ControlLogix Upgrade Questions

[escoplcguy]

Hello All!! this is my first post on the fourm so if there is any info you need or you think i left out just let me know...i am going to do my best to describe everything....

i am at a site that wants to upgrade their existing plc systems to 1 stand alone control setup. currently they have 4 SLC PLC's (2 5/03, 1 5/01, and as i call them the SLC BRICK) they have a PLC 5/40 and a Micrologix 1200. They want the upgrade to be "painless" (not moviing around alot of wiring) quick and cost eficient. Thinking hard about it and doing as much research i can on ControlLogix i have come up with putting 2 control logix 16mb processors in one chassis, the 1st processor will control the SLC 5/03, 5/01, and the PLC5 via remote i/o by swapping processors with rio modules and a 1756 dh+rio card in the CLX chassis. the other processor will be used to control I/O modules in the same chassis and the micrologix and SLC BRICK well be done away with.

with the analog inputs, the 1756 dh+rio card can recieve and transmit this data?? i assume the is straight forward but i seen a few posts saying it can only handle so much analog data before it maxes out.


also, same for the HMI there is a PV 1400e on DH+, a PV 600 and a PV+ 600... the PV+ can be see via ethernet i assume, but the others can be controled via the 1756 dhrio card??


thanks for your time to read all this!!


-anthony

ps...any ladder examples for setting up the dh+rio card would be amazing!!


i also have set up a IAB showing how all this would work...i can post if it would help you to understand more

 

[Paullys50]

Welcome to the forums!

A conversion isn't too bad because Rockwell does have tools to help convert PLC5 programs into ControlLogix programs. I have used the tools, it gets about 85% done w/o issue, but you have manual items you need to fix/re-program. Block Transfers don't exist in a CLX, they must be converted to a series of MSG instructions. We had an issue with ASCII code, due to CPU differences between a PLC5 and a CLX, required coding changes. If strings are used, you'll have to do some HMI work to make these visible. There are other minor items, tedious to fix, but overall converting is not to bad using their tools.

The 1756-DHRIO card does have limitations, and is not as capable of handling RIO load has a PLC 5 is. Read the manual before designing, it is very important you understand the limitations as you may need a 2nd, or 3rd card. Analog is done with MSG instrucsions instead of block transfers. Talk with your AB rep about demonstrating a MSG instruction configured for block transfers.

You have the ability to "Map" CLX tags to memory addressing like you would see in a SLC or PLC. This allows those existing panelview apps to be "unchanged" (granted, there is always a catch to this). Again see if your distributor can give directions/demos of this.

Given the number of PLCs you want to merge, using the Rockwell conversion tools to do this won't be that easy. I say this because B3:3/0 could exist in every PLC you are merging....certainly will cause problems anytime memory overlaps.

In your situation, it makes more sense to me to re-write the code, based on the ability of the CLX platform. It should be "painless" from a physical hardware stand point, but either case it won't be "painless" from a software point of view.

Certainly use your Rockwell distributor as a resource on this during the design phase. You'll have some unique items you'll need help resolving. We are a great forum here, but we don't always see the big picture.

 

[escoplcguy]

Thanks for the reply!

as far as programing goes, they basically want me to rerite ALL OF IT! so i am getting a scope of work for all the operations....some are so small! memory usage is terrible here, they have a 5/03 controlling about 7outs and 2inputs and 3 analog signals. so the reprograming isnt going to be such a task. but rewireing would be, because they want to set up a control room that is much further away from where some of the I/O is currently and this is where the CLX will be.

this is why hardware wise i wanted to make sure that how i was thinking to do this would work. the SLC processors i was going to link the chassis together so it will be one connection to the DHRIO card seeing 2 different SLC chassis.

also, what your telling me is the restrictions as far as messaging for analog inputs is limited by the card not the program?? so if i needed more msg instructions i add another card??

 

[escoplcguy]

in total between all the processors, there is a total of 30 analog signals, and 8 thermocouple signals. how many can each dhrio channel handle

 

[Ron Beaufort]

welcome to the forum ... sorry I'm short on time but here's another angle to think about ...

how FAST does the data have to move? ...

the truth is that although the I/O can be made to "talk" from the existing chassis to the ControlLogix processor, in most cases, the data will NOT move as fast through the Remote I/O link and the DHRIO module as fast as it formerly did through only the chassis backplanes ...

buzzwords for today: "potential bottleneck" ...

 

[escoplcguy]

how much slower are we talking.....

compare....

whats the speed of the plc-5 and the speed of the dhrio module....what are my other options that isnt going to cost the company a lot. i only have a budget of 50,000

 

[escoplcguy]

what about swapping the processors out and replace with control net cards, would that speed it up....can i use the existing chassis as remot I/O controled by the CLX with Cnet.

are there any restrictions on this card???

 

[OkiePC]

Another consideration when combining multiple controllers into one:

What happens to the plant if one PLC pukes as it is currently set up?

How is this going to change when you have the whole operation in one chassis and its power supply gives up the ghost?

You have probably thought about this, but if not, you certainly should.

 

[escoplcguy]

i havent thought about this and now i am going to have to give this project much more thought. very good point.

 

[rdrast]

Why two processors?
For that matter, why 16MB processors? Seriously, you would need a HUGE number of I/O and even a huger program to use up even one of those processors.

I'm looking online at an L62 processor, with somewhere about 3200 I/O points spread out over 18 racks and 20 drives, and still have 402,000 bytes of I/O memory free (out of 505,000).

The program has three tasks, 8 programs, and 110 routines, and has 2,952,000 bytes free out of 4,194,000.

