Juki vs Samsung nozzles

[Marek T.]

Hi Jason,

I guess your new machine is using Juki nozzles and I know that in your older machine you have CP40 of Samsung type.

Can you tell me your impressions in case of the runouts' differences? Is possible to clearily tell that some of systems is absolutely better?

Disappointed with Chinese Juki I consider to change the system into CP40. But Robotdigg made a notice about the holder that has freezed me a bit:

"The concentricity of screw to screw design is Not as good as Press Fit design of JUKI."

br

Marek

[Jason von Nieda]

Hi Marek,

I have not really done any runout tests on the new machine. I'm not using it for jobs - just for development right now. My focus has been on trying to get it up and running with OpenPnP. Once (or if) it's up and running I'll have to go through the normal setup and calibration suite and at that time I'll be able to take some measurements.

For what it's worth, I like the CP40 style nozzles better than the Juki. The Juki is, I think, overly complex and requires a more complex holder and change system. CP40 is super simple and just slides on and off. With that said, I don't have any personal experience with the CP40 so maybe they are junk in practice - but in theory I like them better.

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/994a3778-683c-4a71-ab8a-9585aabfb670%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Marek T.]

Understand, thanks. I also think the solution of Juki is over complex - and hard to produce perfect effect in (Chinese) practice.

The only one thing that seems be potentialy risky in CP is this screw connection holder to the shaft. I'm afraid i'ts not easy to make ideally axial inner thread (In my case I must make all the hollow shaft for my head, can't use some taken from the market with high precisely made threading). Even if it is ideal I'm not sure wether the connection of two ideal threads outer+inner can provide perfect axial set.

[Jason von Nieda]

I think a press fit CP40 adapter, or one where the threads are only used for holding and not location would be ideal. I bought a CP40 holder from RobotDigg and it was trash. The threads were very loose and the result was a very wobbly connection.

Jason

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/43e4f0a1-10b5-47b8-a64b-87ee3ef87f63%40googlegroups.com.

[Marek T.]

Wow, I just was thinking to order it from them to test. So thx for this warning!

And what have you made with this? Have you found other supplier or modified the one you got from Robotdigg?

[Jason von Nieda]

Because of this I have stuck with Juki nozzles. I have not come up with a good solution for CP40 nozzles. I can take some pictures when I get home if you like.

Jason

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/2726ad8d-c3fc-4a39-9175-eb4df48714ca%40googlegroups.com.

[Marek T.]

I was convinced that you use CP for "everyday"...

I think no need the pictures, I don't expect get known something usable from them. But thanks :-).

But I have the question:

https://www.robotdigg.com/crab/image/2016/09/24/2dc797b0066ef492441f43c2f25c37a9-300-300.jpeg

Do you think that this shifting part is easly possible to be taken off (together with the load spring)?

I consider to get off this threaded element and just do the down-inner hole (grinded precisely obviously) in my hollow shaft.

What is the shift rate (probably it should be called with other words, sorry, I hope you know what I mean) of these nozzles, same like Juki or less?

[Bernd Walter]

On Thu, Oct 18, 2018 at 07:57:14AM -0700, Marek T. wrote:
> Understand, thanks. I also think the solution of Juki is over complex - and
> hard to produce perfect effect in (Chinese) practice.

Nozzle changing is really hard with the Juki tips.
Today I've received my new holders from Betz as well as new NEMA11 motors
from Robotdigg.
So far I'm using a Robotdigg cam head with the standard 3 setscrew holders.
My nozzle tips are also all from Robotdigg, but I've recently ordered some
from another supplier, which I'm still waiting for.
I can't say anything about the tip runout because with the previous holders
I had to calibrate them anyway.
The head itself is functional, but not optimal, especially if you want a
changer.
Z height is just 16.5mm, while at least 22mm is needed to get enough
clearance from parked nozzle tips, which means I need to lower and raise
the parked nozzles by themself to use them.
One linear rail has some play, which however never got me into trouble,
although I'm running 0402 with that nozzle.
The cam reduces motor force, which isn't a problem with pick and place,
but now that I have the Betz holders I'm almost sure that the motor hasn't
enough power to release the nozzle tip from the holder.
If I would start building today, I probably would use 2 Betz heads or a similar
gear based mechanism with more Z force and movement.
Maybe on my next PnP machine ;-)

Not sure yet how I will solve my problem to have an autochanger.
I have some basic ideas, but those might be a bit too expensive to be usefull.
Next part will be to mount the new holders first and then run some jobs,
which will benefit from better nozzle usage.

> The only one thing that seems be potentialy risky in CP is this screw
> connection holder to the shaft. I'm afraid i'ts not easy to make ideally
> axial inner thread (In my case I must make all the hollow shaft for my
> head, can't use some taken from the market with high precisely made
> threading). Even if it is ideal I'm not sure wether the connection of two
> ideal threads outer+inner can provide perfect axial set.
>
>

> W dniu czwartek, 18 pa??dziernika 2018 16:24:05 UTC+2 u??ytkownik Jason von
> Nieda napisa??:

> >> email to openpnp+u...@googlegroups.com <javascript:>.

