how to overcome tolerance in automatic alignment

hi, i m college student. i m currently doing my Final year project in PCB drilling. the main task is to design auto feeder which allign the PCB properly in position before drilling take place. the close tolerance becomes a big problem to the precision. anyone who have any idea please kindly share with me

reply »

| hi, i m college student. i m currently doing my Final year project in PCB drilling. the main task is to design auto feeder which allign the PCB properly in position before drilling take place. the close tolerance becomes a big problem to the precision. anyone who have any idea please kindly share with me | Calvin: Look at the conveyors used in placement machines. Good luck. Dave F

reply »

| hi, i m college student. i m currently doing my Final year project in PCB drilling. the main task is to design auto feeder which allign the PCB properly in position before drilling take place. the close tolerance becomes a big problem to the precision. anyone who have any idea please kindly share with me |

Hi Calvin.

You are up against an age-old problem. What you fight is actually tolerance stack-ups. Artwork registration versus panel edge routing tolerances, with the additional tolerance stackups of multiple panels stacked for drilling, coupled with drill runouts, insertion and placement machine tolerances, and Mr. Murphy. The truest way to ensure accuracy may be to utilize fiducials on the panel, and register the drill pattern to the artwork based upon the fiducials. But the dilemma becomes one of speed and efficiency. It is not practical for a board house to drill PCB's one at a time. So the mfg'ers drill panels stacked. This causes registration variances as the panels are drilled. Dull drills compound the runout issue, and FR4 is not exactly kind to metal cutting tools. The automatic equipment for insertion into the drilled and plated holes and placement onto pads has had optical correction for artwork registration problems installed for many years. Most folks utilize this method to overcome the tolerance stackups. Many automatic assembly operations simply use edge registration, and look for the fiducials.

I don't know of a better system out there which is economically feasible.

Scott Cook

reply »

| | hi, i m college student. i m currently doing my Final year project in PCB drilling. the main task is to design auto feeder which allign the PCB properly in position before drilling take place. the close tolerance becomes a big problem to the precision. anyone who have any idea please kindly share with me | | | | Hi Calvin. | | You are up against an age-old problem. What you fight is actually tolerance stack-ups. Artwork registration versus panel edge routing tolerances, with the additional tolerance stackups of multiple panels stacked for drilling, coupled with drill runouts, insertion and placement machine tolerances, and Mr. Murphy. | The truest way to ensure accuracy may be to utilize fiducials on the panel, and register the drill pattern to the artwork based upon the fiducials. But the dilemma becomes one of speed and efficiency. It is not practical for a board house to drill PCB's one at a time. So the mfg'ers drill panels stacked. This causes registration variances as the panels are drilled. Dull drills compound the runout issue, and FR4 is not exactly kind to metal cutting tools. | The automatic equipment for insertion into the drilled and plated holes and placement onto pads has had optical correction for artwork registration problems installed for many years. Most folks utilize this method to overcome the tolerance stackups. Many automatic assembly operations simply use edge registration, and look for the fiducials. | | I don't know of a better system out there which is economically feasible. | | Scott Cook | | Our board shop drills boards stacked. Scott is right and to solve the problem we invested in a tooling system that allows us to align the artwork to fiducials then the tooling holes are puched. The tooling pins are then pressed through the stacks for tight alignment and works very well with the top vs. the bottom alignment. That still doesn't stop the variances with the drilling process so we periodically run test samples and use x-ray to monitor our stacking process.

If you can, visit/contact Excellon. They have an automated conveyor system and may be able to help you out.

Cheers, Scott

reply »

