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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


MPM AP/A Problem

Views: 3221

#51712

MPM AP/A Problem | 10 September, 2007

Having a problem with a MPM AP/A. Once a board is programmed; all is calibrated and the first board screened - each board thereafter drifts and gets progressively worse. The drift can range from barely off to half off the pad. Any ideas?

reply »

#51713

MPM AP/A Problem | 10 September, 2007

Set "FP" mode to 1 in the setup menu 1. It won't fix your problem, but it might help a little. What kind of board support system are you using?

reply »

#51715

MPM AP/A Problem | 10 September, 2007

Chris,

Appreciate the response - we set FP to "1" last week but no help. The drift is only in the X as well - the Y is never off. We have tried switching the cards to see if if the problem goes to the Y but it doesn't - stays in the X.

As for the board support system - we are using dedicated tooling for this board. There are no support pins for this fixture.

The odd thing is that, for example, the first board or two will be good, the next 3 will be horribly off, then the next board will be good again. We had to put in an offset of 160 thousandths in the X just to get it lined up to get the first good board.

reply »

#51719

MPM AP/A Problem | 10 September, 2007

I assume your vacuum suction is good. Is this an above mounted board stop sensor?

-Make sure it is finding the correct fiducial and not a similar object nearby. -Try physically moving the X-axis to check for backlash. There should be no movement. -Also make sure the x axis motor coupling is tight on both ends. -Make sure the stencil is securely clamped. -Try slowing down the conveyor speed(not a true fix, but it could help)

reply »

#51734

MPM AP/A Problem | 11 September, 2007

Are you using fiducials or pads?

reply »

#51888

MPM AP/A Problem | 24 September, 2007

I never saw in any post if your machine uses the camera to find the board edge or if your machine uses the bottom front rail mounted board stop sensor.

Also is this with only one particular job you run on this machine or is this happening with ALL your jobs? If it were all of the jobs then I would check the vacuum on the board itself when moving to the stencil. If it is strictly one job in particular then try the following......

If it is the bottom style sensor, look on your board where the sensor engages the boards edge. Are there any reflective areas, on the bottom side or large cutouts near where the sensor would engage the board? If so, try using a piece of Kapton tape on the bottom edge where the board stop sensor meets the board.

If it is a camera mounted sensor, the only way I am aware of changing the position where the camera senses the board is by going into the "Teach Board" menu and reteaching that particular step so that you get the sensor to locate on a different leading edge area of your PCB.

Chris

reply »


kpm

#51945

MPM AP/A Problem | 1 October, 2007

I've never worked with a MPM printer but I had something similar happen on a DEK printer. The problem was that screens were manually centered in the machine and the operators were not careful enough when setting it up so the settings were very close to the limits of the machine. For some reason this made the machine behave in exactly the manner you've described. All we had to do to fix it was recenter the screen in the machine so that it was father away from the limits and it worked fine. I hope my explaination is clear enough to help. I can't think of how else to describe it. Part of the problem was we also had to load the screen into a fixture so sometimes it would be crooked and misaligned because of that.

reply »

#52000

MPM AP/A Problem | 8 October, 2007

Worked for MPM for many years, and this sounds like a fairly common problem which is easy to fix if you can answer a few questions, several of them already asked... 1. Is this happening only on this board (if the answer is yes, this is a good thing! no need to troubleshoot the mechanics of the unit.) 2. Are you using fids or pads for alignment?

If you are using pads, try the following "teach vision" method;

Enter teach vision and delete all models. Find two devices on opposite corners of the board which will allow you to see at least two of the pads on the monitor screen. (a 20 or 25 mil pitch QFP is VERY good to use!) Center the end pad of the device in the cross hairs and hit next. This brings up a graduated box that you can manipulate both positionally and in size. Make the box large enough to encompass both of the end pads on the device (with a little of the board also) then "offset" the box (reposition it) so that only the end pad is within the box. Repeat for all models being taught. This works because this first box in the teach vision software routine is defining a "search area" within the field of view. When you offset the box and hit next, a bit map of the search area is stored on the HDD.If in production the defined search area contains two pads (the camera moving fast, board positional error or whatever)the vision system sees both of the pads, rejects the second one, and automatically stepps off to the end pad which is a valid target within the defined search area. Works every time if you are using pads/apertures for alignment.

reply »

PCB machines

Reflow Ovens thermal process improvement