New PLC old HMI, PLC5 to CompactLogix Transition

I'm bidding on a project to upgrade a machine that is using a PLC5/30 and a PC based user interface that is running on Windows 3.1. The connection is serial. There is a second PC connected to the first one (also over serial) that controls a bunch of simple proprietary drives used to setup the machine for each new product run. The user interface PC sets the recipe information in both the PLC and the setup PC.

So here's the dilemma. The customer would prefer to upgrade the machine in pieces, and is starting with the PLC. I'm planning to upgrade to a CompactLogix. Assuming I have no ability to modify the PC, is there a way to set up the CompactLogix to emulate the PLC5 communication so that I can simply swap it out? I am aware of the PLC5 mapping option in Logix 5000, which seems to be the way to go. I even spoke with Rockwell, and they said that it might work, but couldn't give me any assurances.

Has anyone out there done something like this?

Thats probably your best bet. did similar but with a SLC and controllogix (with old SCADA)

So you used the mapping function to talk to an old third party SCADA? Not an exact match to my application, but pretty close! That makes me feel a little more confident.

yes thats exactly what it was used for. It worked good. There were no issues. eventually they budgeted a new SCADA and were much happier. It got them through the in-between period

Excellent, that's good news.

Thanks,

Just out of curiosity why the change to the compactlogix? AB has a conversion solution from PLC5 to ControlLogix that utilizes the old PLC5 wiring arms and uses the same footprint as the existing rack. The software is in theory a direct conversion process, but timers and messages need to be checked I believe. We are looking at a similar situation where I work and that is the direction we are looking at going.

The first question is the physical layer. The modern CompactLogix and ControlLogix don't have RS-232 serial ports, so to do this directly with a CompactLogix controller you would need to get a 1769-L32E or -L35E, so firmware support would stop at v20. The alternative is a networked gateway device like those from Prosoft, Real-Time Automation, or HMS Fieldbus.

The second question is which datatypes the PC is using to communicate with the PLC-5.

The "PLC/SLC Mapping" feature in Logix 5000 supports 16-bit Integers (B and N files) as well as 32-bit Floating Point (F files) data arrays. It also supports the old flat-file PLC-2 data access commands.

But it doesn't support any other structures, including Strings and Timers and Counters.

If you have some access to the drivers and datatypes of that PC-based HMI, check them carefully. Maybe the PLC program includes some documentation on the data files that are used for HMI interface.

The only way to know for certain what's being communicated between the systems is intercept and decode the serial traffic. That can be a difficult job; I use Frontline Test Equipment NetDecoder for those kind of projects.

califflash said:

Just out of curiosity why the change to the compactlogix? AB has a conversion solution from PLC5 to ControlLogix that utilizes the old PLC5 wiring arms and uses the same footprint as the existing rack. The software is in theory a direct conversion process, but timers and messages need to be checked I believe. We are looking at a similar situation where I work and that is the direction we are looking at going.

Click to expand...

In a word: cost. The vast majority of the I/O are on remote SLC500 racks with RIO adapters. The remote racks are staying with new Ethernet/IP adapters, so only rewiring will be on the base PLC5 rack, which is only a handful of cards, which I can do myself. Not counting the Ethernet/IP adapters for the remote racks, the price of the ControlLogix rack is twice that of the CompactLogix rack, without even including the conversion rack. So it just didn't make sense.

If I were doing an upgrade that needed the ControlLogix (large I/O count, lots of comms), had a lot of local I/O to rewire, or needed an extremely short downtime, I would absolutely use the conversion solution.

Ken Roach said:

The first question is the physical layer. The modern CompactLogix and ControlLogix don't have RS-232 serial ports, so to do this directly with a CompactLogix controller you would need to get a 1769-L32E or -L35E, so firmware support would stop at v20. The alternative is a networked gateway device like those from Prosoft, Real-Time Automation, or HMS Fieldbus.

The second question is which datatypes the PC is using to communicate with the PLC-5.

The "PLC/SLC Mapping" feature in Logix 5000 supports 16-bit Integers (B and N files) as well as 32-bit Floating Point (F files) data arrays. It also supports the old flat-file PLC-2 data access commands.

But it doesn't support any other structures, including Strings and Timers and Counters.

If you have some access to the drivers and datatypes of that PC-based HMI, check them carefully. Maybe the PLC program includes some documentation on the data files that are used for HMI interface.

The only way to know for certain what's being communicated between the systems is intercept and decode the serial traffic. That can be a difficult job; I use Frontline Test Equipment NetDecoder for those kind of projects.

