GE 9030 losing %m status?

I have a GE 9030, and I am using some "%m" bits to set an "out of service" condition. These bits are set ONLY by an HMI. The HMI requires a login and password, and only myself and my opposite know this.

We shut in for the hurricane, and now the "%m" bits I used are cleared. Note that the PLC power was disconnected, and the PLC was OFF, as in dead-lights. The result was, all except one of these "out of service" bits were cleared.

As far as I know, if you set a bit, it stays set, right? Is there something in the GE PLCs that clears these bits? I'm really confused. So far I haven't found anything in the documentation.

Variables assigned to %M addresses in the Series 90 can be either retentive or non-retentive. If they are retentive, they will resume their previous state following a power interruption. If non-retentive, they will come up in the off state following power interruption unless the logic turns them on.

In programs created with Logicmaster or Proficy Machine Edition software, all %M addresses default to being non-retentive and the programmer must declare them to be retentive.

In programs created with VersaPro, all %M addresses default to being retentive.

AH!

...I should have known that...

I'm using Proficy. Let me check... yep. This program was a conversion from 9070, I don't think I set any retentive status anywhere. Thanks. This is a first.

Steve Bailey said:

In programs created with Logicmaster or Proficy Machine Edition software, all %M addresses default to being non-retentive and the programmer must declare them to be retentive.

In programs created with VersaPro, all %M addresses default to being retentive.

Click to expand...

I didn't know that. Noted for the future, I'm sure it would have tripped me sooner or later. This board is awesome.

Yeah... in the AB world, everything stays set. This is an option that really does have some use. I found the one "out of service" bit that was good WAS set for true; it was done before the out of service screen was finished, go figure.

My advantage is that the regular bypass bits allowed them to operate. Setting these in the GE is still awesome, it rolled the changes right in. This was changed from a LM90 9070 program to Versapro, then Proficy 9030 over the last few years, with only a few hiccups that the 9070 had that weren't supported by the 9030.

Got it fixed, though. Sneaky GE...

Just to continue the discussion of retentive vs non-retentive, you have the same choices for %Q addresses, but not for %G or %T addresses. %G addresses are all retentive and cannot be declared non-retentive, %T addresses are al non-retentive and cannot be declared retentive.

When you program a coil or a latch instruction with a retentive address, the instruction will show up in logic as -(M)- or -(SM)-
If you use non-retentive addresses they will show as -( )- or -(S)-

The retentive nature is not limited to latch coils. The typical holding circuit

Start   Stop          Run

-] [--+--] [----------(M)-

      |

 Run  |

-] [--+

will show the same behavior when you use a retentive address.

I could never get anyone at GE Fanuc to publicly (privately was a different story) agree that the VersaPro default choice of retentive for %M And %Q adresses was a poor choice, but they reverted to non-retentive default in PME.

I've seen the retentive coils. I have actually had Proficy select retentive, even when I specified non-retentive. It has some quirks regarding how it sets up retentive values.

AB should take note of this; while it isn't something I would normally use, in my case it prevents a trip just to change the status bits.