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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Cracking in High Voltage Chip Caps

Gary Simbulan

#15719

Cracking in High Voltage Chip Caps | 26 May, 1998

I know, I've read the "been there done that" stories about this subject, but I still profess ignorance on some aspects. We are using NPO type dialectric caps in potted high voltage assemblies and continue to experience catastrophic vailures. Upon digging them out or cross sectioning there is massive cracking at the bottom (nearest the PWB) near the end terminations. The boards are FR4. I don't have any specs on the potting material. The question raised by our components engineering group is, is it possible to tell whether these failures are thermally or mechanically induced. The focus of design engineering up until now has been on process. Are our reflow temps ok, are the parts hand placed and / or soldered. Does our rework process, when used, follow a controlled profile. We have written a new in house rework procedure specifying a ??expensive?? rework station, and solicited from the vendors rework instructions. Now the components group is asking if we can even identify thermal stress as a cause. Is it possible after failure to be sure of the cause. I know that the different CTEs of the materials must play some role, but again, how to differentiate. Any thoughts? Gary

reply »

Earl Moon

#15721

Re: Cracking in High Voltage Chip Caps | 26 May, 1998

| I know, I've read the "been there done that" stories about this subject, but I still profess ignorance on some aspects. | We are using NPO type dialectric caps in potted high voltage assemblies and continue to experience catastrophic vailures. Upon digging them out or cross sectioning there is massive cracking at the bottom (nearest the PWB) near the end terminations. The boards are FR4. I don't have any specs on the potting material. | The question raised by our components engineering group is, is it possible to tell whether these failures are thermally or mechanically induced. The focus of design engineering up until now has been on process. Are our reflow temps ok, are the parts hand placed and / or soldered. Does our rework process, when used, follow a controlled profile. We have written a new in house rework procedure specifying a ??expensive?? rework station, and solicited from the vendors rework instructions. Now the components group is asking if we can even identify thermal stress as a cause. | Is it possible after failure to be sure of the cause. I know that the different CTEs of the materials must play some role, but again, how to differentiate. | Any thoughts? | Gary I don't, obviously, know your complete situation. However, I recently consulted for an R/F radio manufacturer with similar problems on many chip or passive device types. At first they were sure the cracking was mechanically induced. After reviewing their designs, it was easily determined they were using unacceptable design rules as SMT solder termination areas with no neckdown features. This not only caused serious rework problems and much hand soldering, but it caused thermally induced cracking at the device end caps as you describe. As FR-4, and most other resin system type boards, has a surface TCE of about 15-17 ppm compared with ceramic of about 5-7 ppm, this can cause problems. However, when this factor is combined with no neckdowns, the copper foil (17-19 ppm/C.) termination areas move with great force against end termination areas causing cracks - agan as you describe. All this coupled with whatever potting resin system you are using, and mechanical stress or shock, adds to problems. Also, in the case I mention, some of the devices were sub standard parts exhibiting poor solderability thus requiring too much touchup (once is too much) again adding to the problem. Please tell us more about your findings or other factors you might think of concerning your cracking problem. Earl Moon This

reply »

Jerry

#15722

Some ideas on Cracking in High Voltage Chip Caps | 26 May, 1998

Our experience has been that mechanical crack occur vertically through the chip. In other words perpendicular to the board surface. You can sometimes tell through close inspection that the crack may have propogated from a point that the nozzle may have contacted. You might want to confirm this through close inspection after placement or placing lots on parts on sticky tape. I think the signature of thermal cracks are horizontal propogation usually starting at the point where the end cap solder meets the ceramic body since this is a high stress point. You might want to confirm this by hand placement of a large sample of devices, process, then closely inspect.

reply »

Scott Cook

#15720

Re: Cracking in High Voltage Chip Caps | 27 May, 1998

| The question raised by our components engineering group is, is it possible to tell whether these failures are thermally or mechanically induced. The focus of design engineering up until now has been on process. Are our reflow temps ok, are the parts hand placed and / or soldered. Does our rework process, when used, follow a controlled profile. We have written a new in house rework procedure specifying a ??expensive?? rework station, and solicited from the vendors rework instructions. Now the components group is asking if we can even identify thermal stress as a cause. | Is it possible after failure to be sure of the cause. I know that the different CTEs of the materials must play some role, but again, how to differentiate. | Any thoughts? | Gary Gary, I'd not try to second guess the game. Send samples of these caps (attached to the substrate, if you damage them removing them) to several capacitor manufacturers. They will do destructive testing, take great pics, and write you a report of their findings. Use several vendors in order to keep 'em honest........ This service is free, and they can pinpoint whether you've got mechanical stresses or thermal stresses propagating the microfractures. Scott

reply »

One stop service for all SMT and PCB needs

SMT fluid dispensing