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

> >> <javascript:>.

> >> To view this discussion on the web visit
> >> https://groups.google.com/d/msgid/openpnp/994a3778-683c-4a71-ab8a-9585aabfb670%40googlegroups.com

> >> <https://groups.google.com/d/msgid/openpnp/994a3778-683c-4a71-ab8a-9585aabfb670%40googlegroups.com?utm_medium=email&utm_source=footer>
> >> .

> >> 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+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/43e4f0a1-10b5-47b8-a64b-87ee3ef87f63%40googlegroups.com.

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

--
B.Walter <be...@bwct.de> http://www.bwct.de
Modbus/TCP Ethernet I/O Baugruppen, ARM basierte FreeBSD Rechner uvm.

[Marek T.]

Hi Bernd

W dniu czwartek, 18 października 2018 18:28:23 UTC+2 użytkownik Bernd Walter napisał:

My nozzle tips are also all from Robotdigg, but I've recently ordered some
from another supplier, which I'm still waiting for.

Why have you ordered next nozzles from another supplier instead of Robotdigg? Something wrong with them?

I just need to order some 502-503 and thought to do it at them as I'm not glad of the nozzles' quality that have now (not from Robotdigg).

 

but now that I have the Betz holders I'm almost sure that the motor hasn't
enough power to release the nozzle tip from the holder.

Do you mean generally the holders with 3 balls and springed collar or something wrong with Peter's product?

Your previous holders were just 3 screws centered or 3 balls but Chinese not Peter's?

 

br

Marek

[Bernd Walter]

On Thu, Oct 18, 2018 at 10:38:58AM -0700, Marek T. wrote:
> Hi Bernd
>

> W dniu czwartek, 18 pa??dziernika 2018 18:28:23 UTC+2 u??ytkownik Bernd
> Walter napisa??:

> >
> >
> > My nozzle tips are also all from Robotdigg, but I've recently ordered some
> > from another supplier, which I'm still waiting for.
> >
> Why have you ordered next nozzles from another supplier instead of
> Robotdigg? Something wrong with them?

I've ordered from another supplier to compare them.
The others are also black and not green.
https://www.aliexpress.com/item/hot-sale-smt-nozzle-welding-nozzle-pick-and-place-juki-nozzle-model-500-to-508-for/32837486168.html

> I just need to order some 502-503 and thought to do it at them as I'm not
> glad of the nozzles' quality that have now (not from Robotdigg).

The Robotdigg worked great so far, but I can't say anything about their runout
as my current holders require calibrating them.

> > but now that I have the Betz holders I'm almost sure that the motor hasn't
> > enough power to release the nozzle tip from the holder.
> >
> Do you mean generally the holders with 3 balls and springed collar or
> something wrong with Peter's product?
> Your previous holders were just 3 screws centered or 3 balls but Chinese
> not Peter's?

Peter's holders are amazing.
The tips sit in them with tight tolerance.
Havn't used them yet, but so far they feel great.
My cam based head is just a bad choice to be used with a Juki changer.
That's why I would use a different head design for a new machine.

My current holders are the classic chinese style with have the 3 setscrews,
which you have to fiddle with to center the tip.
Nothing wrong with them if you don't change the tips very often.

[Marek T.]

Understand. I thought that maybe Pete's holder has some extremely strong spring and requires more force than Chinese version.

Don't know what you want to compare if you haven't measured the runouts of old nozzles... Don't you have an effect that the tip is swinging to the sides because of much to large diameter of the pipe holding the tip? I have this problem at 506, 507 nozzles.

[Greg Wroblewski]

Funny thing, I am going to switch to Juki from CP-40 for my second machine. Auto changing with CP-40 is not that easy, I gues it could be if you can get the XYZ precision and repeatability right.

My biggest problems with CP-40 are in vision (apart from the wobbliness, for which I figured out a compensation procedure). They are black, and parts are black, that does not help. Some manufacturers make them more matte black, some more glossy. The matte ones work better for me.

I got it working most of the time though. Only some parts are challenging, not the most common ones.

Greg

[Marek T.]

Consider it twice... Juki is probably good if original Japanese or Korean, but these which we buy have never landed next to the genuine parts.

We pay $10, original cost $150 - probably difference in quality is moreless like difference in price.

[Greg Wroblewski]

Well, in my next machine I want to have enough heads to avoid nozzle changing, so in the worst case I'll just switch them back to Samsung.

Maybe I could even mix them? CP-40 work perfect for me for the small parts with larger shiny areas, resistors, capacitors, not so much for things like SOT-23, 3x3mm inductors, when shiny space overlaps with the refracted light from the nozzle. I suspect that the angle of the light on my bottom vision could also contribute to the problem (LEDs just lay flat on the surface).

Greg

--
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/Pzl7Sk9zeEM/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/83c25256-9e1d-4ac9-be37-6eb7204e7229%40googlegroups.com.

