DeltaV VIM2 vs EIOC Pros and Cons.

Advice and thoughts of the communitity would be welcome. The view of anybody who has used both would be appreciated.

I have personally only used the VIM2.


With the advent of the EIOC device from version 13.3.1, is it still advisable to specify new projects with the VIM2 card. 

Or should we be specifying the EIOC device?

What is the Emerson recommendation?

In particular:

    • do you know lifecycle on VIM2 ie is this is preferred route by Emerson. Will VIM2 be phased out?
    • How long will VIM2 be supported?
    • can the Ethernet io card support multiple simulataneous protocols or just one. In fact could the VIM2 card?

 Thanks,

Neil.

  • Hi Neil,
    Hope all is well with you.
    I would stick with VIM2, the cost of the EIOC is the biggest problem, more expensive than an SX controller.

    „„ You can choose between the 3 different Ethernet based
    protocols included in the EIOC. Each EIOC supports one
    protocol and once the protocol license is assigned, the
    protocol can be selected from the drop down list. Just
    select the one you want to use from the drop down menu.

    copied
  • Here's a suggested process to follow for using EIOC vs VIM2

    • IEC-61850 Required? Yes = EIOC
    • > 32 device/PLC nodes? Yes = EIOC
    • > 6000 Soft IO Tags? Yes = EIOC
    • EM's, SFC's or Batch Required? Yes = VIM2
    • PLC Migration w/ Heavy Control? Yes = VIM2
    • PLC Hardwired IO using Soft IO? Yes = VIM2
    • Use EIOC
  • Thanks Petroban and Matt

    I had performed a trial config of EIOC and I did like the ability to configure signals as per the profibus signal approach.

    I also like the single DV Explorer configuration of EIOC vs the dual VIMNET explorer and DV Explorer configuation required with VIM2 and serial card emulation (that is quite clunky, but appreciate the reason before Mynah were brought under Emerson ownership). Will the need for that ever be replaced with native DV Explorer config for VIM2? I’m thinking system lifecycle savings over 15 years, vs the double cost of EIOC (£4-5k more than VIM2)

    Unless I’m mistaken, the quoted UK cost is nearer SQ controller cost not SX which is approx 2 x SQ here in UK at the moment.


    Still a few other questions I posed not answered, longevity of VIM2 and can EIOC/VIM2 only do one Ethernet protocol per application?

    Thanks.

    Neil

  • In reply to Matt Stoner:

    Could you elaborate on why the VIM2 is preferred over EIOC for EM's/SFC's/Batch?
  • In reply to Chelsea Sanborn:

    Good question Chelsea.

    It does say this in the datasheet,

    But I don’t see why.

    The EIOC seems to be in the IO Network, does this therefore mean that it’s DSTs can be used in the same way as CIOC but other controllers (limit of 4 controllers per CIOC)

    If this were the case why not for EMs Batch and SFCs if the DST are referenced by modules in the controllers executing the EMs Batch SFCs.


    I realise VIM2 is close coupled to one controller as it emulates serial IO signals of that controller.


    If my assumption about EIOC DSTs being accessible to any controller as per CIOC the EIOC is a solution with much more flexibility even in the EMs SFC and Batch environment .

  • In reply to Chelsea Sanborn:

    Modules using EIOC I/O have to be assigned to execute in the EIOC and only allows certain function blocks which doesn't allow EMs, SFCs, and Phases/Units. This means that all the reads and writes will be cross controller writes if using the EIOC modules with EMs, SFCs and Phases/Units which will likely load up the network communications.

    You can find the supported function blocks that can execute in EIOC in Books Online under Configuration -> I/O Configuration -> The DeltaV I/O network -> DeltaV Ethernet I/O Card - Function blocks supported by the EIOC.
  • In reply to Matt Stoner:

    Hi Matt,

    I should have read the datasheet Petroban posted

    From that

    “Batch components such as SFCs and PLMs are not supported in the EIOC. Extensive external referencing to the EIOC for applications such as VFD control and Batch control is not advised. See the PDS for the VIM2 which supports
    those applications.”


