Quick change juki nozzle holder test information

[Peter Betz]

Hello all!

I just thought I would share some information regarding the newest juki nozzle holders. As you may recall, we moved from interference fit to shaft fit. As I was testing some of the nozzle holders I have against a bunch of motors I just received I was a little concerned to see that the motor shafts on the motors vary in diameter very slightly. This resulted in some of the nozzle holders being slightly loose on the shaft. 

For reference, here is a video showing the runout on the actual motor shaft. Every motor I have tested has ZERO runout:

https://youtu.be/NFtHB6LwNPs

I went out and bought some green loctite as suggested by the machinist to try and center the holders. I wanted to take a video to show the worst example of runout when the single setscrew was tightened:

https://youtu.be/OUg9y8eJ3Uw

In that video there is about 0.03mm runout (0.0011"), that is the worst I ever saw during testing. During the course of trying to find the worst possible position for the video I made an interesting discovery:

https://youtu.be/ASCObmrOjvw

The runout practically disappeared.... That is less than 0.01mm (0.00039"). I was actually able to quickly clock every nozzle holder on every motor we have to below 0.01mm with very little effort. Here is another motor and different nozzle holder:

https://youtu.be/vbntKPvccwk

If you think about this it is very confusing, because the shaft itself has no runout, and the nozzle holder is minutely loose on the shaft in some cases. I can only assume there is immeasurable runout on the motor bore, and the nozzle bore, that you are able to cancel by allowing the holder to run slightly eccentric on the shaft (by tightening the screw). Anyways, this is all great news! This was designed to be concentric to 0.0005" (at significant cost) and it turns out that we can realize that actually mounted on the motor shaft!

FOR THOSE OF YOU WHO BOUGHT THESE HOLDERS, YOU SHOULD EXPERIMENT WITH CLOCKING THE NOZZLE HOLDER ON THE MOTOR SHAFT FOR BEST RESULTS. DO NOT OVER-TIGHTEN THE SET SCREW. 

Yes, that is a "cheap" test indicator. So I took another video with a nozzle mounted as a "proof of the pudding is in the eating" kind of demonstration (I tested several nozzles in the holders and optically there is NO runout):

https://youtu.be/v8l3txS41mA

I am going to buy a high quality test indicator, but the video makes the accuracy of the tool plausible. 

Hope you guys find this useful!

Peter.

[Peter Betz]

Whoops, this is the video showing worst possible runout:

https://youtu.be/OUg9y8eJ3Uw

Peter.

[Jason von Nieda]

Hi Peter,

Thanks for posting, this seems really interesting, but I'm afraid I don't understand what you mean by "clocking". Can you explain what this means a bit?

Jason

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/db82c734-3a45-4de6-aecd-94020c6a65b5%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Peter Betz]

Hi Jason,

 Sorry, I should have been more clear. I just mean loosening the set screw and rotating the nozzle holder on the motor shaft slightly to test again (trying a different “clock angle”). Repeat until the results I posted are obtained. Keep in mind that doing noting, on the worst possible combination of holder and motor we have, resulted in runout of just over 0.001”, with a few rotations that was easily reduced to 0.00039” or better. 

Peter.

You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/G1eKJN4OJN8/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CA%2BQw0jzYLaM9YT%3Do5ZCBmuttNQgRXATMRjz-oMU1eUThWuR1sA%40mail.gmail.com.

[Jason von Nieda]

Ah, I see! Nice work! 

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/A9E8B4E6-5EA5-49B4-8CA7-4E1C3B8E24A1%40gmail.com.

[Mark Harris]

Its nice to see that the extra time and money for the 0.0005" spec actually tests out :)

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+unsubscribe@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/db82c734-3a45-4de6-aecd-94020c6a65b5%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

-- 

You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/G1eKJN4OJN8/unsubscribe.

To unsubscribe from this group and all its topics, send an email to openpnp+unsubscribe@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CA%2BQw0jzYLaM9YT%3Do5ZCBmuttNQgRXATMRjz-oMU1eUThWuR1sA%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+unsubscribe@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/A9E8B4E6-5EA5-49B4-8CA7-4E1C3B8E24A1%40gmail.com.
For more options, visit https://groups.google.com/d/optout.

--

You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+unsubscribe@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CA%2BQw0jzQesqZ9qXPP63VgEJS04-Wq73T7GZaJ8-yvh_ZqVPcGg%40mail.gmail.com.

[Олег Перевышин]

Cool! Thank you for your work!

Is the latest version on the repository?

воскресенье, 12 марта 2017 г., 5:52:29 UTC+3 пользователь Peter Betz написал:

[Marek T.]

Amazing zero runout holder and great job!
Remains only the question where to take the same zero runout Juki nozzles from...

[Peter Betz]

I can’t say I’ve noticed an issue with my robotdigg ones. I have a whole bunch so I can do some measuring. Maybe I need to make nozzles too ? 😬


Peter Betz
Betz Technik Industries Ltd.
BETZtechnik.ca

> On Jan 16, 2018, at 2:06 PM, Marek T. <marek.tw...@gmail.com> wrote:
>
> Amazing zero runout holder and great job!
> Remains only the question where to take the same zero runout Juki nozzles from...
>

> --
> You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
> To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/G1eKJN4OJN8/unsubscribe.

> To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.

> To post to this group, send email to ope...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/8f27453a-5ce3-4e82-84ba-081ae10d87d1%40googlegroups.com.