[Marek T.]

I'm using black plate on the nozzle working as the background for the larger parts. And black is not any problem. Yes it looks like the problem with your led light reflecting onto the parts to strong. Try to add some diffusor plate over the leds and it should be much better.

[kevin]

Hi mark

I am using juki nozzle tip and my holder nozzle( i do it myself). I did not see any problem with nozzle runs out. ( it very small, 0.05mm)

My holder supper easy to change, and very tight for vacuum.

If you are interesting, i will provide drawing.
https://www.youtube.com/watch?v=IME216xviCI

[Mike Menci]

Hi - it looks like magnet is holding your nozzle body ?

[Marek T.]

Hi Kevin
Runout that we see is the sum of nozzle runout and holder runout (how centerly and tight it keeps the nozzle). If you have 0.05 total then question is where it comes from, however it's quite good if doesn't grow if you rotate the nozzle in holder.
Where do you buy the nozzles?

If you have some drawing - yes pls share. If you don't but must do this, just tell me:
- what kind of magnet you use? Dimensions and type. Does it have flat or trapezoid surface touching to the nozzle?
- do you use there some oring to minimize an air leakage?

Hi Jason
I made few experiments with magnet today. Only very small and light 0402 coils were reacting on magnet touching to the opposite end of the nozzle. No problem with this in my case if I have blowing ventouries. Nothing for parts bigger from 0603.

Br
Marek

[kevin]

Hi Marek

That is the sum of nozzle run out and holder run out.

I buy nozzle tip from china, that is not origin but i think the problem not belong to nozzle, that belong to holder.

And here is some picture and drawing for your reference.

 

[kevin]

[Marek T.]

Hi Kevin,

I have nozzles from China where the tip os wobbling 0.3mm. So runout belongs to both of the elements.

Thanks for drawing. Is it nedyme magnet or "normal". And is the magnet not touching to the nozzle at all, is there a distance between?

W dniu piątek, 19 października 2018 14:46:45 UTC+2 użytkownik kevin napisał:

[Juha Kuusama]

I tried a few suppliers for LitePlacer. The best was ”Baificar” on AliExpress, and I’ve been happy with those.

For vision, the nozzles are black, but there is no such thing as black color. :-) With bright light, they appear gray; the middle hole is really black. The software needs to look for a known size black circle, but that approach works.

There is very little runout, and I have no idea if that comes from the adapter, the thread inside the pickup tube or the bearings of the tube.

[Marek T.]

Thanks for that info Juha!

Yesterday I've searched about it over net. Surprising a bit that there are suppliers offering it for $10 and $25 as well, both Chinese nozzles not orig. Interesting if it goes about other factors or other sellers' profits...

[Marek T.]

Juha, I see they have both black and green Jukis. I've never worked with black but always green. Is there some real difference between them if it goes about the allignment? Anyway we convert green into gray and find rather pins not part contour in the pipeline, so it seems be not big difference in use. Am I wrong?

[Mike Menci]

Hi,
Just a short note
I have 6 Juki nozzles and only one of them is not metal and non-magnetic.
I have no idea how come. They all look exactly the same!

Mike

[Marek T.]

Interesting, I'll check all of mine for their magnetics on Monday

[Greg Wroblewski]

I'm not sure I understand the problem completely. I bumped my CP-40 nozzle holder in many ways by "accident", and I experimented with it a lot, and as a result it has terrible runout (like 0.5mm?), is overall very wobbly, and general total crap. Yet, because I use vision three times during the placement process, my 0603s and TQFPs are being placed very accurately, so I keep putting away the task of replacing it (which would be simple, but realigning it properly after replacement takes a good hour).

Is it mainly about the faster placing, when you skip one of the vision alignment steps?

Greg

On Sat, Oct 20, 2018 at 4:43 AM Marek T. <marek.tw...@gmail.com> wrote:

Interesting, I'll check all of mine for their magnetics on Monday

--
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/Pzl7Sk9zeEM/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/fae07048-329e-4ece-b346-fbf71d0148dd%40googlegroups.com.

[Bernd Walter]

On Sat, Oct 20, 2018 at 01:12:54PM -0700, Greg Wroblewski wrote:
> I'm not sure I understand the problem completely. I bumped my CP-40 nozzle
> holder in many ways by "accident", and I experimented with it a lot, and as
> a result it has terrible runout (like 0.5mm?), is overall very wobbly, and
> general total crap. Yet, because I use vision three times during the
> placement process, my 0603s and TQFPs are being placed very accurately, so
> I keep putting away the task of replacing it (which would be simple, but
> realigning it properly after replacement takes a good hour).
>
> Is it mainly about the faster placing, when you skip one of the vision
> alignment steps?

Speed is one aspect.
I try to run all of my parts over the bottom camera and with prerotate
runout doesn't matter much, but for 1206 it is usually ok to place them
without, unless your pick location isn't very precise.

