MTBF for AB SLC500

Does anybody have the MTBF figure for the AB SLC500 PLC?

AC

Techsupport or local AB office will give you this number if you know exact part number for SLC.
Each controller as well as each module has own MTBF value.

What does MTBF stand for? Just wondering.

Patrick

Mean Time Between Failures

In other word, how long it should last on average...

The PLC itself should last longer than the job it's meant to do, given the rate that most industries change. I've never heard of any of the processors I've dealt with having problems (Other than a person smashing something into a cabinet, like a Forklift, and actually damaging it...) I don't believe the MTBF should be an issue at all.

I'd say a properly installed and enclosed SLC500 would last 10 or more years. Of course, I'm no where near an official word. Any MTBF over 100,000 hours is kinda over kill if you think about it... 10 years of the same operation, you'd think you'd upgrade a bit or something, make things more efficient. You dont wanna stand still

I see things with 1 million hours, is over 100 years really necessary?

For SLC 5/05 it is about 800,000 hours or 90 years
for older SLCs it's 8-10 times higher, in millions of hours.
Again get exact numbers from techsupport or local RA engineer.
All MTBF numbers are available.

Interesting

We're an AB plant with about 140 processors....I have never seen a 5/04 or older fail. We have had several newer 5/05's fail. The Ethernet daughter card has failed on 6 or 7 in the last two years. They are all properly grounded and behind N-TRON industrial switches. We haven't been able to get any feedback from AB, even when they do a failure analysis on them as to the root cause. They have all been under warranty and they have always exchanged them for a new one....anyone else have this issue?

Greg

Contr_Conn said:

For SLC 5/05 it is about 800,000 hours or 90 years
for older SLCs it's 8-10 times higher, in millions of hours.
Again get exact numbers from techsupport or local RA engineer.
All MTBF numbers are available.

Click to expand...

How do you do the testing to say that MTBF is 90 years? They haven't been around any where close to that long - so how do you come up with that number? I've never worked for a company that has to produce these kinds of statistics, so I'm genuinely interested in knowing how this number was formed. Whenever I'm asked i give a response similar to "If it works for the first three months, you'll probably never see it fail".

MTBF has nothing to do with testing, it is just a pure statistics and probablity
It's calculated based on the amount of devices shipped and failed within specific time interval.

http://en.wikipedia.org/wiki/Mean_time_between_failure

so how do you come up with that number?

Click to expand...

I looked at the actual numbers reported by RA.

Greg Dake said:

We're an AB plant with about 140 processors....I have never seen a 5/04 or older fail. We have had several newer 5/05's fail. The Ethernet daughter card has failed on 6 or 7 in the last two years. They are all properly grounded and behind N-TRON industrial switches. We haven't been able to get any feedback from AB, even when they do a failure analysis on them as to the root cause. They have all been under warranty and they have always exchanged them for a new one....anyone else have this issue?

Greg

Click to expand...

I worked in a 95% AB plant with dozens of SLC 5/03 and 5/04, only one or two 5/05s, about 75-100 PLC5s, lots of 1394 motion controllers, and much much more.

I found that the PLC5 was indestructible, we even tried to drown a few of them (roof leaks) under power in panels full of 3 phase drives, but only succeeded in wiping out the program and burning up a few power supplies. Every PLC5 processor we suspected of failure was easily revived in the office with a known good rack and power supply. Even the I/O cards would tolerate severe abuse and often survive somehow.

The SLC obviously was not nearly as rugged. I have seen about 10 SLC5/04 failures, only one 5/03 failure, and we didn't have enough 5/05s to be fair. The SLC failures were uncommon and weird each time. I had one that would put itself in Program mode spontaneously. I had several that lost their programs inexplicably. I had a couple of them show zero sign of life at all. I even had one that had an LED pattern not described in the book. I never once had a SLC try to run a corrupt program or fail to detect it's own fault. Not as rugged, but still very safe if properly commissioned.

My experience supports what was said about the older stuff being more robust. Kinda like my 350 pound 1972 Maytag washing machine...they don't make 'em like that anymore!