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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Misalignment Issue (CAD Vs. Machine Variables

George

#2483

Misalignment Issue (CAD Vs. Machine Variables | 11 November, 2000

I am tasked to breakdown the causes and to report the best reasons why our everytime we download our CAD Data we need to align or adjust our component alignment. Can anyone help me to identify as I do not have enough experience with SMT equipment. Just share your bits of knowledge. Our machines are using a Centering Chucks and not a full vision system. Our Vision system is using Tresh Hold Level our Gray scale camera.

Thanks and regards,

reply »


JAX

#2484

Re: Misalignment Issue (CAD Vs. Machine Variables | 11 November, 2000

Have you tried running a glass plate?

If you still need to adjust placement it is a machine issue.( calibration )

If everything is good your conversion software needs looked at.

That's all I have without knowing the CAD software and Machines you use.

reply »

George

#2485

Re: Misalignment Issue (CAD Vs. Machine Variables | 11 November, 2000

Hi Jax, CAD Software is Protel and the machine is Yamaha YM84 Hyper Series.

thanks

reply »


JAX

#2486

Re: Misalignment Issue (CAD Vs. Machine Variables | 12 November, 2000

Does the alignment issue only pop up on certain parts or everything on the board? Do you set your origin on the corner of the board or on a fiducial? Do you always have to move the parts the same amount of distance?

reply »

#2487

Re: Misalignment Issue (CAD Vs. Machine Variables | 12 November, 2000

George, You need to calibrate your chucking assemblies! . What you are seeing is poor alignment due to egocentricity in you jawing assemblies. Also make sure that your jaws/nozzles are in good condition and that your assemblies are correctly orientated with the correct jaw on the correct arm, this is crucial as it determines the centering sequence. Also check your centering height and pickup/placement height. Finally check your head offsets. Irregularities can also occur due to faulty vision files and non cad fiducial data - NEVER TEACH FIDUCIALS! Contact your Phillips/Yamaha dealer and get yourself a Yamaha calibration pcb such as 'mntchip' or 'mntqfp'and the relevant programs to run these boards. Check your calibration at least every month. The old 4600/6000/21 and hyper series machines are deadly accurate from cad data with few exceptions, look after it and it will look after you. And no, ther's nothing wrong with Protel either. Happy calib. Darby

reply »

George

#2488

Re: Misalignment Issue (CAD Vs. Machine Variables | 14 November, 2000

Hi Jax, The alignment pop up only on some parts and it's random axes. This happens every prototype run that I need to adjust the alignment of some parts. Fiducials are correct and okay. Also had done calibration on our machines. I really suspect that it is a CAD problem but I've been told that there should be no alignment after generating a program from the CAD. Temporary counter measure that I proposed is to align all parts at 100% alignment using double sided tape prior to release a placement program for production. But people here do not like my proposal. How does a glass plate works ?

reply »

sarry

#2489

Re: Misalignment Issue (CAD Vs. Machine Variables | 14 November, 2000

I have one suggestion why don't you align all parts and take a close look at it if you can find any misaligned component after that. IF SO, that's machine's repeatibility problem.(or mechenical problem-dirty nozzle, wrong pick up and placement height and etc) IF not, cad file is not correct. How glass plate works? you need some sop or qfp components with glass plate after placement, you can look at placement accuracy with microscope.

reply »

#2490

Re: Misalignment Issue (CAD Vs. Machine Variables | 14 November, 2000

Hi George,

you got to find the common factor for the parts that are misaligned. This double sided tape method is IMO a good one for the first shot but you have some waste in parts that should be considered. (Better skip those precious parts for the test run and look for them later when you can correct the placement or use some dummy or test devices for your "sticky test"). We had similar problems but they turned out to be fiducial related problems (HAL :-( ). Sometimes it went from good to bad during normal run, that turned out to be a loss of offsetdata for the heads in the RAM. Nothing helped but a shutdown and restart but that doesn�t seem to fit to your problems.

Good luck Wolfgang

reply »

#2491

Re: Misalignment Issue (CAD Vs. Machine Variables | 14 November, 2000

Without using a glass plate you could check if your jaws could properly align the compomnents by writing a small vice versa program. Place a component in a 0 degree angle and another one right next to it in a 180 degree angle. Built up a line of lets say ten components and check after placement if they are all lined up. I find in CAD programs that the angle for none polarized components are randomly selected 0 or 180 and 90 or 270. Again, if the jaws do not correctly align to the center your vice versa program will show. If the misalignment is off by 0.1 mm the components are shifted by 0.2 mm from one to the next.

Does your CAD program allow a 1:1 print out? Could you make a print out on a mylar and overlay it to your board layout? Stefan

reply »

#2492

Re: Misalignment Issue (CAD Vs. Machine Variables | 14 November, 2000

Hi George! Is your CAD in inches and your equipment metric by any chance? If so, make sure your CAD data is converted to metric using at least 4 digit after the decimal (0.1234" for example)

reply »

Francois

#2493

Re: Misalignment Issue (CAD Vs. Machine Variables | 14 November, 2000

First groupe of questions: Is your machine have one head with multiple nozzles? Do you have separate vacuum tubes for each nozzle? Do you have problem with the same package or with all pakages. In your PNP machine, what are the dimensions tolerances for your parts?

Software question: With protel, do you have to place an offset for each component to determine where is the center 'fiducial' of parts. Do you place a reference target in the center of your parts. How do you determine the center of your component when you create your parts?

reply »

#2494

Re: Misalignment Issue (CAD Vs. Machine Variables | 14 November, 2000

George, Send me your email address, I'll send you some files and programs that will make things clearer for you, and easier to convert your cad data straight from Protel to Yamaha Hyper series format. Darby.

reply »

George

#2495

Re: Misalignment Issue (CAD Vs. Machine Variables --Done!!! | 21 November, 2000

Hi Guys, We found the cause of the problem. It is during the import of the CAD data which on some components the Software give us the Mid X and Mid Y values. We checked from the protel and we found that some of our component coordinates do not much the imported XY coordinates. We still need to identify which between the CAD softwares (Protel and SMS Line Control) is giving us the problem.

Thank you for all your good and fruitful suggestions. George

reply »

Steve

#5835

Re: Misalignment Issue (CAD Vs. Machine Variables --Done!!! | 2 April, 2001

George,

Did you ever get to the bottom of your XY offset issue. I am a user of Protel and SMS Linecontrol, and know the system well. You probably need to find out if the Protel XY is the component center or Pin 1. The SMS Linecontrol Protel module will import the exact data you have supplied.

reply »

Manufacturing Software

PCB equipment