But when you pick a part from a feeder then having a nozzle with 0.5mm offset
can ruin your day, not just for 0402, but also for 0603.
Also keep in mind that pick locations aren't very exact if you have your parts
in plastic tapes, e.g. 0603 LEDs, plus many automatic feeders add variations
as well.
You just don't want you nozzle to miss the part.
I've had my machine placing 0603 resistors sideways because of a wrong pick
location.

> On Sat, Oct 20, 2018 at 4:43 AM Marek T. <marek.tw...@gmail.com>
> wrote:
>
> > Interesting, I'll check all of mine for their magnetics on Monday
> >
> > --
> > 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/Pzl7Sk9zeEM/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/fae07048-329e-4ece-b346-fbf71d0148dd%40googlegroups.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+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/CAF3PUY_Tv_LsJny0YiUpQN8dect9hFMLcqdLDf2UeJh2z%2BxSzQ%40mail.gmail.com.

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

[Marek T.]

Runout is not important if you have all the hardware setup tuned to it, use pre-rotation feature and don't need to change the nozzles.

[Juha Kuusama]

On Saturday, October 20, 2018 at 2:02:15 PM UTC+3, Marek T. wrote:
> Juha, I see they have both black and green Jukis. I've never worked with black but always green. Is there some real difference between them if it goes about the allignment? Anyway we convert green into gray and find rather pins not part contour in the pipeline, so it seems be not big difference in use. Am I wrong?

Sorry, I wasn’t very clear. I was talking about the CP-40 nozzles; I have no experience on Jukis.

[Marek T.]

Understand, no problem. Also good to know where are good CP available if I decide to change from Juki :-).

[kevin]

Hi Marek

Sounds your juki nozzle has big problem.

I use neodym magnet, distance between =1mm

Thanks

[Marek T.]

I have some 8 different size nozzles. Very differential quality, usually as bigger nozzle 506/507 as bigger runout or wobbling - what's irritating as big nozzles we use for precise bga or other fine pitch parts. That's why I'm searching for solid suppliers as I'm convinced it's matter of source only not general problem of every these jukis.

Thanks for info Kevin, I'll build the holder like yours to test it. It's simple, as something more simple as higher chance for good working :-).

[Bernd Walter]

On Sun, Oct 21, 2018 at 03:34:05AM -0700, Marek T. wrote:
> I have some 8 different size nozzles. Very differential quality, usually as bigger nozzle 506/507 as bigger runout or wobbling - what's irritating as big nozzles we use for precise bga or other fine pitch parts. That's why I'm searching for solid suppliers as I'm convinced it's matter of source only not general problem of every these jukis.
>

I never spend much time into calibrating the runout on the bigger nozzles.
As mentioned in another thread I'm currently using holders with 3 setscrews
and always have to calibrate.
A slight runout isn't that critical with big parts as they still have a good
chance to pick the part and placement accuration is done with prerotate
support for the bottom camera.
I do the calibration with the bottom camera and rotate the nozzle to see
in which direction I have to correct.
It works reasonable well with just a few iterations.
With my Robotdigg tips however I've noticed that the rubbery part on the big
one can be noticeable off center to the vacuum hole, which makes it a bit
difficult.
By noticeable I mean on camera, it may just be 0.2mm or so.

> Thanks for info Kevin, I'll build the holder like yours to test it. It's simple, as something more simple as higher chance for good working :-).

Hope that next week I will find time to mount my new Betz holders.

[Marek T.]

There are two problems with these nozzles.
Runout and clearance (difference between diameters of tip and diameter of the pipe leading this tip).
Runout - we can process as you say.
Clearance is worse because the tip stays in some position (stabilised by the spring) which you center and later align the part arround this centre. But when the part is placed and pushed into the paste the tip with part is sliding aside because it's just wobbling in the pipe. It's horror if the clearance/wobbling reaches 0.3mm and you don't have the control on it. I say about new nozzle without any accidents, just from supplier.

[Bernd Walter]

I never had the impression that the robotdigg feel loose, but my machine
has a few unrelated precision issues because of 3D printed parts and
also slightly off rotation in both directions, which may be because I'm
using TMC2100 drivers in stealth mode for rotation and have some torque
from vacuum tubing.

[Mike Menci]

I had neodymium magnets next to nozzles on my nozzle changing holder but I removed them due to nozzles getting magnetized ....
Mike

[Marek T.]

And the parts left on your nozzles after the vacuum switched off?

[Mike Menci]

Yes - hanging sideway down from nozzle

[Marek T.]

How does your vacuum work? Are your nozzles blowing when vacuum is not on or doing puff in moment of turning vacuum off?

[Marek T.]

The same question to Kevin:

How does your vacuum work?  Are your nozzles blowing when vacuum is not on or doing puff in moment of turning vacuum off?

If you have the vaccum pump (it means not blowing or puff) - have you found some problems with nozzles' magnetism?

I have seen some nozzles demagnetisers on Ali but not sure fur whom they are designed...

[Mike Menci]

This was last year - no vacuum ++ reverse at that time

[Marek T.]

