Head design hints/tips?

285 views
Skip to first unread message

Harjit Singh

unread,
Oct 29, 2021, 1:02:52 AM10/29/21
to ope...@googlegroups.com
In another thread, Mark said:

For those building their own heads: make sure your head offset Y (the Y distance from the nozzle tip to the camera view center) is a less than 30mm.

Any other tips?

I'm planning on a head with a single nozzle with the down looking camera next to it (x direction).

Thanks.

ma...@makr.zone

unread,
Oct 29, 2021, 6:08:23 AM10/29/21
to ope...@googlegroups.com
What I would do:
  1. make the head light, i.e. fast
  2. balance head weight above/below and before/behind the rails to reduce shaking and stress.
  3. Z clearance should be generous but still not sacrifice (1) and (2).
  4. camera offset largely in X, faster
  5. camera close to nozzle, but camera lights should still not glare into each other, when the nozzle is on top the bottom camera.
  6. You should be able to mount a shade behind the nozzle to fill the bottom camera view.
  7. take a very long lens and move the top camera up.
  8. Unless you have co-axial lighting: Mount the top camera diffuser near the camera lens front, to make the see-through hole in it small. Make it large enough to fill the camera view when reflected on PCB level.
  9. Put the valve and a sensor on the head.
  10. (If this is still open): put compact Z/C motor drivers (or even the controller) on the head to reduce heavy cabling and drag chain.
  11. Use small motors with high voltages.

_Mark

--
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 view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CAEm8oETChFs%3DoGoah7boSY7m%3Dt2rp-M%3DnfXPc7%2Bqv5d7X%3DO4hQ%40mail.gmail.com.

Harjit Singh

unread,
Nov 6, 2021, 2:34:06 PM11/6/21
to ope...@googlegroups.com
Thank you.

Assume
* Precision is more important than speed. Want to place 0.4mm WLCSP and QFN with 0.4mm pitch.
* Single nozzle machine
* Z-axis zero position is when the tip is at the highest point. I believe OpenPnP operates Z-axis in the negative direction when picking up and placing parts.
* Changer is at a higher plane than the PCB and feeder. This is just an assumption
* PCB and feeder plane are 20mm above the table <- no particular reason for this. I can "sink" the table to make this less.

What would you recommend for the:
a) Z-axis full throw - I was planning 60mm
b) Clearance between PCB and feeder plane and Z-axis position when moving around the table - I was planning 25mm. My components will be approx. 5mm max.So,25mm = 15mm gap with a loaded part.

I've read notes about being careful about crashing the nozzle/head. What should I change to minimize this? For example, sink the changer but would the target be the "engage" Z location is the same as the PCB and feeder plane's? or?

Anything else I should consider?


Mike Menci

unread,
Nov 6, 2021, 2:53:39 PM11/6/21
to OpenPnP

Harjit Singh

unread,
Nov 6, 2021, 3:59:50 PM11/6/21
to ope...@googlegroups.com
Thank you for the links. I'd found and studied all but the last link - looking at it now.

Since everyone makes trade-offs, I was looking for insights/rationale/considerations.

All the best.

ma...@makr.zone

unread,
Nov 7, 2021, 3:17:23 AM11/7/21
to ope...@googlegroups.com

Hi Harjit,

Ideally, everything is approximately at the same PCB surface Z height (say ±1mm), including the bottom camera focal plane, all feeders (specifically the part surfaces) and the sunken nozzle tip changer (surface). This will make the machine more precise, because you don't have any errors brought in by nozzle Z axes not being perfectly perpendicular. The same is true for the cameras' viewing axes (down and up Z) not being perfectly perpendicular.

Plus you have no problems using the camera, as everything is in or near the focus plane (not blurred) and at the same scale (Units per Pixel). 

Note, we have various compensations for such Z related errors built into OpenPnP (and Tony Luken will shortly introduce even more of the kind):

https://github.com/openpnp/openpnp/wiki/Nozzle-Tip-Calibration-Setup#what-is-the-bottom-camera-location-and-how-can-it-vary-between-nozzles

https://github.com/openpnp/openpnp/wiki/3D-Units-per-Pixel

https://github.com/openpnp/openpnp/pull/1297

So it is not a hard requirement to have everything on the same Z level, but still just better design.

Given you have everything more or less on the same Z, the answers to your questions are as follows:

b) Twice the largest part you ever want to place plus some extra, so you can lift one over the other.

a) Take b) plus the "diving" needed for the sunken nozzle tip changer.

I personally started using a sunken changer after I managed to crash my machine one too many times at high speed, so I really recommend it 😁:

https://makr.zone/pick-place-machine-first-simulated-small-test-run/66/

It also (probably) works better for the vision calibration, as the template images are more or less in focus:

https://github.com/openpnp/openpnp/wiki/Nozzle-Tip-Changer#vision-calibration

_Mark

Mike Menci

unread,
Nov 7, 2021, 3:39:48 AM11/7/21
to OpenPnP
Well Harjit, you will need to say which nozzles would you want to use ? 
The nozzle holder and nozzle size with Nema 8 will determinate your Z hight.  
Samsung adapter and nozzle dimensions.png
Nema8-hollow-shaft-stepper-holder-FOR-JUKI-NOZZLE.png

Harjit Singh

unread,
Nov 8, 2021, 1:26:42 AM11/8/21
to ope...@googlegroups.com
I'm using Peter's Samsung CP40 nozzle.

Based on feedback and investigations, here is a picture of what I'm thinking. I still need to add in the down looking camera, the back plate which this mounts on and then attaches to the X-axis.

image.png
image.png


Reply all
Reply to author
Forward
0 new messages