From Rosemount 3420 FIM to Softing's FG-200 HSE/FF Modbus

[AlfredoQuintero]

Hi, a customer has an installation on which a Yamatake DCS monitors about 60 Rosemount FF temperature transmitters. The DCS is a Mobdus RTU master and in between the FF field devices and the DCS the customer has a Rousemount 3420 FIM (Fieldbus Interface Module).

The 3420 FIM has been discontinued years ago and the customer is thinking of a replacement, either with Softing's FG-200, or with a Yokogawa Stardom system (which seems to be kind of overkill, I mean using a Stardom DCS as a Modbus to FF gateway to yet another DCS).

In terms of cost and space, the FG-200 seems to be the right choice, but the customer's 3420 FIM's project was poorly documented by the Emerson system integrator, and the end-user has no control engineer with FF knowlegde. For them it is basically a black box. Hence they expect my company to do the engineering as well, not just supplying the FG-200.

I am familiar with FF technology, the FG-200 and the configuration tool for FG-200, but never have setup an actual system, just demos and simple customer trainings. The actual system is only monitoring temperatures and no valve positioner or any actuator device is being used. So, fortunately, for the commissioning it would be acceptable to have some communication interruptions, but here is the doubt I have, and would be grateful if someone familiar with the 3420 FIM could provide some advice.

The 3420 FIM is the LAS (link active scheduler) and I do not understand very well the rest of the configuration. I have received some screenshots of the system. What I want to know is whether I can configure the FG-200 with the same function block application that the 3240 FIM has, so that say I disconnect segment 1 of the 3420 FIM and connect it to the FG-200's segment 1, then do some testing and then, I really need to know, if I reconnect FF segment to the 3420 FIM, whether "the original system would continue working normally". I have this doubt because in the case of the FG-200, it is necessary to configure the function block application, and download the configuration to "both", the FG-200 and the field devices. But I don't want to change anything in the field devices with the Softing tool, lest the field devices became unable to communicate to the original system. So I am thinking to download the configuration only to the FG-200, without the field devices connected to the FG-200. Of course the configuration tool will say that it cannot download the configuration of the field devices, because of course they are at this point not connected, and a lot of errors will show-up in the log, but I think (although I an not 100% sure) the function block application should be configured in the FG-200.

Once the configuration is downloaded to the FG-200, I would disconnect the segment from the 3420 FIM and hook-up the FG-200, then use the MBPOLL tool to verify the process value, the built-in web server and the configuration tool itself to try to check process values and so on. But if after reconnecting the field devices to the old 3420 FIM the system does not reestablish operation, that would be a disaster.

I am so concerned that I may hire an Emerson engineer to be with me during the commissioning so that he can fix anything I screw-up, and if the customer says they do not want to pay for the Emerson engineer, then I will wish them the best of luck with their endeavour.

Anyways, thanks for reading down to this point, and if you have advice, it will certainly be most welcome.

 

[danw]

Serial-to-Ethernet
You said the Yokogawa was Modbus RTU which a serial connection, presumably over RS-485. The FG-100 is Modbus/TCP over Ethernet. So unless you can make changes on the Yokogawa to change the port for accessing these temp transmitters, you'll need a Modbus enabled serial/Ethernet converter.

Function blocks
My very limited FF exposure lead me to believe that the function block configuration was done in the field transmitter itself> I used Pactware and a Softing MobilLink FF modem to put in tag name and square root function in FF DP transmitters and it appeared that the function block was in transmitter, not in the gateway, because I didn't have a gateway.

Are you sure whatever function blocks are needed are in the 3240 FIM or the replacement FG-200? When I look at the network diagram in the FG-200 spec sheet, it sure looks like a plain old gateway, a protocol converter (FF H1-to-Modbus or FF H1-to-FF HSE converter).

If the function blocks are in the field instrunents, then there's nothing to reprogram, you're changing the gateway protocol converter, right?

Emerson engineer
What's the likelihood of getting a "Emerson engineer" on-site who knows the ins and outs of an obsolete FF gateway? Isn't the guy dispatched to show up likely to be 4 months out of college, energetic and willing and hoping you know more than he does?

Maybe being in Asia, where I've heard there's more installed FF than North America, there's a higher likelihood of a Rosemount/Asia service guy having FF exposure, but the obsolescence factor still leaves some unknowns on the table.

 

[AlfredoQuintero]

Dan, thanks lot for your investment of your time in giving me this advise.

Serial-to-Ethernet
You said the Yokogawa was Modbus RTU which a serial connection, presumably over RS-485. The FG-100 is Modbus/TCP over Ethernet. So unless you can make changes on the Yokogawa to change the port for accessing these temp transmitters, you'll need a Modbus enabled serial/Ethernet converter.