| | | hi, i m college student. i m currently doing my Final year project in PCB drilling. the main task is to design auto feeder which allign the PCB properly in position before drilling take place. the close tolerance becomes a big problem to the precision. anyone who have any idea please kindly share with me | | | | | | | Hi Calvin. | | | | You are up against an age-old problem. What you fight is actually tolerance stack-ups. Artwork registration versus panel edge routing tolerances, with the additional tolerance stackups of multiple panels stacked for drilling, coupled with drill runouts, insertion and placement machine tolerances, and Mr. Murphy. | | The truest way to ensure accuracy may be to utilize fiducials on the panel, and register the drill pattern to the artwork based upon the fiducials. But the dilemma becomes one of speed and efficiency. It is not practical for a board house to drill PCB's one at a time. So the mfg'ers drill panels stacked. This causes registration variances as the panels are drilled. Dull drills compound the runout issue, and FR4 is not exactly kind to metal cutting tools. | | The automatic equipment for insertion into the drilled and plated holes and placement onto pads has had optical correction for artwork registration problems installed for many years. Most folks utilize this method to overcome the tolerance stackups. Many automatic assembly operations simply use edge registration, and look for the fiducials. | | | | I don't know of a better system out there which is economically feasible. | | | | Scott Cook | | | | | Our board shop drills boards stacked. Scott is right and to solve the problem we invested in a tooling system that allows us to align the artwork to fiducials then the tooling holes are puched. The tooling pins are then pressed through the stacks for tight alignment and works very well with the top vs. the bottom alignment. That still doesn't stop the variances with the drilling process so we periodically run test samples and use x-ray to monitor our stacking process. | | If you can, visit/contact Excellon. They have an automated conveyor system and may be able to help you out. | | Cheers, | Scott | All things drill or place start with CAD data. In a board shop it is coverted using CAM and pertinent information (drill entry points)is downloaded to automated drilling equipment. This data would be an ideal place for you to start as it is simple X/Y type as Gerber (dumb data).

Everyone has stated problems correctly concerning drill wander, spindle runout, stacking, artwork, etc. Still, you have to start at the start. Data is that point as it all is referenced from a single datum from which all other locations, features, fids, etc. are positioned.

As far as making an accurate table, fixture, or other locating mechanism goes, it's all up for whatever works best. Original production artwork, when first generation, offers little problem with respect to registration. It's the iterations it goes through as image, etch, etc.

Still, the data is the thing.

Earl Moon

reply »

| | | | hi, i m college student. i m currently doing my Final year project in PCB drilling. the main task is to design auto feeder which allign the PCB properly in position before drilling take place. the close tolerance becomes a big problem to the precision. anyone who have any idea please kindly share with me | | | | | | | | | | Hi Calvin. | | | | | | You are up against an age-old problem. What you fight is actually tolerance stack-ups. Artwork registration versus panel edge routing tolerances, with the additional tolerance stackups of multiple panels stacked for drilling, coupled with drill runouts, insertion and placement machine tolerances, and Mr. Murphy. | | | The truest way to ensure accuracy may be to utilize fiducials on the panel, and register the drill pattern to the artwork based upon the fiducials. But the dilemma becomes one of speed and efficiency. It is not practical for a board house to drill PCB's one at a time. So the mfg'ers drill panels stacked. This causes registration variances as the panels are drilled. Dull drills compound the runout issue, and FR4 is not exactly kind to metal cutting tools. | | | The automatic equipment for insertion into the drilled and plated holes and placement onto pads has had optical correction for artwork registration problems installed for many years. Most folks utilize this method to overcome the tolerance stackups. Many automatic assembly operations simply use edge registration, and look for the fiducials. | | | | | | I don't know of a better system out there which is economically feasible. | | | | | | Scott Cook | | | | | | | | Our board shop drills boards stacked. Scott is right and to solve the problem we invested in a tooling system that allows us to align the artwork to fiducials then the tooling holes are puched. The tooling pins are then pressed through the stacks for tight alignment and works very well with the top vs. the bottom alignment. That still doesn't stop the variances with the drilling process so we periodically run test samples and use x-ray to monitor our stacking process. | | | | If you can, visit/contact Excellon. They have an automated conveyor system and may be able to help you out. | | | | Cheers, | | Scott | | | All things drill or place start with CAD data. In a board shop it is coverted using CAM and pertinent information (drill entry points)is downloaded to automated drilling equipment. This data would be an ideal place for you to start as it is simple X/Y type as Gerber (dumb data). | | Everyone has stated problems correctly concerning drill wander, spindle runout, stacking, artwork, etc. Still, you have to start at the start. Data is that point as it all is referenced from a single datum from which all other locations, features, fids, etc. are positioned. | | As far as making an accurate table, fixture, or other locating mechanism goes, it's all up for whatever works best. Original production artwork, when first generation, offers little problem with respect to registration. It's the iterations it goes through as image, etch, etc. | | Still, the data is the thing. | | Earl Moon |

Be sure your data is formated to at least five decimal points. Often overlooked, tenths of mils add up on larger boards. Go to hundredths (.00001) and the runout slows if the machines are as accurate. Dan.

reply »