[Bernd Walter]

 
On Wednesday, January 17, 2018 at 1:24:15 AM UTC+1, Peter Betz wrote:

I can’t say I’ve noticed an issue with my robotdigg ones. I have a whole bunch so I can do some measuring. Maybe I need to make nozzles too ? 😬

I'm really tempted to buy a set of your changers.

So far I have 4 nozzles with 4 different tips and I've noticed that I absolutely need all 4 tip sizes.
Those are: 502 503 505 and 506.
On the other hand almost everything on my PCBs is done with a single tip, which is a bit sad.

But the problem is that my machine is already setup with the original holders from RobotDigg.
Do you have an idea about the length difference between your changer and the standard fixed one?
I wouldn't mind it to be 0-2mm longer, but it shouldn't be longer or shorter than than, otherwise I'll
have a lot of work to reset my working height.

I also have to remove the current holders, which - I assume - are press fit.
They were already fitted on the motors, when I'd bought the head from RobotDigg.
No idea how easy they would come off.

I also have no idea how the changer support in OpenPnP works.
So far I've got the impression that each tip is assigned to a fixed nozzle.
And if I have multiple identical tips I have to configure the parts on all of them individually?

[Bernd Walter]

On Wednesday, January 17, 2018 at 3:31:26 AM UTC+1, Bernd Walter wrote:

 
On Wednesday, January 17, 2018 at 1:24:15 AM UTC+1, Peter Betz wrote:

I can’t say I’ve noticed an issue with my robotdigg ones. I have a whole bunch so I can do some measuring. Maybe I need to make nozzles too ? 😬

I'm really tempted to buy a set of your changers.

So far I have 4 nozzles with 4 different tips and I've noticed that I absolutely need all 4 tip sizes.
Those are: 502 503 505 and 506.
On the other hand almost everything on my PCBs is done with a single tip, which is a bit sad.

After a bit of investigation, I sadly have to say, that this will likely not work on my head.
With the long cam I already fear that the Z motors don't have enough force to hold the nozzle down.
But a sure show stopper is that head only has 16.5mm Z movement, and the save Z height of the other
nozzles will not give enough clearance over the changers.
It would only work if I keep the spaced very widely, but I also have to keep the out from the
general movement area, which means they have to go in the feeder area.
It is not impossible to solve, but quite complicated.
So in the end, this is all about manual change.

[Marek T.]

Peter's holder doesn't force to use autochanger... I've similar holders but made in China and don't have auto. But can't imagine working without fast nozzles replacement possibility given by boards like that.

Peter, I'll be the first who order your nozzles if you start to make them :-).
Chinese tips have runout 0.02-0.08, that's why I smile through tears what for need the holders so good like <0.01. What's more, having holder with runout and nozzles with runout - can compensate them each other rotating nozzle position in holder.
I've made my own nozzle with ideal runout but using Chinese tip taken from Chinese nozzle. Got nothing good because Chinese tips have runout already... So to have it good you would have to produce full complete nozzle including the tip - what seems be a challenge.

[Marek T.]

"given by holders like that" wanted to write, sorry - my phone's keypad knows better what I want to write :-).

[betzt...@gmail.com]

What about staggering the nozzle holders on the bed? (illustration below assumes you have the robotdigg head with 4 nozzles on 2 cams):

 

__    __

    __    __

 

 

Might take up less room?? (or you can “upgrade” to my head :)  )

Peter.

--
You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/G1eKJN4OJN8/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/0cf055cc-b14c-4481-bbaa-0d63e8812f02%40googlegroups.com.

[Bernd Walter]

On Wednesday, January 17, 2018 at 6:57:20 AM UTC+1, Marek T. wrote:

Peter's holder doesn't force to use autochanger... I've similar holders but made in China and don't have auto. But can't imagine working without fast nozzles replacement possibility given by boards like that.

Yes, I'm aware of that, but I can also manually change the nozzle tips with what I have right now.
Peters's holder is much easier to use of course and I might still buy them, but mostly because I like to have a changer one day.

[Bernd Walter]

On Wednesday, January 17, 2018 at 7:15:47 AM UTC+1, Peter Betz wrote:

What about staggering the nozzle holders on the bed? (illustration below assumes you have the robotdigg head with 4 nozzles on 2 cams):

 

__    __

    __    __

 

 

Unlike with your head, you can't raise a nozzle.
While one goes down, the other stays at max height.

The problem however is that I have no space at all for that.
My machine is running Y with a linear guide on each side.
And a gantry for X.
The head is oriented, so that the nozzles are aligned on X.
My automatic feeders are located at Y-min and Y-max and use the complete X space.
Those are the only sides I can use for automatic feeders as the Y rails and alu extrusions are in the way.
So I can put the nozzle holders at Y-min or Y-max, but loose feeder locations and even a lot of space because the holders can't be tightly packed.
Every other place on the bed is risking a collision.
Say I place it somewhere at the right side on the table at X max.
If the machine picks a part at the rightmost feeder from the Y-min bank and then from the rightmost feeder from the Y-max bank
the head will run in a direction line over the holders and crash.
One possible solution would be to change the code to avoid that area in normal movements, which is a fair bit of work.
Or another to lower the nozzle holders when not in use, which requires some mechanical work, plus a small software change to actuate the holder height.

