Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.

SMTnet
»
Electronics Forum
»
ceramic cap cracking

ceramic cap cracking

Subscribe to Thread

we have a capacitor cracking on one of our boards, there are... - Feb 02, 2005 by

First, we've seen no documentation that even addresses delta... - Feb 02, 2005 by davef

OK, component wide portion runs with the conveyor, so real... - Feb 03, 2005 by

The component orientation was just for the calculation. It ... - Feb 03, 2005 by davef

... - Feb 03, 2005 by Mark M.

Try heating a blank board on a hotplate to 183 deg C. with a... - Feb 03, 2005 by

Lets get to the simple things. Wave or Reflow? Wave solder... - Feb 03, 2005 by

Just a basic suggestion.... Inspect the nozzle placing the ... - Feb 05, 2005 by Etienne Ellul

I see that this problem has been already raised in 1998, but... - Feb 07, 2005 by Etienne Ellul

we are still looking into possible causes, we have rigged up... - Feb 07, 2005 by

Dougs

#32408

ceramic cap cracking | 2 February, 2005

we have a capacitor cracking on one of our boards, there are four of the same cap on the board and only one cracks, we are seeing about 6% of the component cracking. it's not near the edges of the board and he board is 2.4mm thick so i dont think it's to do with the board flexing. one side of the component is connected to a thick ground plane and when i attach thermocouples to both sides and run a profile i've noticed that during the cooling stage the side that has the ground plane cools much quicker than the side without, for most of the cooling process there is around a 20C difference between the sides.

would difference in temp cause the component to crack? anyone seen this type of thing before?

the cap is 0805 size and the crack is between the end termination and cap body.

cheers

reply »

davef

#32428

ceramic cap cracking | 2 February, 2005

First, we've seen no documentation that even addresses delta across and 0805 capacitor.

Second, based on our analysis [that follows], we'd expect to see ~0.5*C temperatore delta between 0805 capacitor pads. Figuring this, say: * 2.5*C/sec is the maximum cooling rate for the part. * Oven conveyor runs at 25 inch/min. * Component is 0.08 inch long. * Component long portion runs with the conveyor.

Here we go: * Changing units for the maximum delta temperature: 2.5*C/sec => 150*C/min * Determining the distance it takes to lower the delta temperature at the current conveyor speed: [150*C/min]/[25 inch/min] = 6*C/inch * Calculating how much of this 6*C/inch would be spread along the length of the 0805 capacitor: [6*C/inch]* 0.08 inch = ~0.5*C

Finally, many times you can guess at the cause of a ceramic capacitor failure by the pattern of the crack. Describe yours in more detail, please.

reply »

Dougs

#32431

ceramic cap cracking | 3 February, 2005

OK, component wide portion runs with the conveyor, so really both sides should heat and cool at the same rate.

The crack is at the termination between end cap and ceramic body and run vertically up the end cap, so when we remove the cap it comes off in two pieces one side is the component body and a end termination and the other side is just a end termination. looks like it's mechanical damage but the board is solid and there is nowhere in the process that wuld be bending it, we dont think it's happening during component placement as there are four more of the same component on the board that are always ok.

reply »

davef

#32432

ceramic cap cracking | 3 February, 2005

The component orientation was just for the calculation. It has no affect on what you're seeing.

As an alternate explanation, it's possible that difference in temperature is causing the board to flex in the z-direction [as you say, it may not be flexing in the x or y direction]. This z-flex could be breaking the capacitor. Consider talking to your: * Designer about respining the board to provide thermal relief on the ground end of the component. * Component engineer about a sturdier capacitor.

Here is a Calce study for more about broken ceramic capcitors: http://www.calce.umd.edu/whats_new/2001/design.pdf

For as much heat as that ground plane sucks from the solder connection, it's a bit surprising that you get a good solder flow.

reply »

Mark M.

#32433

ceramic cap cracking | 3 February, 2005

http://www.smtnet.com//forums/index.cfm?fuseaction=view_thread&CFApp=1&Thread_ID=3628&#Message14433

reply »

Philj

#32434

ceramic cap cracking | 3 February, 2005

Try heating a blank board on a hotplate to 183 deg C. with a board this thick, if the layer construction is poorly balanced, it may bend enough in the region of the ceramic capacitor for the soldered joints to crack it on cooling to ambient.

reply »

Chunks

#32435

ceramic cap cracking | 3 February, 2005

Lets get to the simple things. Wave or Reflow? Wave soldered parts have a higher susceptibility to physical damage until they are waved. That�s were I normally see this type of break.

Usually, the crack is a circular pattern in the cap if it�s from Delta T. But that�s mostly from thermal shock � generally a huge ground plane cause solder issues, not broken parts issue. Why not use a repair station and heat this area up and let it cool to verify your findings?

reply »

Etienne Ellul

#32453

ceramic cap cracking | 5 February, 2005

Just a basic suggestion.... Inspect the nozzle placing the component...because that is the only medium, apart from the PCB design, which can be causing the damage. It occurred in my Department.

reply »

Etienne Ellul

#32466

ceramic cap cracking | 7 February, 2005

I see that this problem has been already raised in 1998, but unfortunately at that time the author raising the problem did not feedback whether he solved the problem or not. So this time I would like to ask Dougs, to feedback any solutions to his problem, because this way we keep learning from each other.

reply »

Dougs

#32467

ceramic cap cracking | 7 February, 2005

we are still looking into possible causes, we have rigged up a test to determine whether the cap is cracked and have ran a controlled batch off of SMT, these were all tested and look OK. I'm looking into buying some small strain gages to place over the area where the cap is cracking to see if there is any physical stress in that area of the board during the manufacturing process, these boiards take a while to feed through our process due to the high amounts of hand assy time and test time, plus they need to be sent out to a contractor to have some work done on them. as soon as i find out what is causing it i'll post it on the thread.

reply »