Afraid don't understand :-(

[Mike Menci]

See enclosed: 

two different size of magnets

On Sunday, 21 October 2018 20:49:52 UTC+2, Marek T. wrote:

Afraid don't understand :-(

[Mike Menci]

Non magnetic Juki - see this link https://drive.google.com/file/d/1FUTcp6mPnlZdmrG59TPddDwmbymi7dnj/view?usp=sharing

Mike

[Marek T.]

I did the same and only 0402 was picked by nozzle with magnet, and it was not easy to pick. Maybe really different materials of nozzles so?
But if my nozzles blows it can be not the problem then. I'll try to test it.

[Mike Menci]

Up to you, note that longer magnet is present to Juki more magnetic field will be transferred to Juki body and it will get worse and worse ...
Mike

[Marek T.]

I know...

[Alastair Young]

I'm still in the build stage, but I was inspired by stuff I found on the internet (https://groups.google.com/d/msg/openpnp/IG9cEUypjtM/bqkRM8PEHwAJ) to build a Juki holder similar to kevin's design. Rather than buying a proper holder I invested in a cheap set of metric reamers. The magnetism does not seem to be a problem if the two magnets are inserted in-line to each other i.e. one north and one south facing into the axis. I'm using 1/2" delrin as a material. Mounting this in a collet in the lathe and cutting with 6-blade reamers looks good to me so far. I'm not planning on placing anything under 0603. Magnets are neodymium from my fridge - 1/4" x 1/16" I think.

first video is a pre-reamer attempt. shows picking but awful runout

https://youtu.be/cSeRZBBpUVo

second attempt looks much better

https://youtu.be/q32qKCoygdU

not used in anger yet but it does hold chips

On Sunday, October 21, 2018 at 1:37:45 PM UTC-7, Marek T. wrote:

I know...

[Marek T.]

It's some inspiration to the test. But not sure if I like it that the nozzle is attracted to the wall of the hole. Have you measured the runout with some sensor?

By the way, we tested today our nozzles and found that one of 10 has magnetically neutral (magnet does not attract it) black tip. All other are attracted by the magnet and attract small parts with different power. But placed longitudinally between two magnets polarised opposite, the tip has lost its magnetism and stopped to attract the parts even when connected to neodyme magnet! Tomorrow more experiments :-).

[Alastair Young]

The nozzle is a close sliding fit in the hole, as it is reamed and not drilled so is pretty accurate. I don't feel andy movement there. As there is a magnet on each side the sideways attraction should cancel out, and as they are aligned NS-NS the nozzle is theoretically in the null point and attracted there. Depth of the magnet holes is not very precise yet - I'll need to build a jig for that.

At some point I will mount it solid and take a dial reading, but I expect that the dial indicator will itself push it to the other side of the hole for whatever slop there is in the hole, so that will measure the runout of the hole, but not the slop in the hole. Light pressure on the other side of the nozzle should give a slop/rattle measurement.

[Alastair Young]

The accuracy I am aiming for is 0603 components (1.6mm x 0.8mm) and SOIC (0.05" = 1.27mm pitch. So I figure I can probably tolerate a small amount of slop. Very interested to read above that the cheap nozzles like I have often have slop in their construction - between the plunger and the tube it runs in. However concentric you make the holder, the slop is always going to be random and impossible to compensate for, I would think. 

[Marek T.]

Exactly! (about the slop problem). So except we need have perfectly concetrated holder we need also good concentrated nozzles...

I just saw something looking like runout on your second (better) movie, that's why asked wether you have measured it or not. You never know when may need 0402 to place, it's not ufo in these days already like 01005 still is.

[Marek T.]

It's true as long as you have every nozzles with the same diameter - what is not so obvious in Chinese nozzles case. Generaly it's construction mistake that there is not any cone used to center the nozzle like in CP40 made. This was a reason why I considered to change the system from Juki to Samsung.

Unfortunately you are right, specialy when the high slop plunger-tupe allows the wobbling... More accuratley you can measure it on the bottom vision when it's mechanically neutral not pushed by anything.

[Mike Menci]

HI, 

Why not try a 7mm collet  https://www.banggood.com/15pcs-ER11-1-7mm-Spring-Chuck-Collet-with-ER11A-5mm-Extension-Rod-Holder-for-CNC-Milling-Lathe-Tool-p-1280637.html?cur_warehouse=CN

I have no problem with this and modified collet 

Mike 

[Marek T.]

I thought about it but It's too large (long) for my machine. I have retrofitted machine and limitations due to original construction.

[Marek T.]

Mike if you have this pls measure for me the total length of it (body with the top nut).

[Mike Menci]

See this data enclosed -38mm from picture min holder but you can machine it  off if  its to long for you

[Mike Menci]

For metric stepper 5mm shaft   - here ebay 5mm hole in holder  https://www.ebay.com/itm/ER11A-5mm-Extension-Rod-Holder-Motor-Shaft-Collet-Chuck-Toolholder-CNC-Milling/282826746218?hash=item41d9c9a96a:g:NwEAAOSwOYFaaAMa:rk:4:pf:1&frcectupt=true

[Marek T.]