 

Might take up less room?? (or you can “upgrade” to my head :)  )

Already looked into this - this would loose feeder locations too.
My current head has a nozzle spacing of about 30mm.
Your head has 46mm.
This isn't a big deal for 2 nozzles, but quite different with 4.
So my head would be be wider by 48mm, which reduces the addressable X range by twice of that, which is 96mm, or in other words six 8mm tapes less on each bank.
I would loose 12 8mm tapes I can feed into the machine.

I played with the idea of doing a tray changer, since I saw that Siplace waffle feeder.
This will be a complex thing on my current mechanic, but I need it to extend the work surface for trays and strip feeders.
I could also use it to temporarily locate a tool changer on the work surface.
Will have to see, right now I'm functionally happy with the 4 tips, a changer would just increase the machine speed.

[Олег Перевышин]

I modeled this solution. 6 bolts, inside the spring-loaded balls 3,175. But I do not know if it's good. Technologically simpler and components are easy to find (AliExpress).

[bearer]

Interesting idea! Could you elaborate a little on how this does toolchanging - by simply overcoming the spring tension on each of the 6 balls by means of z axis pull/push?

And you're refering to the roller type og grub scre / set screw like this?: https://www.aliexpress.com/item/UXCELL-M3-X-6Mm-Hex-Socket-Drive-Spring-Ball-Plunger-Screw-Black-10Pcs/32790065450.html

[Олег Перевышин]

Yes, the total clamp is the same as for a spring with a wire of 1 mm. This is enough to keep the nozzle and not create a big load when replacing.

Still there is an idea to use an electromagnet instead of all this, but while in the process.

Spring

Screw

Balls

суббота, 20 января 2018 г., 2:53:58 UTC+3 пользователь bearer написал:

[Mike Menci]

Hello 

The simple and accurate holder can be done with precision jaws & Collets - 
- https://www.amazon.com/Inch-Spring-Collet-Chuck-Holder/dp/B00ARB8D5G

and 
- https://www.amazon.com/Extention-Holder-Milling-Spindle-Collet/dp/B076T91RPT/ref=pd_day0_469_4?_encoding=UTF8&pd_rd_i=B076T91RPT&pd_rd_r=9XEM27TCY8PB8ZG2FMZ4&pd_rd_w=HHPzd&pd_rd_wg=MUzye&psc=1&refRID=9XEM27TCY8PB8ZG2FMZ4

Modified version you can see here: 

Mike

[Mike Menci]

and the nozzle change works: 

https://www.youtube.com/watch?v=Ctjf5pvbIKI&t=51s

Mike

[Michael Anton]

How is the nozzle held into the collet?  Is it just friction?

[Mike Menci]

Hi, 

Yes correct - in addition "O" ring is placed in the neck of Juki nozzle groove to increase friction and to have better sealing.

 See the green seal here;

[Marek T.]

Really nice idea, pity that has not come few months earlier :-).
Seems be better than Chinese holders with "click" balls where the sealing is capricious and depend on the shape of the nozzle neck (changing ie 502 into another 502 I'm getting other pressure values and diagnosed the reason is this shape not idealy the same in every nozzles). And farther it's cheaper ;-)

[Peter Betz]

Very interesting setup!

 

Marek, what you see as a better fit on the juki nozzle has to be taken in perspective with all of the other mechanical mates present in this setup.... Just finding a collet that will run 0.0005” is challenging and expensive (see hardinge for reference), let alone the holder. Then you have the concentricity of the holder bore for the motor shaft and the tolerance on the hole itself.

 

With our holders that fit on the shaft is incredibly precise, +/- 0.0005”, and if we miss, the run-out is out of limits on the whole holder. There should be room in the base of our holders for an o-ring, so the nozzle comes up and seals at the top of the bore. I will look into this more and order some to try.

 

Peter Betz
Betz Technik Industries Ltd.

 

From: Marek T.
Sent: January 24, 2018 6:55 AM
To: OpenPnP
Subject: [OpenPnP] Re: Quick change juki nozzle holder test information

 

Really nice idea, pity that has not come few months earlier :-).
Seems be better than Chinese holders with "click" balls where the sealing is capricious and depend on the shape of the nozzle neck (changing ie 502 into another 502 I'm getting other pressure values and diagnosed the reason is this shape not idealy the same in every nozzles). And farther it's cheaper ;-)

W dniu środa, 24 stycznia 2018 03:23:00 UTC+1 użytkownik Mike Menci napisał:

Hi, 

Yes correct - in addition "O" ring is placed in the neck of Juki nozzle groove to increase friction and to have better sealing.

 See the green seal here;

 

 

On Wednesday, 24 January 2018 05:49:44 UTC+7, Michael Anton wrote:

How is the nozzle held into the collet?  Is it just friction?

On Tuesday, January 23, 2018 at 1:08:55 AM UTC-7, Mike Menci wrote:

and the nozzle change works: 

https://www.youtube.com/watch?v=Ctjf5pvbIKI&t=51s

Mike

On Tuesday, 23 January 2018 15:01:28 UTC+7, Mike Menci wrote:

Hello 

 

The simple and accurate holder can be done with precision jaws & Collets - 
- https://www.amazon.com/Inch-Spring-Collet-Chuck-Holder/dp/B00ARB8D5G

and 
- https://www.amazon.com/Extention-Holder-Milling-Spindle-Collet/dp/B076T91RPT/ref=pd_day0_469_4?_encoding=UTF8&pd_rd_i=B076T91RPT&pd_rd_r=9XEM27TCY8PB8ZG2FMZ4&pd_rd_w=HHPzd&pd_rd_wg=MUzye&psc=1&refRID=9XEM27TCY8PB8ZG2FMZ4

Modified version you can see here: 

 

Mike

 

On Saturday, 20 January 2018 14:29:09 UTC+7, Олег Перевышин wrote:

Yes, the total clamp is the same as for a spring with a wire of 1 mm. This is enough to keep the nozzle and not create a big load when replacing.

