MTTF Datasheet Value

Hello,
I was looking at a datasheet for a sensor thats going to be used in a safety related application, I was interested in a B10 or B10D value, what I found was an MTTF value in years.

My understanding was that an MTTF value was based on the B10 value divided by the anticipated number of operations when installed.

How can the manufacturer provide an MTTF value in years if they don't know what my intended frequency of operation is going to be?

Any thoughts?




Kind Regards
UKB

It's been a while since I did anything safety related but from memory MTTF is based on (Time Period * Items Tested) / Items that failed.

So if you have 2142 Sensors and ran them all for a year, then 1 would fail.

If you look at the IFM website there is no B10d value as the unit is not relay switching, that's how I read it anyway.

Go ti the more information tab and the MTTF details are there

http://www.ifm.com/products/us/ds/MK502A.htm

Edit: Direct link to certificate http://www.ifm.com/files/mttf/MK502A.pdf

If it is a solid state sensor (in other words, if it has OSSD outputs), the number of operations, and therefore the B10/B10d, is irrelevant. B10/B10d is a measure of mechanical switching lifetime and is measured in number of operations. Solid state devices do not have a mechanical lifetime. Therefore solid state device MTTFd is fine to use directly in safety calculations. By comparison, something like a tongue interlock switch with dry contacts would require factoring in the anticipated number of operations and B10d to estimate the MTTFd.

Thanks I understand the part about b10 pertaining to mechanical parts wearing,
Im still a bit unsure about the MTTF being given in a time base, rather than number of cycles...
If the component in my installation is getting switched more frequently than the manufacturers test procedure I would expect the component in my installation to fail sooner, thus making the MTTF value inaccurate.
An average number of cycles to failure would seem more appropriate?

thanks Janner_10
that makes sense

It doesn't really matter how many times they test the transistor output compared to how many times you use it, transistor wear for the application you are using it will be negligible, and probably difficult to measure. The most likely cause for your failure would be some kind of voltage disturbance, which obviously isn't covered by wear.

If your switching cycles were in the MHz range then you could argue that you might need to take wear into account, but I'm guessing yours is a fairly slow switching process.

Jonnie_R said:

It doesn't really matter how many times they test the transistor output compared to how many times you use it, transistor wear for the application you are using it will be negligible, and probably difficult to measure. The most likely cause for your failure would be some kind of voltage disturbance, which obviously isn't covered by wear.

If your switching cycles were in the MHz range then you could argue that you might need to take wear into account, but I'm guessing yours is a fairly slow switching process.

Click to expand...

MTTFd is statistical data.

Jonnie_R said:

It doesn't really matter how many times they test the transistor output compared to how many times you use it, transistor wear for the application you are using it will be negligible, and probably difficult to measure. The most likely cause for your failure would be some kind of voltage disturbance, which obviously isn't covered by wear.

If your switching cycles were in the MHz range then you could argue that you might need to take wear into account, but I'm guessing yours is a fairly slow switching process.

Click to expand...

For any application, and especially for safety operations, you must also follow the manufacturers limits for usage such ast frequency of switching, voltage supply range, load switched, protection, etc. if you don’t do that, the B10 or MTTFD would be meaningless.

Okay,



so this is from the same datasheet.

So did they arrive at MTTF of 2142 years by taking 2142 units, switching them at 2000Hz and then after a year 1 failed, so they say the MTTF is 2142 years? seems a bit too easy...

UKB said:

...so this is from the same datasheet.

So did they arrive at MTTF of 2142 years by taking 2142 units, switching them at 2000Hz and then after a year 1 failed, so they say the MTTF is 2142 years? seems a bit too easy...

Click to expand...

The short answer is no. MTTFd for solid state safety devices is typically a complex engineering calculation factoring in specifications of all the components in the device. (Similar to how a safety system calculation factors in the MTTFd of all the safety products in that system.) MTTFd for solid state devices is NOT determined through product testing. (For comparison, electro-mechanical safety devices, such as tongue interlocks and traditional contactors, ARE specified through the use of mechanical testing.)

(EDIT: There certainly is product testing done by the vendor on any safety product and the components in it, I just want to emphasize that this is not the primary factor in how vendors specify MTTFd for solid state safety products.)

Whichever third party certification agency certified the solid state safety product (typically one of the TUV organizations) would have verified the MTTFd calculation as a part of the certification process.

ndzied1 said:

For any application, and especially for safety operations, you must also follow the manufacturers limits for usage such ast frequency of switching, voltage supply range, load switched, protection, etc. if you don’t do that, the B10 or MTTFD would be meaningless.

Click to expand...

I never stated you shouldn't merely explained the transistor suffers from negligible wear unlike a relay contact.