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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Question for everyone regarding part placement on the SMT line (feeder position)

Views: 2468

We have had a few large human error, wrong placement in the ... - Jan 28, 2020 by Jede  

#84305

Question for everyone regarding part placement on the SMT line (feeder position) | 28 January, 2020

We have had a few large human error, wrong placement in the SMT set up over the past year(same package size placed in the wrong feeder position to where it goes unnoticed) and we are trying to look for possible process improvements to prevent this from happening. A CM that I worked for many years ago used an in house program to pull information directly from the SMT files to where you would scan the part label and it would tell you the size of feeder and the location in the feeder bank(Juki SMT).

What does your company do to prevent this type of occurrence? How successful is your process? Thank you for your response in advance.

reply »

#84307

Question for everyone regarding part placement on the SMT line (feeder position) | 28 January, 2020

If you refer to having wrong resistor or capacitor value, the only way is to automatically measure the part value. Mycronic pick and Place machines do it on the fly, ASM machines have feeder where part is measured....other main brands probably have similar ways. Any other cheks and verification that I have tried through the years failed me.

reply »

#84308

Question for everyone regarding part placement on the SMT line (feeder position) | 28 January, 2020

Let me rephrase my question. How do you avoid having errors in your SMT set up? IE xxx part goes in feeder slot 55. We currently have our quality department do a machine check but small errors like this can still go unnoticed.

reply »

#84309

Question for everyone regarding part placement on the SMT line (feeder position) | 28 January, 2020

The best I have seen was two people set up and verify and then two other people not involved with that setup verify. It worked quite well but even then a wrong part was put on. Day shift set up and Afternoon shift thought that both verifications had been done. Most people that have changed a part have at some time put a wrong part on. Even a couple managers I knew that probably only changed a few parts total in their lives managed to put a wrong part on. And it does happen where the reel has the wrong part in it from the supplier. Here we have a Cluso and a meter to check each part on the first off.

reply »

#84310

Question for everyone regarding part placement on the SMT line (feeder position) | 28 January, 2020

Thank you for your input. We do something similar but we are looking for something a little more automated so human error is reduced.

reply »

#84311

Question for everyone regarding part placement on the SMT line (feeder position) | 28 January, 2020

I genuinely think component verifier on the machine is the optimum solution. Even then it can probably get caught out, we build some instrumentation that can have 10K 1%, 10K 0.1% and 10k 0.1% 2ppm, a verifier won't catch those getting mixed up,likewise caps wih different voltage ratings- that requires good stock control and labelling.

MyCronic/Mydata both now fit these basically as standard (they might tell you its a special deal just for you ;) ) and that's because a huge section of client base is high mix, high changeover meaning this is more critical. (Well that and they have a decent share of fussy clients who measure every single part).

Other things that help:

barcodes and a scanner, not mouse clicking and reading text based part numbers. Also if you create internal part numbers do it with care.

It should be easy to tell which reel is loaded onto which feeder lane. So single/dual lanes gun feeders are arguably better than banked units. So for instance I have Essemtec CLM feeders which have up to 10 reels in a cassette packed quite tightly, checking them is a total ass.

Smart feeders so you can't put the wrong feeder in the wrong slot, or rather you can but it doesn't matter.

A separate operator who only does load out checks

Autoloading feeders probably increase the risk, so if you have those be extra careful.

Time, this most recently happened to us when running under severe time pressure, don't rush kitting teams. But by all means optimise and examine how they operate.

Above all establish how the load error happened, mis-read, mis-label, a mid-run replenishment etc.

IME its a mid-run replenishment that have typically caused this issue, luckily its quite rare.

reply »

#84312

Question for everyone regarding part placement on the SMT line (feeder position) | 28 January, 2020

We are running a Juki SMT line (3RAL and 2080E). I am not sure there is an option for a component verifier. My thought is a barcode/scanner set up and I am in the process of working on this but waiting on Juki response on a file question. We do use internal part numbers. Our last instance of wrong part placement was 2 parts that had similar MPN's that were swapped placement on the feeder bank and missed during the machine check by quality.

reply »

#84313

Question for everyone regarding part placement on the SMT line (feeder position) | 28 January, 2020

It looks like the new Jukis do..