Still there is an idea to use an electromagnet instead of all this, but while in the process.

 

Spring

Screw

Balls

 

суббота, 20 января 2018 г., 2:53:58 UTC+3 пользователь bearer написал:

Interesting idea! Could you elaborate a little on how this does toolchanging - by simply overcoming the spring tension on each of the 6 balls by means of z axis pull/push?

 

And you're refering to the roller type og grub scre / set screw like this?: https://www.aliexpress.com/item/UXCELL-M3-X-6Mm-Hex-Socket-Drive-Spring-Ball-Plunger-Screw-Black-10Pcs/32790065450.html

 

On Friday, January 19, 2018 at 4:35:04 PM UTC+1, Олег Перевышин wrote:

I modeled this solution. 6 bolts, inside the spring-loaded balls 3,175. But I do not know if it's good. Technologically simpler and components are easy to find (AliExpress).

--

You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/G1eKJN4OJN8/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/13ff49e3-86ec-45c7-817a-f559b6f19361%40googlegroups.com.

[Marek T.]

Peter, so you mean to use the seal on the top of the nozzle (it means above the nozzle) in your holder, instead on the neck like shown by Mike?
My Chinese holders fits to my shafts (I have other shafts, I don't have NEMA motors shafts but pneumatic system) with precise as you write ~0.0005" - I have fitted my shafts to the holders. And there is no runout coming from the fixing holder to the shaft. The problems are other and still the same: runout of the nozzle tips and not repeatable sealing nozzle to the holders depending on the nozzle piece.

Still waiting for your information that you started to produce the nozzles so ideal as your holders ;)

[Peter Betz]

Hi Marek,

 

That “neck line” is intended for retention as designed and as such we have 3 M2 balls that use that recess to grab the nozzle.

 

What I am interested in trying is placing a seal in this area:

 

 

It is an interesting thought to try and make nozzles as well. When all of my nozzles come back from the machine shop I am going to do some measuring to see about the run-out of the tip.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/4e41b46e-26c5-4880-9e91-aef911efc8e2%40googlegroups.com.

[Marek T.]

The same place to put the seal I've had on my mind :-). Will try this too but must go buy some thin o-ring as don't have there to much place. On the neck cannot put any o-ring same as you have there the balls,

Producing nozzles body is nothing complicated, I made some 2 pieces using the brass and they are 3x more precised than Chinese. The main goal is to make "zero" clearance pipes for tips, it's easy.
The problemable are the tips to produce them and make them hard as should be. No big sense to use original tips because they are not "straight" and are source of the runout too (however much lower if kept in the zero-clearance pipe).

[TheCunningFellow]

Has anyone investigated what part of the knock of juki nozzles are not concentric?

May change your ideas about where they need to be located from.

I may sacrafice one and clock it up to see what parts are out of concentric if no one else has already looked.

[Peter Betz]

Good point, an excellent thing to investigate. I don’t think you need to sacrifice anything. There is a set screw that lets the whole thing come apart if I am not mistaken.

[TheCunningFellow]

I thought the green plastic may have been molded on and need to be broken off.

I'll get out a collet chick and the scary clock that moves when you breath on it later and have a look.

[Marek T.]

There is not too much to investigate:
- the pipe holding the tip is not concentric,
- internal diameters of these pipes are to large in acc to the tips diameter. In effect the tip is little leaning to one side instead to be ideally centered.
- the tips are often not straight, what still more increases the runout
The green part is not construction part important for the runout quality. You can take it off and the nozzle mechanically will work the same. You can buy these plastics in China

[Marek T.]

It makes impression like someone produce 1000 nozzles and 100 of them passes the tests as good then sold as "original Juki" at $150, the rest 900 fails while test and sold for us $15 as "Chinese Juki" :-).

[Mike Menci]

Hello, 

I found some old photos from my measurements of Juki nozzles 

I measured nozzle needle part OD = 2.5mm  and Juki body part ID= 3.2mm 

This gives a gap total = 0.7mm or abt 0.35 mm of run-out - depending on SPRING push. 

Security bolt is preventing needle to fall out and it is located on one side of Juki body and spring might push needle to be out of centre. Due to this large gap it makes no sense to measure eccentrically.  

I had only one set of Juki nozzles and they were all the same. As there are many copies of nozzles out on market it would be interesting to see the measurements of original Juki nozzles if sombody could do !?

Based on above I started looking for a solution based on STANDARD parts on market and I came to this collet holder solution which works fine for me with minor modifications: 

- manufactured new holder - smaller in size, 

- new nut /smaller in size 

 

Mike

[TheCunningFellow]

Mike - there is a very obvious step in the inside back of the juki I have in front of me here.

Using a caliper to measure the back hole I get 3.7mm

The hole at the front after taking it apart is 3.20 on the nose with calipers.

The shaft is 3.19 with calipers.  I am not going to bother to mic it because I can't take the internal measurement any better.

once taken apart and without the spring getting in the way of feel - the fit on the shaft into the hole is really quite nice.  Feels like a pneumatic cylinder.

