Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.

SMTnet
»
Electronics Forum
»
First piece inspection, 0402

First piece inspection, 0402

Subscribe to Thread

What in process methods are being used to validate the value... - Jan 07, 2002 by Charlie

Use the same method us use for SMT capacitors with no markin... - Jan 07, 2002 by davef

Which method do YOU use ???? ... - Jan 08, 2002 by stefano bolleri

Dun you have a LCR meter? or mebbe your side calls it a LCZ ... - Jan 08, 2002 by ianchan

Stephano: Since you have not thought through this, it indic... - Jan 08, 2002 by davef

Our standard is: Accept and/or reject decisions must be bas... - Jan 08, 2002 by davef

Since we are on the topic, What we do over here, is to ge... - Jan 09, 2002 by ianchan

Hi Ianchan, you mean you set up a line and wait 6h or 2h ... - Jan 10, 2002 by Wolfgang Busko

For repeated model runs, we go ahead with the SMT run, whils... - Jan 11, 2002 by ianchan

It�s up to you what and how you perform. I was only wonderin... - Jan 11, 2002 by Wolfgang Busko

Charlie

#18583

First piece inspection, 0402 | 7 January, 2002

What in process methods are being used to validate the value of 0402 resistors during first piece inspection?

reply »

davef

#18589

First piece inspection, 0402 | 7 January, 2002

Use the same method us use for SMT capacitors with no marking.

reply »

stefano bolleri

#18593

First piece inspection, 0402 | 8 January, 2002

Which method do YOU use ????

reply »

ianchan

#18594

First piece inspection, 0402 | 8 January, 2002

Dun you have a LCR meter? or mebbe your side calls it a LCZ meter??? whatever...it measures Inductance, Ohms, Capacitance...check-it-out-manz...

reply »

davef

#18601

First piece inspection, 0402 | 8 January, 2002

Stephano: Since you have not thought through this, it indicates that your discrepancy reporting and resolution process probably requires critical review. Consider highlighting this issue to your quality assurance staff.

reply »

davef

#18602

First piece inspection, 0402 | 8 January, 2002

Our standard is: Accept and/or reject decisions must be based on applicable documentation. As an element of this, the marking on the part must identify the part positively.

If the marking on the part does not identify the part positively, the inspector verifies the component by comparing a component on the assembly with a component from of the original reel. Both parts must match (ie, color, termination style, width, length, etc.).

If the part can still not be identified positively, either: * Remove and test a component with the appropriate test equipment. � OR * Use automated testing as viable alternative to visual inspection.

reply »

ianchan

#18621

First piece inspection | 9 January, 2002

Since we are on the topic,

What we do over here, is to get SMT Technical to machine pick&place "sample component layouts" onto double-sided tape, which in turn is placed onto an actual PCB...

QA will take that double-sided tape sample, and using an LCR meter measures all component values. The values must match the specs.& tol% specified within the BOM. Components without measurable values, are QA confirmed by bodymarking or legend-codes. Lastly, QA will confirm the polarity orientation of the same mounted components. Only when QA has approved the double-sided tape sample, will SMT Production proceed with the Lot run.

Suggest having inspection drawings avaliable, that depict accurate PCB layout of components, and orientations, and displays "location-to-value" information (eg. R2 - 10ohms +/-1%)

Typical target time of First Piece inspection, is 6h for new model runs; and 2h for repeated model runs.

Fluke has a range of LCR meters to choose from, of coz they ain't they only meter suppliers around...

PS: Dun advise visual confirmation of components based on strictly only bodymarking methods, coz say you have : 1) SMD 0603, 100ohm, +/-5% tol 2) SMD 0603, 100ohm, +/-1% tol

if both have the same bodymarking, how you gonna know and confirm which SMD has what tol%??? It dun matter in class1 level boards, but it sure can screw up your functional test yield data, if a 5% replaces the original required 1% in a RFID/GPRS board.

When I was with a RF design company, we once couldn't get the correct waveform & signitures, and kept going around in circles trying to debug the buggy...quit making up excuses for frustration generation and do things right...

Happy Inspection!

reply »

Wolfgang Busko

#18623

First piece inspection | 10 January, 2002

Hi Ianchan,

you mean you set up a line and wait 6h or 2h to go on with production ?

Can�t believe that!

Wolfgang

reply »

ianchan

#18638

First piece inspection | 11 January, 2002

For repeated model runs, we go ahead with the SMT run, whilst the QA does the FA inspection. Limit FA timing to 2h or less...(simple boards yield faster time cycles).

For new model runs, yep, we wait for QA to confirm the FA inspection, only with QA green light, does SMT run commence...(the pcba gets to bask in the holy halo of QA approval)....this is limit to FA within 6h or less.

Why is this dwell time, so shocking to y'all???

BTW, we are into small Lot R&D prototype runs.... We usually transfer confirm mass production runs, to our off-shore sister plant.

reply »

Wolfgang Busko

#18639

First piece inspection | 11 January, 2002

It�s up to you what and how you perform. I was only wondering how your figures match with todays performance efforts and predictions made by the industry. The stress today is obviously on change over times and concepts to minimize this time especially with high mix low volume production.

Looks like we are competing in the assembly olympics fighting for seconds, milliseconds, percentage etc. ... at least this is the impression one could have after visiting a presentation or show.

For prototyping things are still different and can�t be compared with production runs. It often shows that the step from prototyping to production looks like just starting all over cause you have different teams and equipment and much tougher limits regarding speed, yield and costs.

Seems reality and advertisement still do not match that very well.

Wolfgang, (who thinks we all can do a bit better)

reply »