On your srries of machines I think you're looking at this http://jukiamericas.com/smt-products/accessories-software/software/ifs_nx/" target="_blank">http://jukiamericas.com/smt-products/accessories-software/software/ifs_nx/

Well that and CVS is also listed on the options page of the brochure. So is bad mark, who the hell thinks that's an "option"? rude.

reply »

#84315

Question for everyone regarding part placement on the SMT line (feeder position) | 29 January, 2020

Barcode/scanner configuration can still lead to wrong parts mostly when changing reels. The root cause is still the line operator, because instead of scanning the newly set-up reel, he/she scans the old reel because 'it's easier' for them. I have worked with ASM Siplace machines both mass and low volume production, and we still had wrong parts built on PCBs.

reply »

#84316

Question for everyone regarding part placement on the SMT line (feeder position) | 29 January, 2020

That is no longer an error.

I guess if your reels had IDs because you were running full traceability, they wouldn't be able to do that. But those options can get quite expensive.

reply »

#84318

Question for everyone regarding part placement on the SMT line (feeder position) | 29 January, 2020

I'd recommend checking with the manufacturer of your SMT equipment and checking into the feeder verification solutions they have.

Combine that with a buddy check and true first articles before the line can continue to run the balance of the job and that'll cover almost all your bases.

reply »

#84340

Question for everyone regarding part placement on the SMT line (feeder position) | 30 January, 2020

A solution that could placed on any machine is a first check that places a component in a fixed location and then the 2 pin component is read with a meter like a HP component measuring system. Similar for transistors and FETS. IC's are probably not such a problem. But this would require software to be written.

sarason

reply »

#84351

Question for everyone regarding part placement on the SMT line (feeder position) | 31 January, 2020

Check out Cogiscan or Bartector. Im sure there are others but these are the two I've heard of. They make solutions for what you are trying to do.

reply »

#84357

Question for everyone regarding part placement on the SMT line (feeder position) | 3 February, 2020

Check into Bartector. Used this solution for 15 years at a previous employer. As long as your parts are labeled correctly (correct part number matching the mfg part number) then Bartector will have the operator scan the feeder to the part label. It automatically interfaces with a variety of brand pick and place equipment and will return a go no go if part is correct or not. Will stop the machine automatically if not correct.

Very cost effective solution also that includes many extras.

reply »

#84372

Question for everyone regarding part placement on the SMT line (feeder position) | 6 February, 2020

We use an AOI machine to check the first-off. This will pick up such errors, and if the line is stopped until the AOI operator has approved it, the problem is solved, although it would require one fix.

reply »


SMTA-Samy

#84779

Question for everyone regarding part placement on the SMT line (feeder position) | 20 April, 2020

After I create the pick & place program on the pick and place machine (Assembléon, recently aquired by kulicke and soffa), we generate setup sheets for the operator which lists what component goes into what feeder. Upon loading a tape & reel in the feeder, the S/W requests that the operator scans the Internal Part Number (IPN) along with the Qty. If this is not done, the part won't be loaded successfully.

This has proven to be really successful.

The only instance where I can envision this might not work is if the tape and reel had the correct IPN but the wrong part in it while having the same package (i.e: we're trying to place IPN 132811 which is a 2 Kohms 0603 resistor and IPN 132811 was scanned but had a 5 Kohms 0603 resistor). In this case, the pick & place machine's optical vision system won't catch it as it's using the same package. As other members mentioned, the only possible way to detect a wrong loaded component in this case is to measure the value (resistance, capacitance, inductance, etc..) on the fly.

This message was posted via the Electronics Forum @ SMTASMTA

reply »

Conformal Coating Machine under 40000

Reflow Oven