Thanks Mike. My hollow shafts are 8mm.

[Mike Menci]

This is not a problem- make 8mm hole in holder

Mike

[Mike Mencinger]

size here

[Marek T.]

Usable drawing Mike, thank you. I have already found it with 8mm internal. Will measure my machine tomorrow if can use it.

Do you maybe know wether the holders like that have hollowed shaft or it is the rod? Can't find anywhere the picture showing it inside. I see there an entrance to the rod is hollowe but never have seen is this only entrance or all the holder is hollowed.

https://www.ebay.com/itm/C8-ER11A-100L-8mm-Collet-Chuck-Holder-Extension-Rod-Straight-Shank-CNC-Milling/201894678387?_trkparms=aid%3D555018%26algo%3DPL.SIM%26ao%3D1%26asc%3D20131003132420%26meid%3Df744572d00544ae68a1cf8659dde1dd5%26pid%3D100005%26rk%3D3%26rkt%3D12%26sd%3D112640811672%26itm%3D201894678387&_trksid=p2047675.c100005.m1851

[Mike Menci]

Marek - look for shorter one like on picture enclosed this are with hole - longer one (your ebay link) have no hole - full material - .

Mike

[Mike Menci]

Here size - Juki and Samsung will work with this Collet E11 as well.... 

Mike

[Marek T.]

Hi Mike
You read in my mind :-). The left body has original length or you have cut it at the upper side?

[Mike Menci]

I fabricated this when I was in Poland last year - Made in Poland!

Nothing is Original only Collet is China made.

Mike

[Marius Liebenberg]

Mike what size collet did you use with the Juki?

[Marek T.]

I just have measured 10 different Juki nozzles for dia held by the holder. Every of them are 0.01-0.03mm below 7.0, none was over 7.0. So we need the collet 7mm.

[Marius Liebenberg]

I tried a 7mm but it seems to go in very tight with the O ring on. And I used a small O ring that broke after a couple of tries.

[Mike Menci]

Hello 7mm is corrrct size - Collet needs a nice radius on edge going onto the Juki. Do not fasten the Nut ! Play with tightening to get best fit. I’m using original “green” O ring that came with Juki - but this one is not really important to use.
More important is that you seal the cuts in collet with some flexible Silicon to have perfect sealing with nozzle - no gaps!
Mike

[Mike Mencinger]

see picture enclosed - this silicon paste is for less friction (it is transparent)  - white silicon is regular silicon for sealing. 

Mike

[Mike Mencinger]

If you are not using nozzle exchange - you can tighten the E11 Collet nut !!

Mike 

On Wednesday, 31 October 2018 13:48:16 UTC+1, Mike Menci wrote:

[Marek T.]

Mike do you use these collets with auto changer? No problems?

[Mike Mencinger]

N1 - I do not change ! Always on - My changer location allows me only to change N2 - no problems. 

Mike

[Alastair Young]

I have my Z-axis and carriage built and mounted and put a dial gauge on the juki nozzle to measure runout and slop on my delrin-and-magnet holder. It is a big plunger dial gauge so it puts a lot of sideways pressure on the nozzle. I measured about 0.006" run-out at the tip, and I estimate slop (pushing back gently against the dial pressure) at about 0.010. So that's 0.15 mm run-out and 0.25 mm slop. The magnets seem to re-center the nozzle after sideways pressure is released so I'm optimistically going to hope that is actually negligible. Likewise the force from the magnet eliminates wobble.

As the smallest component I use is an 0603 cap, which measures 0.76mm wide I think I'm good. Of course errors are cumulative - I have no numbers on how accurate my machine is going to be.

This is as much a learning project as a practical thing, so I'm ahead so far :-)

[Alastair Young]

I have my up-camera and C axis mounted. 503 nozzle runout video. I think I can improve on this a little but I'm hoping it is close enough for 0603

https://youtu.be/mD9IxxU9Z4s

[Mike Menci]

Here the picture how Juki is fixed in real 6 nozzle machine- You can see that Juki nozzle body is actually more inside than in this Open PnP seen forum and its holders

Mike

[Marek T.]

Interresting...

[Marek T.]

Today I have got three nozzles bought from Robotdigg as they told that have best quality.

503 is normal, 506 is absolute scrap, I have sent them video showing the runout and clearance - it's not to use for anything except some large capacitors maybe.

The runout is on the tip not body. So holding the body as shown on your picture (or classic way "seen forum") would not solve anything in this case.

I'm waiting what they will say about it...

W dniu czwartek, 22 listopada 2018 23:18:51 UTC+1 użytkownik Mike Menci napisał:

[SMdude]

Some that I have bought, the hole has not been centered in the body. I used the tips from these and rebuilt some others with bent tips from "accidents".

The other way, if it is just the tip, bend it with a small shifter! I did this after the last time I crashed my 503, I think it ended up better than it originally was!

