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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Barcoding Component Reels

#3300

Barcoding Component Reels | 15 August, 2000

Previously my company manufactured simple 8-10 component through hole parts. Now we have put a mixed technology line in place and are building parts with 50-100 components each. With our older units, functional test and simple visual inspection was enough to ensure that the correct components were on the board. Often our parts are labled by our receiving department, and it has happened twice recently that mislabled parts have entered my production line. We are putting in-circuit test in place to catch anything that does get to the line, but I'd like to prevent that as much as possible. How are parts controlled in large plants that have hundreds or thousands of part numbers? Most (if not all) of our parts are barcoded by the manufacturer. Are they read by barcode as they are received and again as they are placed on the SMT placers? Are "canned" barcode systems available? Any advice would be appreciated as I have just spent the morning changing a thousand diodes.

John

reply »

Charles S.

#3301

Re: Barcoding Component Reels | 16 August, 2000

We have had a similar problem here and looked to using a bar code reader system available from our pick and place machine manufacturer (Siemens) the big downside of this was that we needed to maintain a barcode database for all our components to ensure that when we scanned the barcode it was recognised as the correct part. Our final solution was to buy a simple programable hand held scanner with a display. All we do is compare the barcode of the recieved stock/ reel to be loaded against one already in stock/just run out and if they are the same no problem- if they are different then it beeps and tells the operator to seek advice. Although this system will stop every time you have supply from a different manufacturer (maybe a good thing) it covers most of our supply situations. All I wish now is that I could get tubes of ICs individually barcoded! We have been running this system for a while now (approx 50 million components placed)and in that time I have not had a wrong reel loaded.............famous last words.

reply »

#3302

Re: Barcoding Component Reels | 16 August, 2000

John: It's easy to put parts in the wrong place!!! Isn't it?

Starting at the beginning: check the part at incoming inspection ... measure the value upon receiving the part, if it's possible. Verify the marking on the part.

Next: Prevent things from being mis-labled at incoming receiving inspection. This happens more times than the vendor sending us the wrong part. We label parts tubes, carriers, ESD bags, whatnot with a Brady TLS2200 (www.bradyid.com) or a label printer incorporated in the fancy scale we use.

Next: Most of the time a part being loaded wrong on a board has been because of the operator putting the wrong part in a feeder ... We print feeder labels using the Brady and inspect the feeder set-up as a "first piece".

Next: We always do a "first piece" on a assembly when the first board comes off the line just to make sure the placement program is correct. We run short programs to check just reflowed parts on the MDA. Although hot caps don�t always test well, but hey you can check those visually!!! If everything is verified correct, we run.

Next: We use a "Reel Change Log" to record changes during a run to make sure the right part is loaded into the feeder. When a part runs out on a feeder, there needs to be a "second set of eyes" look at the part (measuring it if need be as in the case of capacitors) and also verifying it to the feeder position on the machine set-up sheet. This just gets people to stop and slow down to MAKE SURE they're putting the RIGHT parts in the CORRECT feeders!!!

Next: We do not splice reels, its too complicated ... to easy for us to mess-up

reply »

#3303

Re: Barcoding Component Reels | 17 August, 2000

Getting things "foolproof" is sometimes a hard effort and can cost a lot of money. Who is able to tell that it actually saves money instead of wasting it. Transfering responsibility from operators to a system may lead to a somewhat less responsible attitude to the job. In my experience operators detected as much errors regarding wrong labeled material as they did put on wrong parts on feeders or changed the direction of IC-sticks. (Wouldn�t it be nice to record not only the errors they make but also the errors they correct?) Systems do help getting the right parts from purchase to the PCB and the use of checklists (Daves list seems perfect)is useful as a reminder for the operators to do their job. Once learned the nessecity and importance of the checklist and other processcontrol measures operators hopefully cease to feel that they were treated as "fools". In the end we all dream of machinery and systems that will look for everything and correct it automatically, check each part for the correct value, orientation and damage, for a price less than anything today. Until than teach your people and treat them nice and make sure they know and understand about the impact errors have on your whole enterprise.

Quality got to be lived not only talked about. The more people know about one thing the better they can do the job. Feed the machine with resistors and not with "Yellow things"

Just some thoughts

Wolfgang

reply »

#3304

Re: Barcoding Component Reels | 17 August, 2000

We had an interesting problem this week that is along this thread. Had a project that had the same resistance value components, but one component was a 1% tolerance and the other was a 1/2%. Guess what? the operator got them mixed up. We have a first article inspection where they bridge the components. Both components passed because the 1% was within the 1/2% tolerances. AOI missed it because the markings on the components were the same... We started the job with the components mixed on the machine. We have the same procedures as Dave describe, but it missed this type of error. Any suggestions for how to eliminate this type of error? Only thing that I've have come up with is to verify the feeder setup on the machine before running.

reply »

#3305

Re: Barcoding Component Reels | 17 August, 2000

In this case a marking of some kind that clearly distinguishes between those different parts is essential. Different stocknumber for example will do and can only be trusted if measurements like in this case do not give 100% clearness. If you have the opportunity to check the value in your machine you could have this switched on for this particular part.

Wolfgang

reply »

#3306

Re: Barcoding Component Reels | 17 August, 2000

Thanks for the responses. Actually we already use a reel change log with two verifications, a first off checklist with two verificaitons, and a first off inspection with two verifications. Unfortunately, the part numbers were pretty close, and the two parts are visually identical. The problem goes back to our receiving department not having a system (no checklist, no reference to correspond our part number to the vendor's, if it looks like this it must be this)in place to ensure that they put the correct part number on components that are received unlabled. Fortunately, it has gotten enough attention that there will be some procedure to get this correct. I was hoping someone has had experience or could recommend a reliable, easy to use method. We receive too many parts without our part number on them to have a quality inspector run verify the part for our receiving personnel.

Thanks for all the comments John

reply »

Jeff

#3307

Re: Barcoding Component Reels | 1 November, 2000

charles,

i can appreciate the hand held "reel to reel" scanner as a good solution that is very cost effective. during my smt supervision days we had a number of part loading checks. 1. load plan check, 2. scan check, 3. reel to reel manual check. we also had the operators stamp the reels we loaded so there was accountability if something went wrong. two separate shifts reduced their wrong part loading frequency 85% and each achieved 1 wrong part due to operator noncompliance in 75,000 part laods within one year (the first six months was quite an adventure - changing attitudes in 50% of the work force is pretty difficult). both these shifts were 12 hour grave shifts. One shift went 75,000 part loads error free including the last three months i supervised them and the first three months of my replacement.

i'm in a much lower tech environment now. the scanners don't exist in my current work place. We use IC tubes.

i REALLY like the automated real to real scanning system you described for my current situation. i take my responsibility to provide a workable process seriously (iow, i can't "correct" the humanity out of people!).

would you be willing to share the details, including mfr, cost, etc.?

tia...

reply »

ICT Total SMT line Provider

Reflow Oven