Printed Circuit Board Assembly & PCB Design SMT Electronics Assembly Manufacturing Forum

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


SMT Connector, Placement Issues

Views: 4812

I have a problem with placement repeatability and accuracy o... - Oct 03, 2008 by Paul Richardson  

#56806

SMT Connector, Placement Issues | 3 October, 2008

I have a problem with placement repeatability and accuracy on this Samtec connector. The customer wants to use a "nubbed connector" which has to be placed in the corresponding PCB holes. When the nubs miss the holes, the part becomes "lifted." SEE BELOW PICTURE FOR DETAILS.

One theory is....In PCB Fab processes, there's going to be tolerance stackup with respect to the copper and the location of the holes after drilling (things will drift). That's why people run into these problems when trying to combine SMT and through-hole since your placement equipment doesn't care about where the through-holes are - it just looks for component centroids.

Anyway, does anybody else have this issue or seen his issue in the past? If so, how'd you fix it? We're thinking a rubber-tip nozzle to eliminate the part "drifting" on the nozzle before placement. We've already tried other stuff, like slowing the mounting speeds down, etc. with marginal improvements.

Attachments:

reply »

#56807

SMT Connector, Placement Issues | 3 October, 2008

Change your connector.

reply »

#56808

SMT Connector, Placement Issues | 3 October, 2008

Change your board to bigger holes.

reply »

#56810

SMT Connector, Placement Issues | 3 October, 2008

On a couple of boards, try remove the pins, let reflow align the connector to the pads

reply »

#56812

SMT Connector, Placement Issues | 3 October, 2008

We agree with Huss. Your clearance 8 to 11 thou between lead and hole make ion difficult. And that assumes the 0.027 to 0.030" hole size is the finished, not drilled size. It just gets worse if this hole size doesn't consider plating.

reply »

#56813

SMT Connector, Placement Issues | 3 October, 2008

What type of placement machine do you have? How is the machines vision system viewing the part, body x,y only? Do you have the option with the vision system to use just the lead information? Can you use the pin itself for placement from an up looking camera? Is there any rotation after the camera before placement, can you pre-rotate the part before vision then place the part. Try using a rubber tip nozzle, slow the travel rate down

reply »

#56823

SMT Connector, Placement Issues | 6 October, 2008

DaveF,

This is a non-plated through hole which these nubs go into. Couple of questions for you:

1. Our rated placement accuracy is 0.05mm -> 0.001969"

You say that 0.008" of "slop" is un-doable; hence, the reasons for our failures. Any basis for that statement? Are rated accuracies from placement equipment manufacturers typically "exgaggerated?"

2. On PCB Fab side of things, what is the likelihood of those holes being in the same place, and repeatably? These hole's center-to-center is measured at 1.27" (30.4mm). What would cause the hole's to drift out of place, or is it more likely that the part placement is "drifting?"

reply »


SME

#56825

SMT Connector, Placement Issues | 6 October, 2008

Hi, can you use the holes in the PCB for this part as extra fiducials but only reference those fiducials to that 1 placement(main and sub)

reply »

#56826

SMT Connector, Placement Issues | 6 October, 2008

Brick (if that is your real name),

Placement machines use centroid data to place parts. Generally this is derived from the "copper" that is etched on the board. Thus why we use the fiducials for alaignment.

During board manufacturing, holes are drilled in a seperate process than the copper, thus will have a different drift and tollerance than the copper.

Now with your part you mix the two. So you have a machine following the copper drift form board to board, while the holes move around on their own. It's simply a stack up of tollerances that are causing your problem.

reply »

#56829

SMT Connector, Placement Issues | 6 October, 2008

Q1: You say that 0.008" of "slop" is un-doable; hence, the reasons for our failures. Any basis for that statement? A1: We're not even considering the accuracy of your placement machine. We base our comments on the gap between the hole and the connector required for successful through hole placement.

Q2: Are rated accuracies from placement equipment manufacturers typically "exgaggerated?" A2: We believe most placement machine fabricators are conservative in their specification of accuracy.