But, yes, if they will supply you with ones without runout, this would be ideal, I might even be tempted to buy some more..

Mick

[Brynn Rogers]

Very interesting indeed.  Mike, do you have more pictures like this one?   I can't quite make out how these are rotated, but I am getting some ideas on Z.

Brynn   

[Mike Menci]

Here the 2 steppers used for each nozzle: 

Nozzle shaft is hollow (vacuum to nozzle), with groves on outside and has special bearing which allows rotating to shaft min. friction with axial movement (all in one). 

Mike

[Marek T.]

I don't believe there are steppers. Are you sure not high voltage servos with encoders? It does not change mechanical idea anyway...

[Mike Menci]

one is Panasonic MSMD-50W and Nema8 for rotation (#20mm body)

Mike

[Marek T.]

Hi Folks,

So it seems I'm after discussion with Robotdigg. With regret must say I'm disappointed with them.

So I have bought from them three nozzles as said priorly, two 503 and one 506, asking them before about their quality. They said that have very good quality and even offered that may deliver sample - what I have not used, just bought the nozzles.

Both 502 are ok like all other that I have, acceptable.

506 - see attached movies and I think no need to talk about it any more, scrap. Runout is like diameter of the hole which you know what is. This runout is 2x higher than other 506 I have.

Ok, runout is not big problem since we have pre-rotate.

Worse that there is a fatal clearance. When lowered part touches to the paste surface and is pushed to be sunk in it, in the same moment slide out from the place where was put.

So I have sent them the movie and got answer:

"I think you need original JUKI Nozzle you need to buy from Japan, thanks."

When asked them about their promised quality the converstaion has broken since Friday.

I'm dispappointed not with the nozzle's quality but their service quality and behaviour on the moment.

So just be warned when order something from them. If you will be luck and get good part - fine, but if not - rather don't count on their good will.

I will keep you informed if anything will change about it.

[Mike Menci]

Marek
It is unrealistic to expect for 12usd a 1st class nozzle. When you buy from China it becomes a lottery of quality.
Bor a new hole - place a brass sleeve in the nozzle and life goes on....
Mike

[SMdude]

Marek, is it the body of that one that is bad?

In cases like that, I use the tip in another body. It is normally the tip that gets damaged, the body can be reused.

The little bit of play hasn't bothered me(yet!!) As long as it always springs back to the same position I am happy..

Mick

[Marek T.]

Mike, so I understand it's in your oppinion not realistic to produce for $12 good 506 but realistic to produce good 503 - am I correct or understood you wrong ? ;)

503 has normal small runout max 0.05mm

But this 506 has ~0.3mm. So how to use it for precise large BGA if I don't have control where it slide out while the spring squeezing? It's real problem and coming back to the stable position is not any enjoy,

They definitely are able to produce smaller nozzles but bigger. And this I don't understand, would expect otherwise rather.

I have some 3pcs 506 and every of them are visible worse than 503 but not so much like this one.

Yes Mick, I know, unfortunately I don't have broken 506 body to exchange it :-). But I will make new body here, already made it before.

Mainly I expect to get from seller realistic information before selling.

Don't tell me it's normal, it's Chinese. I trade with many Chinese guys for many years and it's definitely not that all they lies - mostly they are solid or I was lucky by now.

[Mike Menci]

Marek its a quality lottery - next time you might get opposite - good 506 and bad 503 - you get my point?
But yes in past it was worse and it is getting better (the Chinese quality) but still unreliable - you need to pay first and than you get out of the bag something...

[Mike Menci]

Someting like this would be nice to have 16 heads https://m.youtube.com/watch?v=tt4930ybuBo&ebc=ANyPxKpXSe6_EbVxbx9C0EN-8gVV95OHuGcpOqUevZWDTNY6aCg8jJppMvQ-7h-2RJMq7kIAAKdZ5ilrglV84FnXEhHI0UcpWg&time_continue=8

[Marek T.]

Yes I get what you mean, I have joked only. But I'm surprised that every nozzles I bought from different suppliers have the same characteristic: as bigger as worse. Maybe it's some genius philosophy that only small parts must be precisely placed :-).

[Mike Menci]

Hello Folks   (& Marek)

I have found a good supplier for PnP stuff from China: 

You might test him out (feeders, PnP stuff, good communication.... ) 

Jacky Gao (Sales Manager)

**************************************************************************

Jacky-JM saohar <sa...@jmsaohar.cn>

Shenzhen JM Saohar technology Co.,Ltd
Motink Technology Limited

http://www.smt2000.com/

Cellphone/Wechat: +86-18503055169

QQ: 55016787            Skype: tinkeijacky

Address: B1F, Xindeju, Lixin Rd 96#, Nanwan, Longgang, Shenzhen 518114, China

[Marek T.]

Hi Mike

What's good with him?

[Eagle Media]

Which Ideas do you have got?