    But how is CIOC comms any different to this.

    Are you saying that EIOC DSTs cannot be read/written directly from other controllers on the network.(Perhaps this is the key point). But obviously you could put intermediate mapping modules into the EIOC and take the external reference hit (perhaps not a good idea)

    Perhaps I’m beginning to realise that Controller to CIOC comms has a higher quality of service (QoS) than external reference comms?

  • Hi Neil,
    Chris Lloyd here forgot about petroban username.
    Vim2 now supports profinet and is S series hardware compatible.
    As far as I know S series is will have many years to come.
    1 protocol per EIOC allowed. Same as Vim2.
    Don’t know why they built it with all that functionality and can’t do SFC’s.
    I was told the book price was 15k dollars.
    Me I prefer Vim’s just don’t like the DST license costs. You get first 15 free just charged for the most expensive DST type.
    Mynah is now owned by Emerson.
    Thanks


    Sent from my iPhone
  • In reply to petroban:

    Oh Hi Chris, didn’t realise that was you.

    Thanks for the info.

    Don’t ask about Emerson pricing!
  • In reply to Matt Stoner:

    Thank you! That's a very good explanation. I appreciate all the discussion here, this has been very informative.
  • In reply to IntuitiveNeil:

    Yes you cannot reference IO directly on an EIOC, so it has to be a module executing in the EIOC reference the IO of the EIOC. Putting a mapping module in is an option but now your control will be slow...

    I'm pretty sure the communications for Reads and Writes of values to other controllers are handled much differently than the communications of the I/O values from CIOCs.

    You can find this in the Cross Controller Sends and Receives which have limits. Maybe can confirm if CIOC comms affects this or not.

    The parameters for Cross Controller Sends and Receives are located under the Communications in Diagnostics with descriptions Unsolicited parameters sent (Path:NodeName/ROC/UNSATTSNT) and Unsolicited parameters received (Path:NodeName/ROC/UNSATTRCV)
  • In reply to petroban:

    This is a very relevant question. And the landscape is moving.

    The EIOC was conceived to offer a natively configured Ethernet integration product with superior performance to the VIM/VIM2. In its v13 initial release, I can only say it was poorly positioned based on its capacity for 256 devices and compared to an SX with VIM solution. It was positioned to integrate any and all devices, including MCC, VFDs, PLC's, IEDs(with IEC61850), and if you managed to have 256 devices on an EIOC, it wasn't too exorbitant. But forcing customers to purchase 256 device capacity priced the node out of reach for most budgets.

    In v14, the EIOC is repositioned for subsystem integration and users can purchase the number of devices they plan to connect. So price is less of an issue, and would be less than an SX with VIM.

    The EIOC sits as an independent node on the DeltaV network, and counts as an IO Node, like a CIOC or WIOC, of which there is a limit of 300 per system. It does not count as one of the primary 120 nodes, like controllers and Workstations. However, it is not assignable to any controller, like the CIOC or WIOC. The EIOC natively runs the control modules that access the Ethernet signals in the configured logical devices. The EIOC does not support local DeltaV IO. When integrating motors, there may be an emergency stop hardwired to the DCS, or users may prefer to control the motor via hard IO and use Ethernet data for diagnostics and secondary data. In this case the EIOC alone does not work. Also, when motors and VFD's are integrated in the process control functions, interlocks and automated sequences may be sourced in the controller, and it is preferable to connect these devices to their controller via VIM.

    In v14.3, DeltaV also released the PK controllers that have an integrated Ethernet port for these devices. It runs the same configuration approach as the EIOC, but does not support the IEC61850 protocol or the OPC UA Client that the EIOC has.

    In v14.3, the EIOC was enhanced with the OPC UA protocol and also with the PRP (parallel Redundant protocol) so that it can connect to the IEC61850 network architecture. The PRP can be used with any protocol to create a high availability network with PRP aware switches or RED Boxes. The PK and VIM do not support the PRP, yet.