Click to expand...

Thanks for the info. I'm going onsite in about a week to do some testing, assuming I can beg, borrow, or steal an L32E or L35E. I'll get a pretty good look at the system then.

I've worked on a piece of similar machinery from the same OEM, and as I recall, the HMI is a proprietary program, so I don't think I'll be able to pull the data types from the project, and I doubt that info would be part of the manual. So it will be trial and error!

This is why I pushed the customer to let me do the testing now, before starting the project, rather than wait and find out that we need a huge change-order to make things work.

Rockwell doesn't make these upgrades easy or inexpensive. SoftPLC does - you can leave all the racks & I/O in place, convert the logic & documentation (most times 95% or better), and the HMI won't even know you made a change - then you can upgrade it later to any HMI on the market. The SoftPLC can provide DF1 serial, DH+ or Ethernet comms out of the box. Total cost under $2000 for everything you need. check it out at http://softplc.com/products/migrations/

Cody.
You must be in sales!

I have done some similar upgrades, though not on AB.
My experience is that possibly you should upgrade BOTH at the same time, PLC as well as HMI.

If you reuse the HMI unchanged, it goes without saying that the code will be reused unchanged. This can be OK, but can also be a straight-jacket.
The memory concept of Logix is so different to the old PLC5, that it affects the code. You will at least partially miss out on some of the advantages of Logix.

In your case you are also being affected on the hardware (L32E or L35E, rather than just the CPU that would be the most optimal).

If you put in a legacy CPU, and reuse the old PLC code, they have in reality only gotten temporary relief rather than a real upgrade.

One potential advantage of reusing the HMI unchanged could be that you can switch back to the old CPU. Probably not really needed, but some customers like to have that as a safety, just in case.
Since you have to rewire the CPU rack, you will also miss out this quick change back to the old CPU.
So you are missing one of the advantages of keeping the old HMI, and therefore the reason to do so.

You write that the customer wants to upgrade in pieces. Possibly if you present your case well enough, you can win him over to do a complete switch-over in one go.

If you do a full change over of PLC and HMI in one go, you should meticulously test your programs in advance by simulation.

JRoss said:

I've worked on a piece of similar machinery from the same OEM, and as I recall, the HMI is a proprietary program, so I don't think I'll be able to pull the data types from the project, and I doubt that info would be part of the manual. So it will be trial and error!

Click to expand...

This is an argument for ditching the HMI now, not for keeping it.
If you keep the HMI, it locks your hands for everything else. And if you need to change or add just a small thing, you are stuck.
To change the HMI without the source will mean you have to reverse-engineer it. That is more work, but far better in the long run.

edit: You dont have to reverse-engineer the new HMI to be a 1-to-1 copy of the old HMI.
You can ask the customer "with the experience from the old HMI, how do you want the new HMI to work ?".
It can be a good argument for doing the complete switch-over that not only is hardware and software new and improved, but there can be better functionality of the machine after the upgrade.

PLC Pie Guy said:

Cody.
You must be in sales!

Click to expand...

Nope, we just do a lot of A-B PLC upgrades using SoftPLC and it is a really nice solution. In and out of the job in less than a day, happy customers. If the machine just needs to do what it did before with supported, non-obsolete equipment/software and a path for the future there isn't a reason to re-engineer anything.

When major upgrades are desired, SoftPLC gives us a platform for lower cost with more capability than Logix.

So in all cases, it is our preferred solution.

Sorry for the delay, been busy on a startup.

If it were as simple as just changing an HMI, I'd push the customer to do it now. Unfortunately, it isn't that simple. The PC that runs the HMI also controls parameters in a second PC that controls about 20 inverter drives over some kind of serial network. The drives are rack-based, made by Aros, and they and network are proprietary to the OEM. That's what I'm talking about reverse-engineering.

So, if I have to upgrade the HMI, I also have to move the drive control to the PLC and replace all 20 proprietary drives with standard VFDs, including adding a whole new control enclosure since the old drives are extremely small.

This is all something the customer wants to do eventually, but if we can avoid doing everything at once, that would be preferred, because of the cost and downtime.

It is possible to setup a DF1 connection with a CompactLogix with the same memory mapping as the PLC5, and it will be identical as far as the communication to the proprietary PC is concerned.

But, even if it is just the part between the HMI and the PLC, I would dread the reverse-engineeering part, without source code to the propietary PC program. It will be tedious work, and there will be a risk that you dont get it 100% right.