Am Donnerstag, 18. Oktober 2018 18:28:23 UTC+2 schrieb Bernd Walter:

On Thu, Oct 18, 2018 at 07:57:14AM -0700, Marek T. wrote:
> Understand, thanks. I also think the solution of Juki is over complex - and
> hard to produce perfect effect in (Chinese) practice.

Nozzle changing is really hard with the Juki tips.
Today I've received my new holders from Betz as well as new NEMA11 motors
from Robotdigg.
So far I'm using a Robotdigg cam head with the standard 3 setscrew holders.
My nozzle tips are also all from Robotdigg, but I've recently ordered some
from another supplier, which I'm still waiting for.
I can't say anything about the tip runout because with the previous holders
I had to calibrate them anyway.
The head itself is functional, but not optimal, especially if you want a
changer.
Z height is just 16.5mm, while at least 22mm is needed to get enough
clearance from parked nozzle tips, which means I need to lower and raise
the parked nozzles by themself to use them.
One linear rail has some play, which however never got me into trouble,
although I'm running 0402 with that nozzle.
The cam reduces motor force, which isn't a problem with pick and place,
but now that I have the Betz holders I'm almost sure that the motor hasn't
enough power to release the nozzle tip from the holder.
If I would start building today, I probably would use 2 Betz heads or a similar
gear based mechanism with more Z force and movement.
Maybe on my next PnP machine ;-)

Not sure yet how I will solve my problem to have an autochanger.
I have some basic ideas, but those might be a bit too expensive to be usefull.
Next part will be to mount the new holders first and then run some jobs,
which will benefit from better nozzle usage.

> The only one thing that seems be potentialy risky in CP is this screw
> connection holder to the shaft. I'm afraid i'ts not easy to make ideally
> axial inner thread (In my case I must make all the hollow shaft for my
> head, can't use some taken from the market with high precisely made
> threading). Even if it is ideal I'm not sure wether the connection of two
> ideal threads outer+inner can provide perfect axial set.
>
>
> W dniu czwartek, 18 pa??dziernika 2018 16:24:05 UTC+2 u??ytkownik Jason von
> Nieda napisa??:
> >
> > Hi Marek,
> >
> > I have not really done any runout tests on the new machine. I'm not using
> > it for jobs - just for development right now. My focus has been on trying
> > to get it up and running with OpenPnP. Once (or if) it's up and running
> > I'll have to go through the normal setup and calibration suite and at that
> > time I'll be able to take some measurements.
> >
> > For what it's worth, I like the CP40 style nozzles better than the Juki.
> > The Juki is, I think, overly complex and requires a more complex holder and
> > change system. CP40 is super simple and just slides on and off. With that
> > said, I don't have any personal experience with the CP40 so maybe they are
> > junk in practice - but in theory I like them better.
> >
> > Jason
> >
> >
> > On Thu, Oct 18, 2018 at 9:12 AM Marek T. <marek.tw...@gmail.com
> > <javascript:>> wrote:
> >
> >> Hi Jason,
> >>
> >> I guess your new machine is using Juki nozzles and I know that in your
> >> older machine you have CP40 of Samsung type.
> >> Can you tell me your impressions in case of the runouts' differences? Is
> >> possible to clearily tell that some of systems is absolutely better?
> >>
> >> Disappointed with Chinese Juki I consider to change the system into CP40.
> >> But Robotdigg made a notice about the holder that has freezed me a bit:
> >> "The concentricity of screw to screw design is Not as good as Press Fit
> >> design of JUKI."
> >>
> >> br
> >> Marek
> >>
> >> --
> >> 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 <javascript:>.
> >> To post to this group, send email to ope...@googlegroups.com
> >> <javascript:>.
> >> To view this discussion on the web visit
> >> https://groups.google.com/d/msgid/openpnp/994a3778-683c-4a71-ab8a-9585aabfb670%40googlegroups.com
> >> <https://groups.google.com/d/msgid/openpnp/994a3778-683c-4a71-ab8a-9585aabfb670%40googlegroups.com?utm_medium=email&utm_source=footer>
> >> .
> >> 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+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/43e4f0a1-10b5-47b8-a64b-87ee3ef87f63%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.


--
B.Walter <be...@bwct.de> http://www.bwct.de
Modbus/TCP Ethernet I/O Baugruppen, ARM basierte FreeBSD Rechner uvm.

[Peter Betz]

On Monday, November 26, 2018 at 2:22:35 PM UTC-8, Marek T. wrote:

I'm dispappointed not with the nozzle's quality but their service quality and behaviour on the moment.

So just be warned when order something from them. If you will be luck and get good part - fine, but if not - rather don't count on their good will.

I will keep you informed if anything will change about it.

I've been very pleased with them myself. I've spent thousands of dollars there. They just supplied me a couple custom machined pulleys, top notch. About the nozzles, any I have had that look like they have run-out I give them a quick bump in (in Z) with my finger and they center. I've bought quite a few of them myself, all sizes. 

I've been tossing the idea around of designing a spring loaded CP40 nozzle holder. Would combine the best of both worlds in my opinion....

Peter.