Question about FIT figure ceramic capacitor real and MIL HNBK 217

[jeem]

Hello newsgroup readers,

We do Reliability MTBF calculations for our products. I have a specific
question regarding the FIT figures of ceramic capacitors.

In MIL HNBK 217 are given basic FIT figures for ech type of components also
for ceramic capacitors. But if we search for real figures presented by the
manufacturers of ceramic capacitor these real FIT figures are MUCH better
(so lower FIT figures).

How do you handle with this?
If you use the real FIT figures than you can not stated calculations made
according MIL HNBK 217F, but these calculate figures are much realistic than
with the data from the MIL HNBK 217.

Thanks for your reply's

Jan Mobers

 

[Tam/WB2TT]

Hello newsgroup readers,

We do Reliability MTBF calculations for our products. I have a specific
question regarding the FIT figures of ceramic capacitors.

In MIL HNBK 217 are given basic FIT figures for ech type of components also
for ceramic capacitors. But if we search for real figures presented by the
manufacturers of ceramic capacitor these real FIT figures are MUCH better
(so lower FIT figures).

How do you handle with this?
If you use the real FIT figures than you can not stated calculations made
according MIL HNBK 217F, but these calculate figures are much realistic than
with the data from the MIL HNBK 217.

Thanks for your reply's

Jan Mobers

What kind of numbers do you have? I think the last time I had to do this, we
used 5 - 10 FITS for resistors and ceramic caps. Kind of got lost in the
noise when you add up the real stuff.

Tam

 

[jeem]

What kind of numbers do you have? I think the last time I had to do this, we
used 5 - 10 FITS for resistors and ceramic caps. Kind of got lost in the
noise when you add up the real stuff.

Tam

I dont have the just figures with me at home, but with about 200 - 300 caps
( used as EMC caps, decoupling) its a huge difference if we must use for
instance 2 or can we use real figures of 0.4 FIT.

Jan Mobers

 

[Tam/WB2TT]

this,
I dont have the just figures with me at home, but with about 200 - 300 caps
( used as EMC caps, decoupling) its a huge difference if we must use for
instance 2 or can we use real figures of 0.4 FIT.

Jan Mobers

Jan,
What else is on the board? In my case there were not quite 100 caps, but the
board had three Motorola 68340 CPUs, 6 TI DSPs, an 8051, 10 PALS, two DS1
type interfaces, a 32 bit system interface, and the usual collection of
74ACT245 and 74ACT540 type stuff. Like I said, the caps got lost in the
noise.

You might want to worry about how the caps fail. If 50% of them open up, the
board will probably still work. If 1 shorted, it would not work. Also, the
two sets of FIT numbers might be at different temperatures. I suspect the
amount of AC current through them would also make a difference, as would
pushing the DC voltage rating, or continuous vibration. I have seen leads
pull out of radial mount caps; so, surface mount might be better.

Tam

 

[jeem]

Jan,
What else is on the board? In my case there were not quite 100 caps, but the
board had three Motorola 68340 CPUs, 6 TI DSPs, an 8051, 10 PALS, two DS1
type interfaces, a 32 bit system interface, and the usual collection of
74ACT245 and 74ACT540 type stuff. Like I said, the caps got lost in the
noise.

You might want to worry about how the caps fail. If 50% of them open up, the
board will probably still work. If 1 shorted, it would not work. Also, the
two sets of FIT numbers might be at different temperatures. I suspect the
amount of AC current through them would also make a difference, as would
pushing the DC voltage rating, or continuous vibration. I have seen leads
pull out of radial mount caps; so, surface mount might be better.

Tam

There are a lot of components more on it, nearly as you described above. But
we found by calculations that the ceramic cap's present nearly 20% of the
total failure rate. So its a figure to examine.
We use already for all components surface mount components.

Q
Is it allowed to skip in the calculations those caps, if they become open
during fault condition?
From literature we found that the most common fault condition of a ceramic
capacitor is a short circuit situation.

Jan

 

[Roy McCammon]

Hello newsgroup readers,

We do Reliability MTBF calculations for our products. I have a specific
question regarding the FIT figures of ceramic capacitors.

In MIL HNBK 217 are given basic FIT figures for ech type of components also
for ceramic capacitors. But if we search for real figures presented by the
manufacturers of ceramic capacitor these real FIT figures are MUCH better
(so lower FIT figures).

How do you handle with this?
If you use the real FIT figures than you can not stated calculations made
according MIL HNBK 217F, but these calculate figures are much realistic than
with the data from the MIL HNBK 217.

I used have a copy. Long gone. But my recollection is that somewhere
in it it said that if you have actual data, then use the actual data.