 

[Ken Roach]

About speed:

The speed difference you need to be concerned about is the difference between backplane I/O and I/O that's on a network. Check all programs carefully for local interrupts (DII and STI), and in general think about how fast the processes you are controlling need to react to I/O signals.

Scanning RIO adapters with a 1756-RIO takes about 3 ms per adapter at 230.4 kb/s, and about 7 ms/adapter at 57.6 kb/s. Add a few milliseconds for the Block Transfers to run as well.

About RIO:

The full name of the protocol that runs on blue-jacketed twinaxial wire from PLC-5 controllers is "Allen-Bradley Universal Remote I/O". Everybody calls it "RIO", and whether that converts to "Radar Intercept Officer" or a Duran-Duran song depends on how old you are.

The 1771 and 1746 chassis can be scanned by a 1756-DHRIO or 1756-RIO module by using an ordinary 1771-ASB or 1746-ASB adapter.

ControlNet runs a lot faster and more deterministically than RIO. You can scan a 1771 chassis with ControlNet about as fast as you could over the PLC-5 backplane.

But there's no 1746 adapter for ControlNet. If you needed ControlNet for speed on the 1771 racks, you would have to change out the I/O at the 1746 racks and that might blow your budget.

About Block Transfers:

The limitation on legacy analog modules is a limitation of the ControlLogix itself, rather than the network modules. At the very most a ControlLogix can have 32 "cached" messages to analog modules (these replace the BTR and BTW instructions in the PLC5).

Count the actual number of analog modules; it sounds like it's only 6 or 7.

About conversion and consolidation:

It's harder than it looks. You're getting some good advice here. A handful of little things I've tripped over during PLC conversions:

Inaccurate wiring diagrams. Over a decade or two of operation, 10-20% of the I/O devices get replaced, rewired, or removed and the prints seldom get updated. Do your best to get a master set of prints and verify that the addressing in the existing programs matches what's on paper.

HMI overlap. This has already been mentioned and should be repeated: you may have to re-address one or more of the HMI/OIT devices because they may use the same addresses for different controllers. The ControlLogix can only emulate PLC/SLC addressing for N, B, and F data file types on DH+, so find out if your HMI devices are also addressing String, ASCII, Timer, Counter, or Status data tables.

Intercontroller Messaging. Be sure you understand what every single MSG instruction in each controller you're converting does. I've found messages that were abandoned but still running, and messages that included a watchdog payload that caused the triggering of another message in the target. On DH+, the PLC-5 and SLC-5/04 had a "Global Status Data" feature that the 1756-DHRIO doesn't support, so watch for that.

If there are HMI systems or a DCS that you're not converting that might touch these control systems, look into that communications. I've found a few systems where they didn't even know that a DCS was communicating with the PLC system until we took the PLC system offline and the DCS comms failure horns went off.

Scope of supply. Are you an in-house engineer who will have time to work on the items that grow in scope, or are you being asked to supply a "quick, easy, efficient" conversion at a fixed price ? If you go with a fixed budget and an open-ended scope, you're signing up for an educational experience. That can be fine if you are going to use this technology or skill later, but go in with your head up.

 

[escoplcguy]

@rdrast - i was only chosing that processor as part of the IAB portion of the project, is it overkill ABSOLUTELY!! most likely the lowest CLX processor will do exactually what i need. its just always better to start huge and take steps back...or so i've learned in the past...the worst is to start a project decide on a processor then find out there is more involved than the customer thought and now you need to buy a new processor because you have just maxed out its memory.

@ Ken - Thank you so very much for your input on this....i only started my project this way because i seen your post on RA forums to someone else about this matter. My experence is minamul with control logix, i have had only 6 years "4 of which were in school" and they were all with SLC & PLC-5 controllers and networks....

The processes arent anything highspeed, its cement mixing and dumping. as long as timer done bits and i/opointer arent skipped then a few ms i dont think are going to make a difference. most the outputs are drawn off of timers anyway. the HMI's do use strings and in the future they want to add scanners to scan where batches are in the system and what they should be..so i guess ASCII is in the near future as well, also we do use a DCS which tells the PLC-5 how much to mix per batch........

just to clarify everyting will be controlled via "univeral remote i/o" meaning the control logix processor tells the plc-5 chassis or the SLC chassis existing I/O what to do and when. DH+ would only be used for communication with a PV 1400e...but it seems like you are saying this may be impossible because of the commands that the 1400e sends to the PLC. But if i am not mistaking the 1400e also has a spot for RIO as well....could i send the strings and comands that way? or does the 1400e belong in the trash with the rest its family haha. I have already recreated all the screens via rsview32, it can control the PLC-5 from my desk via eithernet. would this be something more logical to look into implementing??

as far as scope, i am a new in-house engineer, i only been here for 8 months and this is one of the projects i recieved. being fresh from school, and other jobs i had after school were working with guys that have been in the field for years and years i am learning TONS!! i read alot and learn as much as i can. I love the world of logic controlling and the programing part i feel i am really great at it, its the networking part that can over take me at times. its a shame its not all just ehternet and rs-232 haha.

 

[escoplcguy]

our DCS is talking to our plc via data highway.

 

[JeffKiper]

If you are going to get into ASCII. You may want to plan for that ahead of time. Put an ENBT card in the chassis.

 

[The Plc Kid]

@ken

Ken could he se this http://samplecode.rockwellautomation.com/idc/groups/literature/documents/in/1747-in017_-en-p.pdf module to do controlnet on the slc I/O?

I saw a display for a AENT adaptor for slc at automation fair but i think it comes out Q3 or Q4 2011not sure.