    For me, I position the EIOC for subsystem integration, where the subsystem is performing the control, like a PLC. Or for integrating IEC61850 or OPC UA subsystems. In v14.3, new engineering tools integrate AB PLC data structures that simplify PLC integration. These are only supported on the EIOC. So if there is a significant number of registers coming from PLC's, the EIOC would provide the best solution.

    The EIOC can process significantly more data than the VIM as the Ethernet data lands directly in the processor, with no need to move through the DeltaV IO bus used by the VIM. The PK also has better throughput than the VIM for the same reason.

    For M-series and S-Series IO, the VIM/VIM2 is required to integrate Ethernet Devices lime MCC's and VFD's to the controller. Adding the VIM2 to the existing controller is much more cost effective than using an EIOC for these devices. But remember that the control modules used for Ethernet IO devices use just as much controller CPU as physical IO. The capacity of an SX controller to handle its normal hard IO is limited, and adding thousands of soft IO to the same controller will impact CPU loading. Integrating a VFD means one module for all the data registers of that device. Integrating a PLC with 20 datasets and up to 2000 signals could add 500 or more control modules. Dedicating an SX controller with VIM for a group of PLC's would work, but I would move to the EIOC for such a situation, and the increased bandwidth.

    The PK controller in v14 changes the game, again. Although I still prefer to use an EIOC for subsystem integration, the PK offers the same performance as the EIOC and for applications that require no more than 100 or 300 "Logical Devices", the PK 100 or PK300 would be better value, as long as I don't need IEC61850, OPC UA client or the PLC Integration tools. The limits on SFCs and EMS and functions blocks are not there. PRP is not supported on the PK in v14.

    The way I see it, the VIM2 is for the M series and S Series controllers, but can also be added to a PK (if you needed both Modbus TCP and EthernetIP on one controller.) With v14 PK controllers the VIM will mostly not be needed.

    The EIOC is only available on v13 or later, with changes in licensing in v14. Both the EIOC and PK use a device license to enable the right number of devices. The PK controllers limit the number of devices possible based on their size: PK100=16 devices, PK300=32 devices, PK750=64 devices, PK1500=128 devices.

    If I had my druthers, I would clearly position the EIOC for Subsystems and enhance that product to serve that market. The PK controller would provide optional Ethernet IO integration for MCC's and VFD's for tight integration in the control strategies. The VIM provided these capabilities in pre v13/v14 systems. The VIM also supports a host of other protocols that can be integrated into the PK, M-series and S-series controllers as well. It continues to be developed to serve our installed base of M and S series systems.

    2 cents.

    Andre Dicaire

  • In reply to Andre Dicaire:

    In my opinion, a very important feature on EOIC not present on VIM is that EOIC allows to define DST's on data exchanged even into bit level, similar to deltav profibus. This is could remove the need to deploy intermediate encode/decode control modules (a.k.a landing modules) when discrete signals are packed into registers. Using this feature possible to use same discrete device classes for conventional and communicated device signals because DST's defined into EOIC may be consumed on function block IO_IN/IO_OUT parameters. Inmmediate benefits are less controller loading and smaller project library size.
  • In reply to gamella:

    All great replies Andre and Gamella.

    At the moment just investigating options for a client.

    One swaying factor is that my client has two VIM2 devices on site already (on two other systems). If they went EIOC on this new system they’d have to stock an EIOC spare so it would make it an expensive option.

    The EIOC is a very attractive solution for the reasons stated. And for another high integration application it would certainly be worth another visit.

    My main reason for the enquiry was to ensure I wasn’t leading my client down a path of obsolescence.

    Many thanks,

    Neil
  • I've recently discovered a few limitations of the EIOC with Modbus TCP/IP that would push me toward using a VIM for Modbus applications.
    1. No support for encapsulated serial Modbus (i.e. supports RTU TCP, not RTU via TCP or RTU via UDP), which the VIM has.
    2. No control over polling rate - it polls as fast as it possibly can. This may have changed in V14.
    3. No support for Modbus registers over 9999 (e.g. holding registers 410000-465536) which does not appear to meet the current Modbus protocol specification (V1.1b3). The VIM supports the full address range.