In fact if I plug the holes on the back and put a thin layer of blue tak over the front it creates suction as I slide the "piston" in and out.

If there really was a total gap of 0.7mm in your holder I don't think you would be able to apply vacuum to pick up parts.

[TheCunningFellow]

Didn't bother getting out the accurate one micron per division DTI - just used the chunky 0.01mm one that lives in the dust.

Anyway - the only face I could measure any runout with the coarse 0.01mm instrument was the inside bore at the back of the holder.  And that bore you can see is out with your eye.

I now think the 2nd chucking is just to drill the wider hole at the back and that all the outside faces are done in one set-up.

Maybe I got lucky with the Juki nozzles I bought from robotdigg and other people have runout between the body and tip.  I really doubt it could be caused by poor fit of the bore of the holder and the moving shaft.  That would just let air leak and the nozzle would not pick parts.

I think if other peoples robotdigg nozzles are as well made as mine and runout is still a problem then the solution might be to change the design of the nozzle holder to locate on the 9.97mm D as well as the 6.97 one

The screw on the 9.97mm face is recessed.  That might be for a reason.

[Marek T.]

This 0.7mm I haven't measured in my holders. But don't think it is so many. Rather would tell it's closer to 0.2mm as feel the gap.

[Mike Menci]

Hello, 

I checked again and the top neck has a larger outlet and this is what I was measuring on photo above abt 3mm ... 

The needle should be 2.5mm diameter and the neck is about 2.7 or 2.8 mm which is 0.2-0.3 mm gap.. still a lot in compare to 0.0xx mm 

Mike

[Marek T.]

This internal diameter depends on the piece! I get other runout at the tip's end when replace the needle with taken from other nozzle! And as said, this boring is not made centerly in my opinion, at least not central in the pipe. Take off the needle and see the walls of the pipe that are not equal from every sides.

[TheCunningFellow]

I clocked the inside and outside of each part.  Everything apart from wider bore at the back were pretty good.

I cant believe there is a 0.2mm gap between the shaft and the bore and the part being able achieve enough vacuum to hold a large IC.

I just re-measured mine to make sure.  And granted this is only with calipers and not a micrometer.  But the LCD display says 3.20 for the internal bore and 3.19 for the shaft.

That puts it in the range of between a "close running" and a "sliding" fit - and that makes sense to me.

0.2 to 0.3mm on a 3mm shaft puts you outside what is acceptable for a "loose running" fit.

[TheCunningFellow]

Some numbers to see if the 0.2mm clearance makes sense

My smallest nozzle has a 0.6mm hole in the centre to pick up parts.

0.6mm diameter = 0.3mm radius so Pi()*R^2 = 0.28mm2 for air to flow when picking a part

A 3.2mm hole has 8.04mm2 area.  A 3.19mm hole has 7.99mm2.  The gap around the edge is 0.05mm2.  That would work.

A 2.6mm hole and a 2.5mm shaft (being generous and only making the shaft 0.1mm smaller = 0.4mm2 path for air to flow around the outside.

Having the leakage path 2x the size of the desired path doesn't sound workable to me.

Lets go full bore and say the shaft is 2.5 and the hole is 2.8.  That makes a full 1.2mm2 worth of leakage.

[Marek T.]

The sizes wich I told here above were from my memory... it was some time ago when I've measured it.
Read all the thread on the list:
https://groups.google.com/forum/#!searchin/openpnp/juki|sort:date/openpnp/iP5f4fOFwag/WPQOwVe4AAAJ
At the beginning I have put there some time ago everything that I have measured about my Chinese Juki nozzles, together with some simple drawings.

[TheCunningFellow]

OK - I get it now.  Was a language barrier problem.

You where talking about the runout of the molded tip at the front being 0.1mm not any of the metal parts.

I can see how the getting the holder runout perfect is a "loosing game" if the plastic tip at the front is terrible anyway.

[TheCunningFellow]

BTW - just checked my small ones from robotdig.

Again - think I must have got lucky.  mine is only out by 0.04 at the tip.

[Marek T.]

If you say about runout 0.4mm not 0.04" and you mean the runout on the absolute end of the nozzle needle - you are really lucky.
The best one that I have is ~1/2 of sucking hole diameter runout - I see it on the bottom camera when rotate the nozzle, and know the runout doesn't come from the holder.
So as this diameter you told is ~0.6mm then the runout is some 0.3mm. Pre-rotation allows to live with this, without per-rotate I couldn't place any parts requiring bigger rotation :-(.
I have some 7 nozzles and all them are worth the same sh**. All they are not from Robotdigg, I see must buy something from them to test!

[SMdude]

I found with my nozzles that most of the runout was in the nozzle body.
I could take the nozzle tip from a bad body nozzle and put it in a good nozzle body and runout was good again.
There will be a bit of play between the nozzle tip and nozle body, but the most important thing is that its resting location is the same each time the tip is depressed into the body.

I have fitted an oring to the top of where the nozzle sits in the holder. It needs to be just right otherwise it pushes too hard against the nozzle when it is inserted and it is hard to seat the nozzle properly. I found a generic oring in my kit and cut a small spacer to go behind the oring to get just a small amount of pressure on the nozzle.

Another problem is some of the nozzle bodies are not quite right, as in the depth from the mounting shoulder/face to the groove that the balls sit in are different, so with these nozzles, you either need to machine the nozzle body or machine the holder a little. Argh!!!! :D

[Mike Menci]

Hello 

There is a solution to machine out a bit more in nozzle body and insert a brass tube to compensate for gap- google for  "Trumpeter Hobby Brass Pipe Set - 1 to 6 different size..."  Precision tube with very small gap.... 

