Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Strange behavior of SMD Ceramic Cap (MLCC)

Status
Not open for further replies.

hm_fa_da

Full Member level 5
Joined
Sep 16, 2003
Messages
287
Helped
10
Reputation
20
Reaction score
4
Trophy points
1,298
Activity points
3,217
Hi,

Recently i'm facing a problem in a device (many ones) that the input Key of it seems to be pressed.
the device has 4 arrow keys as inputs. the malfunction is like one or more than one of the keys are pressed. i checked mechanical structure of the keys and found no problem, the problem exist even after removing cable between board and keys.
then i checked input circuit, one 470K pull up resistor + 10nF bypass cap is used in the input circuit as below:
cap.png

I measured the voltage on input pin of mcu, in the inputs without problem the voltage is 3.3V (through pull up resistor). but in the inputs with problem, the voltage is very low, near 0V. so the MCU senses it as SW is pressed.
The only route to 0V can be the 10nF MLCC.

it was right, i tested in many devices (same device):
1- When i use soldering Iron on the cap by just touching the solder pads and melt the solder, the problem is fixed !
2- When i remove the cap by breaking it by a tool (I didn't want to use soldering heat to remove the cap), the problem is fixed !
3- When i cut the route between the cap and input, the voltage on MCU input pin rises to 3.3V (through pull up resistor).
4- I measured capacitance of the cap by meter, it shows around 10nF. but when i use ohm meter on the cap (which is electronically isolated from other components) in the failed caps, it shows around 1K resistance !

The point is that all these problem will be fixed if i only use (not even change or remove the cap) soldering iron on two pads of the cap ...

seems there is locked malfunction in the cap that will be removed by soldering iron toughing it. note that there is no dead soldering problem, even if there was, it would make disconnection not short or low resistance connection.

Another point is that the all devices with this problem are installed in outdoor conditions for +5 years. in a humid weather area.
I doubt for moisture problems ...

What do you think about the problem reasons ?
 

MLCC caps do go bad. It's possible that the cap(s) were bad from the very beginning. You don't tell us the size of the SMD, but know that it is possible to have a solder bridge underneath the cap due to too much solder. If the cap is cracked underneath one of the terminations, it could look like what you are seeing and also possibly "heal" itself with the reflowing of the solder.

Did you try to replace the cap with one from a different batch?

If this were my failure, I would have the offending caps sectioned and photographed every 5 to 10 mills. Everything you have tried seems to implicate the cap.
 
A 470k pull up resistor seems to me to be much to high in value, 10K would be more reasonable value. 470K would mean that the pin is almost floating and would account for your problems.
 
Hi,

Yes, I see the 470k value as too high, too.

Flux residuals, humidity, corrosive environment, salt...may cause leakage currnts on the PCB
Micro cracks in the capacitor, caused by bad soldering temperature profile, mechanical stress from breaking V-cutted PCBs or vibration...plus the environment above may cause leakage currents in the capacitor...

Klaus
 
Thank you all,
The size of the cap is 0805 and size of resistor 0603.
Replacing the cap solves problem, even without replace as i described above.
being of solder bridge underneath the cap is not right here, these devices have failure after +5 years of working ... the visible quality of soldering seems fine.
About the pull up resistor, i am also agreed with you, but even i saw another point (maybe design mistake ?) is that only 3 of 4 inputs have pull up resistors !
the fourth doesn't have it at all ...
The MCU used in device is Renesas HD64F2239 and according to datasheet i checked it has pull up resistor inside the chip. however i can't realize the reason of not using pull up resistor in fourth input pin like the others.

The device i am working on is Elster EK220 volume corrector, another point is that it doesn't have ANY change or on board design at least since 2009 up to recent produced ones.
 

1- When i use soldering Iron on the cap by just touching the solder pads and melt the solder, the problem is fixed !
2- When i remove the cap by breaking it by a tool (I didn't want to use soldering heat to remove the cap), the problem is fixed !
3- When i cut the route between the cap and input, the voltage on MCU input pin rises to 3.3V (through pull up resistor).
4- I measured capacitance of the cap by meter, it shows around 10nF. but when i use ohm meter on the cap (which is electronically isolated from other components) in the failed caps, it shows around 1K resistance !
Could be that the area around the capacitor pads, or the body of the capacitor itself, was dirty (old rosin, flux, etc), forming a path to ground, and simply reflowing the joints burned it away.

Case #1 is the most curious to me. I've seen this happen in a couple cases, and it turned out they were from families which are prone to solder leaching. Do you know the exact part number?
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top