What happens if the PLC is turned off?

Ive heard people in my field say not to do a cold restart on PLCs as it could lose its memory/program. Is this true?

As long as the battery is good, or the charged capacitor that holds the RAM, cycling power to a PLC is not a problem.

However, I do not recommend powering the PLC with the K1/MCR power that shuts off every time the machine is stopped or an E-stop is pressed. I always found that a hindrance working on machinery - having to wait for the PLC to boot back up, and losing comm's when online and the machine was stopped. I have rewired quite a few PLC's from MCR/K1 power to main power.

Seems to be that many companies in the States wire the PLC power via the Estop circuit, We had a number of machines from the states, all were wired like this, all AB hardware and had many failures of PLC's, Inverters & HMI's, I re-wired all PLC's & HMI's to normal supplies, fed the outputs from the safety relay and did not have that many failures, Inverters were different, had to remove power to drives, replaced them with Mitsubishi & never had any more failures apart from some numpty leaving the panel open & hygiene washing it.
One thing to bear in mind is that some PLC's have retentive & non retentive memory areas so if you need to keep last position before power loss then use retentive memory, however, some bits may need to be reset on power loss.

parky said:

One thing to bear in mind is that some PLC's have retentive & non retentive memory areas so if you need to keep last position before power loss then use retentive memory, however, some bits may need to be reset on power loss.

Click to expand...

One thing related to that is incremental encoders to HSCE inputs. Cycle the power and the drive(s) need homed every time.

When I did my first Micrologix 1000 PLC, that's what I did.
E-Stop mean't kill all power to everything: Outputs, Solenoids PLC, the whole lot.

It's kind of embarrassing now when I look at it.
Some of these machines have been running and stopping this way for decades without doing any harm to the little Micrologix PLC's.

I wouldn't ever do this on a new setup though.
I think that I've learned that you can make an intrinsically safe machine without killing power to the PLC.

I wasn't particularly talking about position on servos but yes your right what I meant was really steps so that it will re-start (yes I mean on a resume by human before posters start going on about safety) from the last position before power loss. It depends on the process & safety some processes may need to be homed others not. Many PLC's have initial power up (or Stop to Run) or power loss (or run to stop) configurable routines which can be called to do some processing. A system I did some years ago copied non retentive bits into retentive words on such event & returned them on re-start included in the power loss and run to stop was setting a hold bit that prevented any operation until resumed by the operator.

parky said:

... some numpty leaving the panel open & hygiene washing it...

Click to expand...

ROFL, this is why the common language that separates us is called English - you own it.

Gbradly: yes the ML's seem to be bullet proof as was the SLC and prior but the later hardware is **** when it comes to power ups, in every case on panelview + it was the PSU board that failed on power down & up. Got so sick of AB failures that we made a decision to replace them over time with Beijers, Mitsi drives & PLC's, there was another reason RW's strong arm tactics regarding supply sources & tech connect contracts sorry my rant for the day.

If you have output cards, you can wire the power to those output cards using e-stop and safety relays, especially if you have a card or two dedicated to motion power. This way, PLC always remains on, but outputs don't function.

Already mentioned in post #3

MattMatt9,

ANYTIME you get a plc into your plant, you need to do the following.

1. get with those installing the system and learn all you can about the system.
2. ask about the plc and how it is programmed.
3. GET A COPY OF THE program on disk, cd, jump drive... and open it to verify its there and is the actual program.
4. before the guys leave, have them back up the software (hmi, scada,plc) and save it to your jump drive and verify its there. then go open the program to verify its correct. you may not have the latest version of software.
5. get the machine drawings, electrical, pneumatic, hydraulic prints, and any other prints, manuals, instructions....
6. if they refuse, get with the project manager and let him know what is going on.

now to your question.
before you turn off the power,
1. use your pc and connect to the plc using the software.
2. copy the program to a safe location as insurance
3. open the program, go online, and upload the program to the pc. this program is the master copy, regardless of what your pc files are, this is the working copy and therefore it is the master copy.
4. save it to the pc and compare the files, changes may have been made without anyone knowing it.
5. it is now safe to turn the plc off, you have a program copy.
the reason for this is because even though you have a good battery and know the plc is in good condition, things can happen. this is from personal experience. we have the master copy on hand on all plc's and our backup software does a routine backup of all plc files.
james

I cringe 100% of the time a power cycle happens.

Can it loose its program? Yes.
Can a module fail? Yes.
Can it load from flash or CF card and undo critical changes that weren’t saved? Yes.
Can a FS instruction overwrite values? Yes.

Lots of bad stuff can happen no matter how prepared you are. “It shouldn’t have done that”, is never an acceptable excuse.

If you are going to kill PLC power with an E-Stop you had better make sure you keep a fresh, charged battery in place if that's how your PLC retains its memory.
If I were in charge of accepting a new machine, I would insist the startup person from the OEM cycle PLC power and demonstrate that the machine could be restarted without any problems and without having to connect the PLC programming software.
Then I would have them remove any batteries, wipe the PLC memory, restore it from a backup and demonstrate the ability to restart the machine. If there were any parameters that had to be initialized, the procedure for initializing them would have to be documented.
I used to joke about invoking my "taillight warranty" when I left the plant. As long as you can see my taillights it's under warranty.

Steve Bailey said:

I used to joke about invoking my "taillight warranty" when I left the plant. As long as you can see my taillights it's under warranty.

Click to expand...

That's a longer warranty than I have given on some bodge-jobs I have had to work on.


Mine would have been more like "shoelaces" not "taillights"

I've seen PLCs have their memory scrambled by rapidly switching power on and off. The first time I witnessed this, it was caused by someone flipping circuit breakers in the panel that fed power to the PLC in an attempt to find which breaker powered a particular lighting circuit. It can also happen after a power outage, when power is momentarily restored only to drop out almost immediately afterward.

It used to be such a common problem with old Westinghouse NumaLogic PLCs that we would install a hardware timer in our PLC panels to only apply power to the PLC after 10 seconds without a blink in the source voltage upon power up.