Mike

[Marek T.]

To machine the nozzle body is difficult. The material is very hardened, if you would try drill out, it's hard to find the bore hard enough - I've tried and get only many sparks only.

[Marek T.]

The hole in the pipe leading the needle is not drilled axially through the body. So reduction the gap is for nothing, the needle will still have the runout even if zero gap. All this bodies which I have are not worth to machine them, best place for them is the garbage.
Also these needles are not straight, so producing new bodies to use them with Chinese needles is also not giving warranty to get amazing final result.

[SMdude]

LOL the last couple of nozzles I bought I used the tip only, in another holder because the body was garbage and fit for the bin!!
We spend so much money on trash sometimes... Last year, I was a Chinese dumping ground :(

[Marek T.]

Exactly Michael, I have the tip 502 in body 501, 503 in 502 - this way collected the best pairs needle-body :-). Also have some body with hole so non-centered, actually even seems be oval, that it's not for use at all.

[Mike Menci]

Hello 

With 7mm  (or 1/4"  up to 9/32") Collet header you can use as well Samsung nozzle holder and any other nozzle with cylindrical shape of body up to size of your conical part in the header. 

Here photo of Samsung holder CP40 inserted in collet in compare with Juki. 

There are several types of collets available and different manufacturer ..... 

A mix fof some here; 

https://www.ebay.de/sch/i.html?_from=R40&_trksid=p2380057.m570.l1313.TR0.TRC0.H0.XCollet+7mm+.TRS0&_nkw=Collet+7mm+&_sacat=0

Mike

[Bernd Walter]

On Friday, January 26, 2018 at 10:18:28 PM UTC+1, Marek T. wrote:

If you say about runout 0.4mm not 0.04" and you mean the runout on the absolute end of the nozzle needle - you are really lucky.
The best one that I have is ~1/2 of sucking hole diameter runout - I see it on the bottom camera when rotate the nozzle, and know the runout doesn't come from the holder.
So as this diameter you told is ~0.6mm then the runout is some 0.3mm. Pre-rotation allows to live with this, without per-rotate I couldn't place any parts requiring bigger rotation :-(.
I have some 7 nozzles and all them are worth the same sh**. All they are not from Robotdigg, I see must buy something from them to test!

My tips are all from robotdigg.
I can't say anything about runout, since I'm using the simple grub screw based holders and had to calibrate them anyway.
So far they run perfectly on my machine, but I have to say that they all have a bit of play.

I also wouldn't say that runout is completely without problem when using prerotate.
Prerotate only handles placement error, but you still have to pick the part.
And if speed is an issue, you want to run less parts over the bottom camera.

W dniu piątek, 26 stycznia 2018 21:38:09 UTC+1 użytkownik TheCunningFellow napisał:

BTW - just checked my small ones from robotdig.

Again - think I must have got lucky.  mine is only out by 0.04 at the tip.

On Saturday, January 27, 2018 at 6:35:38 AM UTC+10, TheCunningFellow wrote:

OK - I get it now.  Was a language barrier problem.

You where talking about the runout of the molded tip at the front being 0.1mm not any of the metal parts.

I don't think runout of the molded tips is a real issue, the molded tips are big and pick position isn't a major deal with them.
You usually pick them from trays or plastic tapes, which have a lot of play in part position, so you end up with bottom vision anyway.
It is the metal end tips for small parts which matters most.
You can't pick a 0402 with a runout of half the part width.

[Marek T.]

Don't understand relation of runout and picking. I have the bottom camera looking to the centers of the nozzles when they are at 0 rotation. The picking is also always performed at 0 rotation. So when the nozzle is lowered for the pick the end of tip is landing in centre of the part and not any problem with the picking then, nor with later allignment.
I must only care to mount the nozzles always in the same position that pointed with marker on the body and on the holder. And I have the endstop on nozzles rotation.
Prerotation is eliminating an issue of xy not intended moving while the bigger rotations, it's obvious it's "only" for this.
I don't know why but as bigger nozzles I have as bigger runouts they have. So the same problem both for small and big parts.

[Bernd Walter]

You have to home the rotation, otherwise you pick at a random orientation, although it stay the same during the runtime of your machine.
Most machines have no support for that.

In a machine with a feederbank on each side one is 180° different than the other.
Don't know if they are picked at 180° or OpenPnP tracks the 180° Offset internally.
This matters if you want to setup your pick position with the head cam or even use multiple tips for the same parts.

[Marek T.]

Homing of nozzles has remained from original setup of my philips machine, so I have adapted it to be served by Openpnp too. Even haven't checked if it is used by everybody using Openpnp or not, I've taken it as standard... And when I made it haven't suspected it maybe later usable for "runouted" nozzles.
Don't know about feederbank problem you say, just don't have a feederbank.

[Marek T.]

Hi Peter,

Have you cheked it maybe?

I have noticed another disappointing effect with Chinese Jukies today:
I have two the same 503 nozzles.
If mount one of them, plug end of the tip with finger (or part) and measure the vacuum - get value 500.
If do the same with another 503 - get only 300! If press up the nozzle body to the holder - get visible a bit higher vacuum, like stable 330, but never 500.
So there is the leak between nozzle and holder. Rather not on the body diameter, and not on the tip which I have later removed and plugged just the tip leading pipe. So it seems be on the neck or on the balls.
Unfortunately in my holder there is no room to place an oring where you have pointed. It seems that the nozzle body just touches to the holder in this point, maybe end of this one nozzle body is a bit too long and the balls can't clatch properly - are little "opened". Hard to find it but I will try :-(.