Q3: On PCB Fab side of things, what is the likelihood of those holes being in the same place, and repeatably? A3: We expect holes to be located with the fabricator's tolerance [maybe+-3 thou]. Also, we expect diameter of the holes to be located with the fabricator's tolerance [maybe+-3 thou but could be as much as 10].

Q4: What would cause the hole's to drift out of place?" A4: Drilling holes through stacks circuit boards is tricky business.

Q5: Is it more likely that the part placement is "drifting?" A5: If the placement of the connector is drifting, then we'd expect broader issues with placement accuracy than just the connector. We are not racing to throw your placemnt machine overboard, before understanding tolerance stackup between the board holes and the connector tabs.

So, what's story on those tab nubs? Is the size of nubs controlled? Are burrs cleaned off the nubs? If there nubs at both ends, how well is the spacing controlled? Are the diameter and the spacing holes on the board nubriffic, according to the connector fabricator's drawing? In your picture, the portion of the nub closest to the connector frame looks more narrow than the portion further away. If that's true, why?

reply »

#56831

SMT Connector, Placement Issues | 7 October, 2008

We had the same problem last year. You need verify these key points. 1- Holes are drill, NOT PUNCH. 2- Inspect 100 Boards at microscope, and verify that there is not rest of fibers in the holes. 3- Meassure the dimensions of the holes. a) virgins boards b) after 1st reflow c) after 2nd reflow With temperature, the holes reduce the dimensions and the connector is not correct seated. 4- Meassure the connector pin. 5- Verify both tolerance, marginal situation probably are not compatible ( pin and Hole )

At the end, our solution was change the design, increase the hole dimension and work with the supplier to get a complete clean hole ( fiber). SMT connector has a very sensitive placement.

reply »

#56832

SMT Connector, Placement Issues | 7 October, 2008

Hi - maybe FINETECH (http://www.finetech.de ) can help you. They did evaluate the reworkability of SAMTEC connectors. The FINEPLACER Pico offers 5 �m placement accuracy (http://www.finetech.de/enid/picors ) - the precise placement should not be much of a problem. Attached is a short notice about the evaluation. If you need more help just send me an email.

Attachments:

reply »

#56837

SMT Connector, Placement Issues | 7 October, 2008

Thank you rafa.

This is the type of "been-there-done-that" reply that I was looking for, versus some of the other empty "theoretical" rhetoric.

reply »

#56842

SMT Connector, Placement Issues | 7 October, 2008

Just curious...which components of the previous replies did you consider to be "theoretical rhetoric" as opposed to a sound scientific approach to a problem?

reply »

#56848

SMT Connector, Placement Issues | 8 October, 2008

Short answer is yep, if you are centering on the nubs and not the leads. Depending on the registartion between nubs and leads you may then have misplacement. Holes as fids will usually let you down at some stage due to vision issues. Blunt drills blah blah blah. Watch like hawk. I hate nubs unless you are using mechanical centred force placement. Oh, and place them first. Darby - yeah - I'm still alive.

reply »

#56851

SMT Connector, Placement Issues | 8 October, 2008

All posts were well-intended, and I understand the, "have you tried this", or "have you tried that" approach.

Rafa's was the only person who actually "DID IT." That's what i was looking for.

In the heat of battle when you're getting chewed out for yields and defects, these types of answers are the most helpful.

reply »

#56898

SMT Connector, Placement Issues | 10 October, 2008

We have several similar parts in our factory and they usually don't give me any problem. One part has a 0.050 pin that goes into a 0.055 hole, and we place it with Panasonic MSR turret style chip shooter, which has a placement accuracy of only +/- 0.003 in. The numbers don't really work, considering the accuracy of the machine and then adding in some tolerance for pcb drill hole to copper alignment. 0.008 that you have should be plenty of extra room. Does the pick and place machine recognize the part accurately?

reply »

ii-feed SMD Intelligent Feeder

PCB Routers