[SOLVED] Parasitic DC series resistance of 10uF SMD capacitors (0402 and 0603)

Hi guys,
we use an array of 20 linear voltage regulators (LDOs) with 5.5 Volts at the input. Each LDO has a 10uF SMD capacitor to ground in 0402 or 0602 size.
After assemby, the DC input resistance of the array, measured with standard ohm-meter, starts at 300kOhms and reaches 600kOhms or more after seconds (caps charging!).
The board did well for two days, now the DC input resistance of the array is 10 Ohms.
Any ideas? Could it be degradation of the capacitors?
Thanks you, best regards
-e

If the caps now really have that much leakage, then when applying 5.5 volts, they would draw over 1/2 amp and the magic smoke would come out.

But, to your question: MLCC degradation is very real, very often caused by flex cracks.

A google search of "capacitor flex cracks" yields over 250k hits. There is a lot of good info there, way too much to summarize in a post. Google some. I like in particular Kemet's app notes.

How do you measure capacitor "series resistance" with the capacitors and LDO in place? A leakage resistance (if actually existing) would be described as resistor in parallel to the capacitor, by the way.

Hello again,

thank you for your replies.
This seems to be a big problem for many of my modules, hence I would like to share more data.

@schmitt trigger:
Due to the current compliance of the HMP4040 power supply it cannot draw 0.5A.

@FvM:
Yes, we measure both, the LDO and cap. And yes, a resistor in parallel to the cap is what I meant.

--
Anyway, I am not sure if the caps are the problem, because the same effect did also happen with different LDOs (LDK130M-3) that use
different/smaller caps (1uF) at the input and output.

Attached you can find a description of the module. It seems, that some of the LDOs break after some hours of operation.
So, I am looking for reasons. I also thought about kind of a chain reaction. If one LDO breaks, it kills the others ?
The datasheets of both LDOs (TPS7A7001 and LDK130M-R) deal with stability issues.
I am very thankful for any hint/idea/thought/...

Thanky very much.
-e

- - - Updated - - -

View attachment MIMI_CTRL_LDO_problems.pdf

Ok. I think we wont solve this problem here.
I will try it with a new post and correct headline!

I will try it with a new post and correct headline!

Click to expand...

If your new post is referring to the same problem (unknown reason for LDO damage) it should be appended to this thread. A moderator can rename the thread on request. You might propose a more appropriate title in the post title.

Hello FvM, ok. I think this type of problem is difficult to solve with forums. It was worth trying ... Lets mark it as solved and close it here. -e.

Ok. For the sense of completeness, we think the reason for LDO failure is reverse current flow. When turning off the input voltage, the output capacitor is fully charged and forces a reverse current. Vout < Vin.
We will include Schottky diodes in our new design. As it is shown here: https://www.e-devices.ricoh.co.jp/en/products/power/ap_note/ldo/reverse.html