My head design, 2+1 head ?

[Mikael Bohman]

Hi

I need some help regarding my head design. (I have access both to a CNC mill and a 3D printer)

I want my head to work with feeders where the head moves the tape as in this video youtube.

The head should also be able to automatically change the nozzles.

I'm considering adding a third pneumatic head for this, maybe with this cylinder https://www.dold-mechatronik.de/Pneumatikzylinder-DM-MAL16X50 with an 1.4 mm tip that fits the tape holes.

Is that a good or bad idea? Or should one of the head always have a nozzle that is < 1.5 mm, and only use 2 heads as in the design shown below?

Can the nozzle be damaged if they are used to drag the tape ? 

The most common parts on my PCB boards i typically 0603 (1608M)

What about limits (Zlim) , what is the best solution? I'm consider to park the head in the holder for all the nozzles before turning off the machine.

That way the machine will know it x,y,z starting position within +- 0.5 mm - and should not have any problem with collision during the startup procedure. 

[Anthony Webb]

Nice work!  I would say that using a nozzle tip for advancement is probably less than ideal.  A little pin would do the trick if you can find an elegant way to raise/lower it consistently.  I've seen designed that bump an arm to advance a part which seems to get the job done as well.

--
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/577e599d-803b-4de3-9d36-8d61a123744f%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Cri S]

Juki nozzle don't work for that.
Tm220 or tm240 nozzles works, and are used on jtronics PNP. The XS S M nozzles works, L not.
You can use XS nozzle making adapter for that and juki nozzle changer or homemade changer for tm220 nozzles . the changer must be on first nozzle.

[Cri S]

You should reconsider the design.
When you need retension the belt, you need to
remove the rails. Usually mounting the rails
at 90 degree in order there is no xy offset when
Changing Z is the most time consuming thing when mounting Z axis and usually rails cannot be mounted on machine. If hovewer you don't need that precision, OK.

[Oz-Ron]

Hi Mikael,

FWIW I consider your head design is perfect! :-)  (Thanks for your snap…) :-)

 

As for your PnP design, as Cri S has pointed out, you have made it very difficult to re-tension the belt, as will be needed in regular maintenance for sure.

 

Easy fix is to mount the motor fixed position and just slot the idler pulley mount for belt adjustment.  

 

I too am in the process of converting a cam head machine to a similar design as yours to facilitate nozzle changing.

 

Keep up the good work!

 

Cheers,

Ron

[Mikael Bohman]

Thanks a lot for your input. I will make some changes and post a new revision. 

I prefer to make the errors in a simulated environment instead of IRL. :)

[Mikael Bohman]

What about this solution with a 10 mm air cylinder!? (My dragchain has arrived and is added to the design)

[mojalovaa1]

That look nice , I m all so think add  something like  you .

On my first machine I us electromagnetic actuator but  sometimes spring is not enough strong and then have problems with tape  .
You are make that head or only design ?

[Mikael Bohman]

I will make the head in CNC Al + 3d printed plastic, when the design is mature enough to work IRL. 

[Mikael Bohman]

I bought this ebay rails to the head, being scared of the Chinese ones.

They where delivered within 2½ days to my door.

[mojalovaa1]

That is my  design head with 100 m rails with this  bearing and rail : http://www.ebay.com/itm/162023991258?_trksid=p2060353.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

[SMdude]

Hi Mikael

For the homing switch, you might want to make it so it triggers when the nozzles are centred, not fully extended as it currently is. Then when the homing operation begins, z is brought back to centre first without the risk of driving the nozzle into something while homing. Setting it this way will then make your head compatible with the setup for Peter Betz head, and the smoothie firmware will be the same also as the firmware needed to be modified to correctly home.

Mick

[Mikael Bohman]

What parts are needed for the idler pulley ? I'm not sure how to make the idler work smooth.

Den måndag 3 oktober 2016 kl. 23:51:25 UTC+2 skrev mojalovaa1:

[Mikael Bohman]

I bought a set of optical detectors, and they have now arrived. What is the precision of the optical detector?

