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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.

Placement Program Control


Placement Program Control | 3 August, 2001

What do yall do to control the configuration of placement programs written for a specific product?

We always just ran what was on the machine, If the first piece looked good, we ran it. If the first piece looked bad, we fixed and then ran it. Now, we're talking about archiving a "good" program off-machine and la-de-da.

reply »



Placement Program Control | 3 August, 2001


We archive programs as follows.

Step1. Perform a successful first piece article

Step2: Copy the Program to a password protects network drive.

Step3: Run the program in production, from the local drive

Step4: After the production run, delete the program from the local drive

Step5: When the program needs to be run again copy it from the network drive to the local drive.

Step6: Religiously backup the network drives.

With this method I don�t have to worry about a corrupt programs or worry about operators changing the program. Every time I run the job I�m working with golden data

Good Luck Jay B

reply »


Placement Program Control | 6 August, 2001


so what�s a good program anyhow? What about the setup configuration ? ... and who�s gonna say that your so optimized program will run after month and month while your equipment has gone through several aging and whatever processes?

So we all know that a first testrun is always necessary ... and some finetuning.

Better file the good known production data and not only the program and give it the actual revision of that product. You won�t check your filed production data if a revision is changed but you have to do that when the job is on your schedule and make the necessary changes than. (that�s our daily live, ... up to 6 revisions between jobs)

Once we have incorporated the new revisions the old data works quite well, a few adjustments of visionfiles or a slight offset for the new PCB batch that�s it.

The main part with old data turned out to be the revision record to get it right fast.

We don�t generate complete new data but we fear to put old ones on the line without checking as if it were the first time.

Best wishes


PS: You wouldn�t, if you have found the right tire treatment for your bike that prevents you from looking like a pizza after this 50 ton truck has ..... , put the tires in your safe, get them out after a couple of month (years) and use them as they are when going that same road where this truck happens to push you over the cliffs.

reply »

Chuck N.


Placement Program Control | 17 August, 2001

I like the idea that files are stored on a Lan system. Typically lans are backed up daily or weekly by the IT department. If not they should be backed up on tape or disk weekly, monthly, etc. depending on how often new programs are released. The more programs that are released the more often the machine programs should be backed up.

The Mfg. Eng. or programmer should be the only one allowed to change the program and set up. The revision levels are controlled by some sort of manufacturing documentation system. I have seen manufacturing revision levels controlled by Doc Control, Quality, or Mfg. Eng. or some combination of the above. Typically the current program and set up revisions are shown on a documentation traveler or process router.

Below are guidelines I use.

There are typically two parts of the program that will require naming when programming. a) Set up file (AR file) b) Placement file (LA file)

Set Up The set up file shall follow the naming convention shown below. Where X is the line (A, B, C) Where Y is the next available revision starting at A and continuing to Z and then AA, AB, AC, etc.

Set up revisions shall increment each time a part number in the set up changes or the location of feeder(s) change. The revision will not change if a correction is made to fix an input error.

Placement The placement file shall follow the naming convention shown below. A �T� for top or a �B� for bottom, the last 5 digits (more than 5 numbers could be used depending on how many characters the machine allows for a name) of the part number of the circuit assembly, REV_X where X is the revision of the BOM and a digit from 1 to 9 for any process changes or change to the panel layout. E.G. The revision letter for the placement file will change only when the BOM changes per ECR. The process number shall change only for process improvements or changes to the panel layout and not for BOM changes. E.G. BOM changes from L to M. Using the example above, the new name would be E.G. The panel layout changes to show a 10 mm scrap edge instead of a 9 mm scrap edge. The new name would be The revision will not change if: A correction is made to fix an input error Minor adjustments are made to the placement coordinates of components for more accurate placements.

I hope this helps you out.

PS. Your comments are always good for a laugh.

reply »


Placement Program Control | 20 August, 2001

Hi Dave, for several reasons the machine computer(s) should be connected to the LAN. 1. The machine computer can be as old as a 486 with a couple of hundred Mb hard drive and the more programs you store in there, the slower it gets. 2. If you put all the programs on the server and download the files when required, you keep the computer at least at this speed and deal only with the files you need for this program. 3. I recommend to empty all files, including the package library, component library, fiducial library and management information data when you start a new job. 4. You can back up the entire machine software via LAN, which is much faster than tape back up or RW CD.

I agree with Wolfgang, that a �known good program� may have been tweaked several times (fiducials, pick up position, component handling parameters and placement position), together with different revision level of software, that it is sometimes difficult to say when it was really good. I guess, we will have to differentiate between the placement program with its X,Y theta values, set up files and machine specific handling data like nozzles and camera lights. Unfortunately, handling data are in a pretty fuzzy area and of course different from machine to machine. All changes to the handling data are usually only temporary and should be done on the line (station) computer, but should never overwrite the data on the server, which were created on �day one�. Please also note that the changes you make to a feeder pick position are not be worth keeping because when you run the job again, you may use a different feeder instead. Finally, I must admit, that I don�t trust too much in the server and keep a hardcopy of the program together with an old fashion disc in a folder, from the time on the job run the first time successfully. Stefan

reply »

Panasonic CM402, CM602, DT401

Six Sigma Training