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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


100nF 0805 ceramic cap with 25 Ohm resistance? Why?

Views: 5272

#62097

100nF 0805 ceramic cap with 25 Ohm resistance? Why? | 22 June, 2010

Hi,

I have a 100nF 0805 ceramic cap (the most common type!), Yageo brand (20+ pcs. on the module). In our production run I get some errors due to conductive caps (can be anything from 25 Ohm to 5K - most are open of course). Only one cap really effects the functions, others don't matter too much.

I presume the roll may had bad caps in it (how common is that??).

But beside that, could there be other reasons?

Process is: Glue attach > Pick & Place > Wave-solder > other parts > test > conformal coating

Is there any production error that could make a ceramic cap conductive?

PS: want to add one point: The caps don't seem to be cracked. We changed a few 100 and didn't notice any broken caps after taking them off.

reply »

#62109

100nF 0805 ceramic cap with 25 Ohm resistance? Why? | 24 June, 2010

> Hi, > > I have a 100nF 0805 ceramic cap (the most > common type!), Yageo brand (20+ pcs. on the > module). In our production run I get some errors > due to conductive caps (can be anything from 25 > Ohm to 5K - most are open of course). Only one > cap really effects the functions, others don't > matter too much. > > I presume the roll may had > bad caps in it (how common is that??). > > But > beside that, could there be other > reasons? > > Process is: Glue attach _ Pick & > Place _ Wave-solder _ other parts _ test _ > conformal coating > > Is there any production > error that could make a ceramic cap > conductive? > > PS: want to add one point: The > caps don't seem to be cracked. We changed a few > 100 and didn't notice any broken caps after > taking them off.

It has been our experience that cracked caps test as shorts or low ohms. That error is usually caused by some sort of mechanical strain on the part, such as a cap near the edge of a board that gets singulated.

reply »

#62112

100nF 0805 ceramic cap with 25 Ohm resistance? Why? | 25 June, 2010

I got a report from Yageo yesterday with microsections and and and..... result: Cracks

I did a test run on boards after ICT, but before coating. They are all ok. Seems the cracking happens somewhere after the ICT. But at that stage all assembly is already finished and we do just:

Conformal coating Heat cure Function test Packing

The cap in question is somewhere in the middle of the PCB bottom side. Seems I have to do more investigation.

reply »

#62113

100nF 0805 ceramic cap with 25 Ohm resistance? Why? | 25 June, 2010

Do you have to 'break' the pcbs from a panel or biscuit or are they individual pcbs form the start ? Poor pcb separation techniques or processes will damage ceramic caps.

reply »

#62114

100nF 0805 ceramic cap with 25 Ohm resistance? Why? | 25 June, 2010

They are panels of 2 and we break them before ICT. Have to say that we do not check that 100nF cap in ICT since it's parallel with some eCaps.

reply »

#62140

100nF 0805 ceramic cap with 25 Ohm resistance? Why? | 29 June, 2010

For the capacitors that Yageo analyzed... Were they taken out at representative points in the process?? ie. were samples taken from pre-assembly, pre-ICT, post-conformal coat, post-final test?

You note that the parts 'seem' to be ok up to ICT. And that Yageo's analysis indicated micro-fractures. It'd be important to analyze parts from representative points in your process to determine if your process is causing the issue.

Potential items in your process that could cause this issue: 1. Glue curing. The smt glue could be presenting uneven strain on the component. 2. Wave solder. The glued components may not be centered on the pads, or the board could be thermally unbalanced, presenting uneven stress on the components. 3. Excess moisture in the parts...same effect in either of the above two process steps. Check your material handling, and ambient environment. 4. ICT. Not terribly likely for an ICT machine to cause 'micro-fractures.' If the cause were ICT, it would more likely be obvious to the human eye. And the components would fail. But, I suppose there's a chance that the target pin is too close to the component, and is presenting some increased stress on one side of the cap. 5. Conformal coating/cure. If the coating is placed by machine, ensure that the gantry system isn't hitting any of the parts. UV cure of uneven amounts of conformal coating could cause uneven stress on a component in the same manner that un-balanced copper can causes undo stress during a reflow evolution.

cheers ..rob

reply »

#62144

100nF 0805 ceramic cap with 25 Ohm resistance? Why? | 30 June, 2010

Thanks for your reply. This production lot is already finished. We produced in 2 lots, first was no-clean (and was ok), 2nd lot was cleaned (aqua based), and I need to look into the cleaning further.

Since it's all finished I can't get parts from several steps. Since most 100nF's are parallel with other caps most are not tested in ICT, except 1, and that was always OK. So it's more that particular location.

Your points: 1. We do that board with glue for 5 years, so I feel it's a bit unlikely.

2. The look well centered.

3. Moisture - Assembly is in humid China, need to look into that, although assembly was always in China.

4. ICT. Also no real change in the ICT fixture, although I did change to higher force pins recently.

5. Conformal coating is spray coated by hand and then heat cured at 70C or so.

I will have to observe the assembly line for a few days I guess......

reply »

#62148

100nF 0805 ceramic cap with 25 Ohm resistance? Why? | 30 June, 2010

In general, if the process has been consistent for a period of time, then one should look at any new variables in the process.

Therefore...the supposition that there's a bad lot of components is a very good one. Did Yageo offer any possible explanations for the damage they found in the failure analysis? ie. stress fractures, impact fractures, burst fractures, etc. If so, that could point you in a direction to look.

Failing that, I'd say determine which variables recently experienced changes in the process. You noted new ICT pins, that's one point to look at. Also, fatigue in ICT fixturing is a possiblity, affecting the center point over time. Is the conformal coating process new? It's an additional thermal cycling step in the process, and could have an affect.

cheers ..rob

reply »

#62224

100nF 0805 ceramic cap with 25 Ohm resistance? Why? | 12 July, 2010

After a detailed check the cracking was most likely from the ICT and a variation in the length of the probes (due to wear and tear) that press down the PCB. I am very certain that was the reason.

reply »

ii-feed SMT Intelligent Feeder

SMT Equipment Service