Is only one centred detector enough , or do I need a Micro-switch in the bed as well to tune the nozzle height?

[Michael Anton]

It needs to ride on bearings.  There are special toothed pulleys available that have a couple of ball bearings inside.  Those work, or you could mount it to a shaft, and have the shaft running in a pair of bearings.

[mojalovaa1]

This is pulley with bearing of course .

[Peter Betz]

The thing you want to think about is the machine homing from an unknown state at power up. The controller needs to have a way to know which way to go in order to NOT drop a nozzle any lower than it was when the machine powered on (due to risk of collision with an object below). This is done by having the switch tripped continuously while it is on one side of center on one nozzle. Then your homing direction varies depending on if the switch was already tripped when homing was requested.

Hope this helps.
Peter.

[Mikael Bohman]

Hi Peter

I do not understand the meaning of "having the switch tripped continuously while it is on one side of center on one nozzle."

Can you explain it in other words?

My idea was to park the head in the holder for the different nozzles before power off. That way the machine will know it's xyz position within 0.5 mm at power on, and can make a safe homing.

Den onsdag 5 oktober 2016 kl. 16:49:10 UTC+2 skrev Peter Betz

[Mikael Bohman]

I found and ordered the idler pulley with bearings from ebay.

Here is my latest revision of the head bolted to the y-rail.

I changed several things. What needs to be changed here?

[Mikael Bohman]

 I'm consider to use this lightning for the downward looking camera.

 

It is the LED list (giving the screen background brightness) from a broken 15.6" laptop. 

The LEDS are in series-parallel, and full brightness is around 24 Volts.

I also got the all the diffusion materials from the screen.

The list is bendable and you can make a ring out of it.The yellow part of the LED is where the light is emitted. 

The LEDS are of a diffuse type, and I can only see a white light from them (no RGB colors is visible).

Anyone tried this before ?

[mojalovaa1]

This look very good , rendering is fantastic .
You have that pneumatic actuator ?

[Mikael Bohman]

It is this type of cylinder, but I have not bought it yet.

http://www.ebay.com/itm/10-x-75mm-Dual-Acting-Mini-Pneumatic-Air-Cylinder-M5x0-8-CDJ2B10-75-CNBTR-/361629506414?hash=item5432cc5b6e:g:1ZEAAOSw9NdXvhL4

[mojalovaa1]

I m look this : http://www.ebay.com/itm/162114112595?_trksid=p2060353.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT

[Wireb]

On the home limit switch topic some setup is needed. 

First assume that when the two heads are centered (both at the same height) that is a Z of 0. From there each head can go either +Z or -Z which causes the other head to go -Z and +Z respectively.

The point he was trying to get across was to have the switch set up so when the head (does not matter just pick one) goes into the -Z range the switch is tripped and stays in that state the entire time the head is in -Z range. And when the head is +Z it is open and stays open the entire time the head is in +Z range. 

This way when openPNP powers up it can just look at the state of that switch and know instantly which way it needs to move to get back to zero. 

If you seem my post here:

https://groups.google.com/d/msg/openpnp/GdQ2Wvt0qGY/XNkjA_o9DQAJ

The Z flag on my head runs half the entire range of the Z travel. In my case when the right head moves up the slot is unblocked turning on the output. When the head moves down the slot is blocked at 50% of the total travel (Z = 0) and the output turns off. If I keep moving down the output stays off (because the slot is blocked) until I change directions and move past the 50% travel (Z = 0) point. 

Hope that helps. 

[Peter Betz]

On Sunday, October 9, 2016 at 2:02:54 PM UTC-7, Mikael Bohman wrote:

Hi Peter

I do not understand the meaning of "having the switch tripped continuously while it is on one side of center on one nozzle."

Can you explain it in other words?

My idea was to park the head in the holder for the different nozzles before power off. That way the machine will know it's xyz position within 0.5 mm at power on, and can make a safe homing.