The FG-200 does have one serial port. It can be used to setup a redundant system with another FG-200, or it can be used for the Modbus RTU slave interface. So the Ethernet to serial converter is not needed. The DCS is Yamatake, not Yokogawa.

Are you sure whatever function blocks are needed are in the 3240 FIM or the replacement FG-200? When I look at the network diagram in the FG-200 spec sheet, it sure looks like a plain old gateway, a protocol converter (FF H1-to-Modbus or FF H1-to-FF HSE converter).

If the function blocks are in the field instrunents, then there's nothing to reprogram, you're changing the gateway protocol converter, right?

Exactly. But it is necessary to configure the FG-200 in order to tell the FG-200 the address of the Modbus database in which the FF data has to be mapped. Hence my idea of configuring the FG-200 without the field devices so as to not screw-up the configuration in the field devices.

Emerson engineer
What's the likelihood of getting a "Emerson engineer" on-site who knows the ins and outs of an obsolete FF gateway? Isn't the guy dispatched to show up likely to be 4 months out of college, energetic and willing and hoping you know more than he does?

Good question and I do not have an answer. The customer is one of the largest refineries in Japan and is located very close to one of Emerson's offices in Chiba prefecture. I think this customer can request a knowledgeable engineer, and Emerson Process Japan can provide him/her. Since you mention this, I will change the conditions in my offer and state the requirement of the presence of an engineer for the FIM in case the configuration of the field devices is corrupted, as opposed to my original idea of my hiring this engineer. Thanks for the hint!

Maybe being in Asia, where I've heard there's more installed FF than North America, there's a higher likelihood of a Rosemount/Asia service guy having FF exposure, but the obsolescence factor still leaves some unknowns on the table.

I think that Asia needs to be qualified a little bit. India does have a huge installed base of FF thanks to Reliance Industries which has adopted widely and who has invested in the training of human resources. In the case of Japan which developed its process industries after WWII, there is a significant amount of analogue first and foremost, then there is HART, and a Yokogawa thing called BRAIN, then further down there is some FF, and I have never seen it but there may be some Profibus PA somewhere.
Again, thanks very much.

 

[danw]

Great that the FG-200 gateway has an RS-485 port, because that eliminates another box configuration.

My FF experience was talking point-to-point from Config software to a single FF transmitter. The Softing modem 'scanned' the 'network' (point-to-point) with a "Live List" function and reported what it was attached to. Pactware had to have the DTM FF file for the field transmitter already installed in its catalog.

The Softing website has 5 videos on the FG-200.
https://industrial.softing.com/us/p...fieldbus-fg-200-hseff-modbus-for-control.html

It might be worth your while going through them because that's Softing's tech support. And it gives you a basis for asking for clarification before you're stuck on the job site trying to decipher a given step.

If I correctly interpret the first video, Creation of a Configuration based on a Network Scan, I don't see the field instrument configurations changing at all. They're already configured. The FG-200 scans the existing wired H1 network with its "Live List" function and creates a segment with the devices the FG-200 'sees' when it scans the network.

The good news is that the Live List function does that when you're on site and connected to the H1 network. But it means you have to be on-site to get the FF segment populated with device info in the FG-200.

The Modbus video, FG-200 Adaption of the Modbus mapping generated by the Communication Configurator infers at 00:50 that the PV is automatically assigned a Modbus register (3)0001 as a floating point value.

It isn't clear to me which transmitter this is for. Is there a complete Modbus map for each transmitter, requiring a 485 slave node address ID for each transmitter (and a separate Master polling message for each transmitter)? The FG-200 Modbus map cannot be custom mapped so xmtr 1 PV is at 30001, xmtr 2 PV is at 30003, xmtr 3 PV is at 30005? I don't know.

The rest of the video goes through how to use a .csv file to offset the set of Modbus values for that instrument from starting at (3)0001 to starting at (3)0101.

The bad news is that I suspect the Modbus mapping function needs an established, defined segment in order to configure the Modbus slave register table.

I'm not quite clear on where the Device Description files shown on the right in the video, FG-200: Maintenance of device types within the Communication Configurator come from. Are these descriptions part of the configuration software that goes along with the FG-200? I suspect not.


When I used the Softing Modem with Pactware, I had to import FF DTM device descriptor files from the transmitter manufacturer and install the DTMs in Pactware in a somewhat similar procedure as shown for the FG-200. That's how I knew where the file was, was because I had to obtain the file and make it available for installing the FG-200 catalog. You might want to clarify with Softing where the device descriptor files come from.

The Modbus configuration was light duty enough that it isn't clear if anything can be mapped without a valid FF segment with all its field devices and device descriptors, but I suspect those are all needed, leaving the Modbus mapping as a task after connection the H1 network and creation of a segment.

 

[AlfredoQuintero]

Dan, shame on me! I had totally overlooked these videos. Thanks for the explanation above as well. I will be studying all this information before I submit my proposal.