Cap Short

Subscribe to Thread

I'm experiencing an impedance breakdown on an 0805 ceramic c... - Nov 21, 2004 by

On the dendritic growth over the top of the capacitor: We ag... - Nov 22, 2004 by davef

How could I determine the effectiveness of the cap vendor's ... - Nov 22, 2004 by

Assemblers often assume that the board and component supplie... - Nov 22, 2004 by davef

John

#31456

Cap Short | 21 November, 2004

I'm experiencing an impedance breakdown on an 0805 ceramic cap. In the application, the cap is at a 16V bias at all times. The unit failed in an extremely high humidity condition. We are using a NC flux and have determined there was flux underneath the cap. There were also metallic fragments under the cap composed of lead, oxygen, tin, nickel, possibly aluminum, and possibly carbon. No halides were detected. In severe humidity testing conducted on new units (95% humidity 65C completely unsealed), dendritic growth was observed over the top of the cap, but nothing underneath. Could this material have come from the PCB vendor's process? Any thoughts would be greatly appreciated. Thanks John

reply »

davef

#31463

Cap Short | 22 November, 2004

On the dendritic growth over the top of the capacitor: We agree that your capacitor supplier probably did not do a very good job of cleaning process residues from their device.

On the metallic fragments under the capacitor: This sounds like solder balls, probably from your attachment process. We doubt the aluminum was present. The carbon could be part of flux residues.

reply »

John S

#31468

Cap Short | 22 November, 2004

How could I determine the effectiveness of the cap vendor's cleaning process? Thanks again, John

reply »

davef

#31469

Cap Short | 22 November, 2004

Assemblers often assume that the board and component suppliers provide them with clean product. This was not a problem when everyone used water washable fluxes, because the water washing used to clean the flux residues also cleaned the board and components.

No-cleaners can not be so blase, because they never clean these components. They need to monitor the ionic contamination on their in-bound material. This becomes urgent where there's ionic material, humidity, and a bias voltage; the three factors that affect dendrites.

Water soluble fluxes leave ionic residue detectable by: * ROSE testing (ie, Omega meter, Ionograph....). This test is useful to determine whether the cleaner is in control. Commonly used throughout the industry. * Ion chromotagraphy. This test determines what specific residue causes failure. Only large shops can maintain the equipment. So, generally, smaller shops use outside laboratories. * Surface Insulation Resistance. This test actually has a correlation with field failures. None relating to residue if resistivity is above a certain level. Guaranteed if its below a certain level. The level depends upon environment, panel design and circuit design (voltage, current, impedance, spacing, indoors vs. under the hood...). This test should be done on test coupons in an environmental test lab. It is rarely done as a process control, since the test takes a week, and is not usually done on production panels.

reply »