Hi Mikael,

 If your machine is a one-of design (just for you) then that is probably fine, personally I would rather not have it park and cold start 0.5mm over a tool holder. I'm used to seeing steppers jump a little when you enable them. I assume that is because when you enable it, it isn't energizing the same last set of windings at the same micro stepping it was using when you parked and shut off the machine. So it jumps slightly to the windings that are energized first. Homing allows you to be precisely certain of the machines position.

Hope this helps!
Peter.

[Mikael Bohman]

Thanks for your answers. I understand what you mean with the limit switch now.

I will try and add the limit switch to my design.

Mikael

[Mikael Bohman]

OK, next revision here.

I have changed to and bought a 15mm wide HTD-3 belt.

The stepper can be moved +-1.5mm to stretch the belt.

I have bought the TCRT5000 IR diode + phototransistor to detect the mid position of the nozzle.

  v

[Jason von Nieda]

Beautiful renders Mikael, really nice work. Head is looking good!

--
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/02da1346-8de9-455a-a245-0a691d8789cd%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Anthony Webb]

Looks amazing!

Sent from my iPhone

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CA%2BQw0jwyPXWSNmTGEuJmYrDhkaWn5OXcPWu43yAcw0zgJCo5FQ%40mail.gmail.com.

[Frank Herrmann]

Mikael, can u short explain from which links/sites you get the 3d modells for i.e. the linear slider? 

You use Fusion360?

[Mark Harris]

Hiwin have the models on their website.

Looks like solidworks rendering to me :-)

--

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/e0983da4-12db-4e15-8e93-f1d32e4964cd%40googlegroups.com.

[Mikael Bohman]

I'm in Europe, so I use https://www.dold-mechatronik.de/Profilschienenfuehrungen (google translate might help you translate).

They have the step-files.

It's Fusion 360

[Mikael Bohman]

This is my idea of the Y-axis close to the stepper.

PS. I have no background in MCAD or design. DS.

[Reiner Schmidt]

I think someone mentioned it at the beginning of the thread but i don't think a pneumatic cylinder is ideal for the drag pin. That being said it will work but there are some draw backs.

• Large energy consumption (you have to fill the cylinder with air twice for every part placed)
• more force = more damage if there is misalignment or bad setup
• If limits are not used to check if the cylinder is up or down you could have your machine move with the cylinder down (assuming air failure) that could be pretty damaging.
  

On Monday, November 14, 2016 at 11:23:08 AM UTC-5, Mark Harris wrote:

Hiwin have the models on their website.

Looks like solidworks rendering to me :-)

On Nov 14, 2016 8:17 AM, "Frank Herrmann" <xpi...@gmail.com> wrote:

Mikael, can u short explain from which links/sites you get the 3d modells for i.e. the linear slider? 

You use Fusion360?

Am Sonntag, 13. November 2016 01:44:01 UTC+1 schrieb Mikael Bohman:

OK, next revision here.

I have changed to and bought a 15mm wide HTD-3 belt.

The stepper can be moved +-1.5mm to stretch the belt.

I have bought the TCRT5000 IR diode + phototransistor to detect the mid position of the nozzle.

  v

--
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.

[Trampas Stern]

A electrical solenoid might be a good solution, provided it is attached to one of the PNP heads. Specifically the head would go down, then solenoid enabled to extend the pin. Then to retract the pin the solenoid would be disabled, but since spring on solenoid might not be strong enough the head would retract to make sure pin returns. 

[Frank Herrmann]

Cool, i'm come from germany. Mikael, can you share your Fusion360 Project over the Autodesk portal? 

I'll just learn how can i build a big projects and how do this other peoples :)

[Matt Brocklehurst]

Impressive renderings :)

-- 
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/97c2fe94-7c85-4932-b3d7-6dc32d33f4c9%40googlegroups.com.

[bobgee...@gmail.com]

Mikael,

Looks very good, very very good!

