8th inch I believe), and having them precision-tapped for m3 thread, might then provide a replacement for the current ball studs. If the range of movement is restricted configured like this, the addition of a small stack of carefully measured washers under the balls will simulate the stand-off we currently have, and provide some adjust-ability for total rod length, ball centres should be precisely the same if suitably high quality bearings are used.
To say the effect will be small and covered by other imperfections is true of some printers, but I spent a lot of time and money to make sure there are few if any other imperfections in my printer, so I am willing to see if we can find a reasonably simple workaround, to iron out this blip, which is cheap, easy and which doesn't place too much of a burden on Haydn who does an amazing job making these arms and doesn't charge a huge amount for them, for which I am very grateful. Perhaps if this approach works then that might be the way forward.
I presume I would need to carefully chose a steel (thinking chromed steel) to optimise its magnetic properties? Does hardening increase or reduce its magnetic attraction?
-
I am wondering if getting hold of 12 precisely made ball bearings (3/8th inch I believe), and having them precision-tapped for m3 thread, might then provide a replacement for the current ball studs. If the range of movement is restricted configured like this, the addition of a small stack of carefully measured washers under the balls will simulate the stand-off we currently have, and provide some adjust-ability for total rod length, ball centres should be precisely the same if suitably high quality bearings are used.
To say the effect will be small and covered by other imperfections is true of some printers, but I spent a lot of time and money to make sure there are few if any other imperfections in my printer, so I am willing to see if we can find a reasonably simple workaround, to iron out this blip, which is cheap, easy and which doesn't place too much of a burden on Haydn who does an amazing job making these arms and doesn't charge a huge amount for them, for which I am very grateful. Perhaps if this approach works then that might be the way forward.
I presume I would need to carefully chose a steel (thinking chromed steel) to optimise its magnetic properties? Does hardening increase or reduce its magnetic attraction?
Don't think this way is worth of efforts. The threads are not precise at all. They are loose and depend on many things as quality of threading, torque applied when assembled etc.
To make the final assembly precise we need some sofisticated measuring device, optical I think. A much simpler to make ball stud binning as relative sizes have matter, IMO.