When started measure these two nozzles to compare them, found they have a bit different shapes and dimensions. Differencies are +/- 0.05-0.1mm.
So they are the same but not the same...

br
Marek

W dniu czwartek, 25 stycznia 2018 18:21:50 UTC+1 użytkownik Peter Betz napisał:

Hi Marek,

 

That “neck line” is intended for retention as designed and as such we have 3 M2 balls that use that recess to grab the nozzle.

 

What I am interested in trying is placing a seal in this area:

 

 

It is an interesting thought to try and make nozzles as well. When all of my nozzles come back from the machine shop I am going to do some measuring to see about the run-out of the tip.

 

 

Peter Betz
Betz Technik Industries Ltd.

 

From: Marek T.
Sent: January 25, 2018 9:12 AM
To: OpenPnP
Subject: Re: [OpenPnP] Re: Quick change juki nozzle holder test information

 

Peter, so you mean to use the seal on the top of the nozzle (it means above the nozzle) in your holder, instead on the neck like shown by Mike?
My Chinese holders fits to my shafts (I have other shafts, I don't have NEMA motors shafts but pneumatic system) with precise as you write ~0.0005" - I have fitted my shafts to the holders. And there is no runout coming from the fixing holder to the shaft. The problems are other and still the same: runout of the nozzle tips and not repeatable sealing nozzle to the holders depending on the nozzle piece.

Still waiting for your information that you started to produce the nozzles so ideal as your holders ;)


W dniu czwartek, 25 stycznia 2018 17:51:41 UTC+1 użytkownik Peter Betz napisał:

Very interesting setup!

 

Marek, what you see as a better fit on the juki nozzle has to be taken in perspective with all of the other mechanical mates present in this setup.... Just finding a collet that will run 0.0005” is challenging and expensive (see hardinge for reference), let alone the holder. Then you have the concentricity of the holder bore for the motor shaft and the tolerance on the hole itself.

 

With our holders that fit on the shaft is incredibly precise, +/- 0.0005”, and if we miss, the run-out is out of limits on the whole holder. There should be room in the base of our holders for an o-ring, so the nozzle comes up and seals at the top of the bore. I will look into this more and order some to try.

 

Peter Betz
Betz Technik Industries Ltd.

 

From: Marek T.
Sent: January 24, 2018 6:55 AM
To: OpenPnP
Subject: [OpenPnP] Re: Quick change juki nozzle holder test information

 

Really nice idea, pity that has not come few months earlier :-).
Seems be better than Chinese holders with "click" balls where the sealing is capricious and depend on the shape of the nozzle neck (changing ie 502 into another 502 I'm getting other pressure values and diagnosed the reason is this shape not idealy the same in every nozzles). And farther it's cheaper ;-)

W dniu środa, 24 stycznia 2018 03:23:00 UTC+1 użytkownik Mike Menci napisał:

Hi, 

Yes correct - in addition "O" ring is placed in the neck of Juki nozzle groove to increase friction and to have better sealing.

 See the green seal here;

 

 

On Wednesday, 24 January 2018 05:49:44 UTC+7, Michael Anton wrote:

How is the nozzle held into the collet?  Is it just friction?

On Tuesday, January 23, 2018 at 1:08:55 AM UTC-7, Mike Menci wrote:

and the nozzle change works: 

https://www.youtube.com/watch?v=Ctjf5pvbIKI&t=51s

Mike

On Tuesday, 23 January 2018 15:01:28 UTC+7, Mike Menci wrote:

Hello 

 

The simple and accurate holder can be done with precision jaws & Collets - 
- https://www.amazon.com/Inch-Spring-Collet-Chuck-Holder/dp/B00ARB8D5G

and 
- https://www.amazon.com/Extention-Holder-Milling-Spindle-Collet/dp/B076T91RPT/ref=pd_day0_469_4?_encoding=UTF8&pd_rd_i=B076T91RPT&pd_rd_r=9XEM27TCY8PB8ZG2FMZ4&pd_rd_w=HHPzd&pd_rd_wg=MUzye&psc=1&refRID=9XEM27TCY8PB8ZG2FMZ4

Modified version you can see here: 

 

Mike

 

On Saturday, 20 January 2018 14:29:09 UTC+7, Олег Перевышин wrote:

Yes, the total clamp is the same as for a spring with a wire of 1 mm. This is enough to keep the nozzle and not create a big load when replacing.

Still there is an idea to use an electromagnet instead of all this, but while in the process.

 

Spring

Screw

Balls

 

суббота, 20 января 2018 г., 2:53:58 UTC+3 пользователь bearer написал:

Interesting idea! Could you elaborate a little on how this does toolchanging - by simply overcoming the spring tension on each of the 6 balls by means of z axis pull/push?

 

And you're refering to the roller type og grub scre / set screw like this?: https://www.aliexpress.com/item/UXCELL-M3-X-6Mm-Hex-Socket-Drive-Spring-Ball-Plunger-Screw-Black-10Pcs/32790065450.html

 

On Friday, January 19, 2018 at 4:35:04 PM UTC+1, Олег Перевышин wrote:

I modeled this solution. 6 bolts, inside the spring-loaded balls 3,175. But I do not know if it's good. Technologically simpler and components are easy to find (AliExpress).