I am trying to build a machine based on Anthony's design and was looking at Peter's head, but now I am not sure which way to go because things have gone a little quiet on his latest version and you have made a lot of progress in the meantime. I am looking for something which can be built from 3d printed parts and laser cut acrylic becuase I have to keep my costs down. Your design looks very promising. May I ask:

Will you release the designs when you are happy with it?
Can all the plates be cut from a laser cutter or are do you need to tap any holes etc?
When do you expect to have the design mature enough to build one?
Are they Nema 8 motors?
Are the rails Hywin MGN9? What length?
Does the head mount directly to a MGN12H block?
When you mentioned a 15mm wide belt, are you talking about the belt on the head itself or the belt that moves the head? What pulleys are you using?

Sorry for so many specific questions, I am about to hit checkout on a robot digg shopping cart which assumes Peter's head, and I just want to see if most of the parts in my basket are transferable, or whether I remove the head items and sit tight.

Many thanks and good luck!

Bobby

[Mikael Bohman]

Thanks, I think that the drag pin will be in phase 2 of my project.

I do not know what solution to finally use yet foor feeders

. I will probably test different ways. At least I have made treaded holes to connect something to the main plate futher on.

[Mikael Bohman]

Try out http://a360.co/2fDuDIz 

I have some unknown problem with my joints.

It missbehavies when open in the browser, but becomes correct if you open it in fusion an moves the head a little.

[Mikael Bohman]

Den tisdag 15 november 2016 kl. 22:42:14 UTC+1 skrev bobgee...@gmail.com:

Mikael,

Looks very good, very very good!

I am trying to build a machine based on Anthony's design and was looking at Peter's head, but now I am not sure which way to go because things have gone a little quiet on his latest version and you have made a lot of progress in the meantime. I am looking for something which can be built from 3d printed parts and laser cut acrylic becuase I have to keep my costs down. Your design looks very promising. May I ask:

Will you release the designs when you are happy with it?

Yes, you can check it out now, but it is work in progress.

http://a360.co/2fDuDIz , as written above. Some unknown problem with the joints.

 

Can all the plates be cut from a laser cutter or are do you need to tap any holes etc?

Yes, you need to tap holes, but that could be done by hand or

 

When do you expect to have the design mature enough to build one?

I have ordered most of the ebay part. Will soon order the main parts from rhttp://www.robotdigg.com/ and https://www.dold-mechatronik.de/

 

Are they Nema 8 motors?

YES

 

Are the rails Hywin MGN9? What length?

NO, http://www.ebay.com/itm/201550369288?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

 

Does the head mount directly to a MGN12H block?

NO, a ARC15FN block https://www.dold-mechatronik.de/Linearwagen-ARC-15-FN-Flanschmodell. I do not trust Chinese Hiwin copies. I hope whis one is better.

When you mentioned a 15mm wide belt, are you talking about the belt on the head itself or the belt that moves the head? What pulleys are you using?

This moves the head,15 mm belt , Idler pulley, Timing belt bulley. GT2 6mm for the nozzles.

[Mikael Bohman]

The latest status.

Can openCV understand the design solution with pins in the tape-holes ? I ordered a bag of 1.5 mm steel pins from ebay. 

[Matt Brocklehurst]

Some other machine also uses this idea of pins in all the holes to hold the strip tape down... i like the simplicity of it, but haven't tried to see if the tape feeder as it stands will work with this (not sure if it looks for black holes or if its just looking for circles... in which case the shiny pins will probably appear as a circle and it will magically just work). Does having the pin in every hole make it easier to deal with loose bits of tape... i.e. better than sticking it down with double sided tape and carefully trying to peel back the cover?

--

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/3bf9b802-32f6-477a-aad2-0c0da3e88b0c%40googlegroups.com.

[Mikael Bohman]

Can all the plates be cut from a laser cutter or are do you need to tap any holes etc?

 I have holes with a chamfer.

[bobgee...@gmail.com]

Mikael,

Thanks for the info. Chamfered holes might scupper my plans. It also looks like Peter's revisions will help me get to where I want to be a bit quicker than trying to bodge your design. I am hoping to get a basic substitute for the robot digg head, fast and cheap. Later on, I will consider upgrading to a better head with more features such as yours.

Thanks again,
Bobby