How long?

[mamluk 0]

How long does the large-type model (AnACPU, AnNCPU, AnUCPU) last? 15, 20, 30 years? i also noticed that A61P (power unit, mostly used in my dept) is among the components that love to fail.

[Tech Fred 3]

They have not been around that long. I have customers that are still using SquareD and TI PLC's fromt he 80's so it is feasible that any PLC fould last that long in a healthy elctronic environment. However, poor quality power and contounuous use of elctromechanical relays does take its toll on a device. I think that in a continuous production environment 8-10 years would be on the high end of the survivability curve. the other thing to consider is that PLC's are usually packaged with the hardware they control so replacing the hardware alse replaces the PLC.

[Rodney 21]

I have a number of A series P.L.C.s still in operation for more than twenty years. Many of which have had no part replacement in all that time including the power supplies. I have however had a couple of power supply failures on the Q series. Regards Rodney Edited 13 Jan 2009 by Rodney

[Crossbow 296]

A Series was launched in 1985. It's officially obsolete (the 10" models, not the 5" models) as of 2005/2007. A 20 year run for any computer is unheard of, and PLCs are nothing more than dedicated PCs. The big question to be asked is would you still use a PC you bought 20 years ago? Back in the days of the 5 1/4 floppy and DOS 5.0? Remeber EGA graphics? I doubt it. It's a PC, just with slightly more robust hardware, but then it is subjected to much more demanding environments in many cases. I know many installations of A Series PLC running for at least 18 years now (installed 1990). They are as robust as anything else out there. Of course years of abusing relays at their switching capacity and bad incoming power and brownouts can degrade the hardware faster. But I have installed another PLC from another vendor and lost a power supply the following week, and it was on an on-line UPS so the incoming power was clean.

[mamluk 0]

to summarize, the factors related to the life span of melsec-a & PLC in general, 1. total working hours (usage) 2. stability of incoming power 3. environment in my case, the oldest have been running for about 15 years..life span is one thing, but actually i'm bothered by the Obsolete Status for melsec-A declared by mitsu few years ago...anyway, thx for the replies..maybe i should proceed with the transition-to-Q-series plan.

[RTRBURN 0]

We have alot of old A series plus 10+ AOJ2's still running Medoc. I have never had a A series go bad...just the Memory Cassette or Battery and the A61 and A62 power supplies. We have upgraded some of the A series up to a Q series and used a MEAU backplane board that looks like a A series backplane board and connected the A series modules to them and they now run off the Q02HCPU. It worked out really well on the conversion.

[mamluk 0]

yeah we have like 100+ units of A61P at the back of 75+ m/c..so its quite natural they topped the list for replacement...after more than 15 years of service, running 2 shifts for the whole year with 1 week of shutdown & several holidays, they need upgrading.. Edited 14 Jan 2009 by mamluk

[andybr 5]

We are currently upgrading 20 year old A Series systems to use Control Logix. The driver for this is the lack of spare A Series processors as we have had units damaged (but no unprovoked failures). The ironic thing is that we have had more failures with CLX processors but we can, of course, get replacements for these.

[Gambit 326]

Why didn't you use new Q processors with the old modules then? Now you have to reprogram the software. Edited 14 Jan 2009 by Gambit

[andybr 5]

Gambit. The decision to move to CLX was due to corporate standardisation. We do have Q processors as spares but using these involves major software changes.

[panic mode 246]

standardization is fine but if we only look at software changes - i am not sure i follow... they are very minor when migrating from A to Q, while going to completely different brand means rewriting everything from scratch. if the power supplies keep failing, something is wrong and need to be addressed instead of just swapping out failed ones.

[andybr 5]

panic mode. I work for a large multinational company. Standardisation is not up for discussion, we just have to live with it. Software changes can be surprisingly involved when going from A to Q if there is extensive use of special devices. The real issue, however, is that large volumes of data (thousands of points per machine) need to be manually re-entered into file registers which are handled differently in the Q processors we have (Q3A). Since the installed software is poorly documented and written by a long since defunct OEM there are benefits to be had by re-writing from scratch.

[Crossbow 296]

Q3A is not a new Q processor. It's QnA. That's also obsolete. That's like replacing a 78 Ford truck with a 79 Ford truck. Q Series is a completely new and improved platform, and it does not require vast amounts of code changes to move from A to Q with the conversion support tools and manuals which Mitsubishi Electric has provided. Use of special registers has nothing to do with the conversion forward, as all the old addresses exist point for point in the new processors.

[Gambit 326]

You can read all devices with the software and then shot them into a QnA or Q. Unless you have used the rset instruction to create more than 8192 fileregisters.