--
You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/G1eKJN4OJN8/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/13ff49e3-86ec-45c7-817a-f559b6f19361%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

 

--
You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/G1eKJN4OJN8/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/4e41b46e-26c5-4880-9e91-aef911efc8e2%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

 

[John deGlavina]

Hi Peter, 

Can these holders be used on cam type heads where there is only a spring to pull the nozzle back up? In other words, once the collar is pushed up, is the force required to release the nozzle significant? 

On Saturday, March 11, 2017 at 9:52:29 PM UTC-5, Peter Betz wrote:

Hello all!

I just thought I would share some information regarding the newest juki nozzle holders. As you may recall, we moved from interference fit to shaft fit. As I was testing some of the nozzle holders I have against a bunch of motors I just received I was a little concerned to see that the motor shafts on the motors vary in diameter very slightly. This resulted in some of the nozzle holders being slightly loose on the shaft. 

For reference, here is a video showing the runout on the actual motor shaft. Every motor I have tested has ZERO runout:

https://youtu.be/NFtHB6LwNPs

I went out and bought some green loctite as suggested by the machinist to try and center the holders. I wanted to take a video to show the worst example of runout when the single setscrew was tightened:

https://youtu.be/OUg9y8eJ3Uw

In that video there is about 0.03mm runout (0.0011"), that is the worst I ever saw during testing. During the course of trying to find the worst possible position for the video I made an interesting discovery:

https://youtu.be/ASCObmrOjvw

The runout practically disappeared.... That is less than 0.01mm (0.00039"). I was actually able to quickly clock every nozzle holder on every motor we have to below 0.01mm with very little effort. Here is another motor and different nozzle holder:

https://youtu.be/vbntKPvccwk

If you think about this it is very confusing, because the shaft itself has no runout, and the nozzle holder is minutely loose on the shaft in some cases. I can only assume there is immeasurable runout on the motor bore, and the nozzle bore, that you are able to cancel by allowing the holder to run slightly eccentric on the shaft (by tightening the screw). Anyways, this is all great news! This was designed to be concentric to 0.0005" (at significant cost) and it turns out that we can realize that actually mounted on the motor shaft!

FOR THOSE OF YOU WHO BOUGHT THESE HOLDERS, YOU SHOULD EXPERIMENT WITH CLOCKING THE NOZZLE HOLDER ON THE MOTOR SHAFT FOR BEST RESULTS. DO NOT OVER-TIGHTEN THE SET SCREW. 

Yes, that is a "cheap" test indicator. So I took another video with a nozzle mounted as a "proof of the pudding is in the eating" kind of demonstration (I tested several nozzles in the holders and optically there is NO runout):

https://youtu.be/v8l3txS41mA

I am going to buy a high quality test indicator, but the video makes the accuracy of the tool plausible. 

Hope you guys find this useful!

Peter.

[Mike M.]

John,

You can use this kind of slide inn nozzle changer where there is no need for Z torque for changing nozzles - X or Y will do all the torque required movement.

https://www.youtube.com/watch?v=KRR3bC9j_3s&app=desktop

If you have two nozzle head - have in mind that you need free space for 2nd nozzle not coliding with other nozzle park location!!

I hope this helps.

Mike

[John deGlavina]

Thanks Mike. I saw that video before but wasnt sure if there is more force required to actually pull the nozzle up and off of the head.

The one problem with my head is the amount of travel is limited, so I can probably only do one nozzle change anyway.

[Mike M.]

John,

I do the same - my nozzle changer is on Y left and I auto change only 2nd head nozzle  :-)

What I came up is servo turning nozzle changer,.... this way there is no colision with 2nd nozzle - still not in function - code missing...

But I have Z torque to do the push and pull .....

Mike

[John deGlavina]

Oh wow, nice. That's a good idea. Are you planning on just hooking the servo up to a MCU to run the code? 

--

You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/G1eKJN4OJN8/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/a982b093-a5ee-469f-bcbd-3d312047ab3e%40googlegroups.com.

[Mike M.]

See Chapter Hoby Servo here; http://smoothieware.org/switch

Most sloved by Configuration just details of deviding 360 deg to 6 servo positions and Jason might do his magic ? We see how it works out...

At the moment servo I have has gap in bearings which rotates nozzles out of center - mechanical issue to solve first...

Mike

On Saturday, 7 December 2019 16:43:47 UTC+1, John deGlavina wrote:

Oh wow, nice. That's a good idea. Are you planning on just hooking the servo up to a MCU to run the code? 

On Sat, Dec 7, 2019 at 10:05 AM Mike M. <mike...@gmail.com> wrote:

John,

I do the same - my nozzle changer is on Y left and I auto change only 2nd head nozzle  :-)

What I came up is servo turning nozzle changer,.... this way there is no colision with 2nd nozzle - still not in function - code missing...

But I have Z torque to do the push and pull .....

Mike

On Saturday, 7 December 2019 15:49:30 UTC+1, John deGlavina wrote:

Thanks Mike. I saw that video before but wasnt sure if there is more force required to actually pull the nozzle up and off of the head.

The one problem with my head is the amount of travel is limited, so I can probably only do one nozzle change anyway.

--
You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/G1eKJN4OJN8/unsubscribe.

To unsubscribe from this group and all its topics, send an email to ope...@googlegroups.com.