Is anyone using a HART network for process data?

CCZ

Member
Join Date
Mar 2009
Location
FL
Posts
25
I am doing a bit of research before the next project hits.

For what I am finding to looks like you *should* be able to set up a multidrop HART network with a gateway to get all the data back to the PLC.

Is anyone doing this in the real world? If so, how many nodes do you have on a network and what kind of update times can you get?

Of course I can read the available design guides, but experience is invaluable.

Thanks,

Chris
 
HART multidrop is a daisy chain 'loop/network' of multiple HART devices and the receiver, where each HART device is powered by the current in the loop and where the receiver is a HART multiplexor (MUX) that polls each HART device one at a time. The MUX is usually a Modbus Slave/Server that a Modbus host polls to get the contents of the Modbus registers which are populated with the HART data.

In HART multidrop, the analog loop signal is limited to 4mA, which powers the transmitters. The loop wiring also provides the signal transmission path between the HART MUX and the field devices on the network. The HART values are polled by the MUX, one at a time, each field device in turn at a max rate of a 2-3 reads per seconds. HART addresses are 1 through 15.

HART multidrop reduces AI I/O cost at the controller by not requiring HART enabled AI's, but the trade-off is the cost of a HART multiplexor and its configuration.

The host system (the controller) needs software that can handle the output of the HART multiplexor. Modbus has been common in the past, but the future might see a HART Mux use HART-IP, an Ethernet protocol that can be used by DCS Asset Management software ($25k software). As of February 2019, Moore Industries, who manufactures a HART module that uses HART-IP, could not name for me any software other than $25k DCS asset management software that uses HART-IP. Be certain you understand which host software (for communicating with the MUX) is available on the host and what its requirements are.

HART multidrop throughput data rates are definitely more limiting than Foundation Fieldbus or Profibus. A rule of thumb is two PV's per second in a HART multiloop. If you have 8 devices on the loop, then each the MUX get an update from any given field device once every 4 seconds. That works fine for a tank farm but is too slow on most flow loops.

You don't say why you're considering HART multidrop. If you need multiple outputs from a single transmitter, like a Coriolis meter (mass flow, temp and density), there are break-out boxes that take the HART variables off the 4-20mA signal and provide a 4-20mA output for each variable. Moore Industries model HIM is one such device.

I have read 'opinions' that the HART signal is not as robust as Foundation Fieldbus (FF) or Profibus PA signals in very electrically noisy environments. There are reports of HART being swamped by electrical interference. I have no experience one way or the other.


Be aware that the field wiring is a continuous current loop. If the current loop is opened anywhere, at any point in the loop, the open circuit brings down the whole loop.

Some HART devices can take as much as 45 seconds to initialize when powered up from a cold start.
 
Last edited:
FWIW, I've never had a problem with electrical noise.

Keep in mind that not all Hart devices are created equally. The TDR level sensors that were chosen for me used there own Hart software, and could only use their own specific Hart interface.
 
That's an interesting point when involving data acquisition.

I agree that not all HART devices are offered with a HART-approved DD file. Honeywell offers only a DTM (for Pactware) for its Guided Wave Radar, it does not offer a DD file.

If the host device needs a DD file to communicate with a field device, then one can not assume that a DD file is available, it's a case-by-case basis. The vast majority probably do, but there's no guarantee.

I doubt that a MUX requires a DD file. It's just issueing conventional polling commands for variables and expecting a conventional response.
 
Correction to my earlier summary:


The receiver in HART multidrop is NOT a mulitplexor.


In HART Multidrop, the field instruments and the receiver are wired in parallel to the DC power supply dseries connection between all field devices and the receiver.



1.png


HART multiplexors use conventional home run loop wiring where the 4-20ma signal is the process signal at the Analog Input, but the 4-20mA signal is monitored by a HART MUX which provides HART data as configured.






HART-mux.jpg
 
Interesting, you don't put the "+" of the Hart communicator between the "+" and the resistor? I always did, but the Hart programming devices that I've used have their own power supply.
 
Viator manufactures HART modems. Their recommendation - connect either across a loop resistor or the transmitter terminals.



Viator-suggestion-to-connect-across-the-resistor-of-the-transmit.jpg
 
Thanks so much for the information!

My goal for this would be to decrease cost of I-O hardware and wiring, I was hoping to be able to get more information (fault bits?) from individual sensors, but with a throughput of only 2 PV's per second that is probably not reasonable.

Our control system for this would be CompactLogix 5480. From what I understand these do not yet have HART AI cards available yet. Communications to the gateway would be Ethernet/IP.

A couple of possibilities for the gateways would be the following:

http://www.prosoft-technology.com/P...Net-IP/EtherNet-IP-to-HART-Multi-drop-Gateway

https://www.aparian.com/products/hart4


Sorry for the slow reply. I was offline over the weekend and Monday..., well it was a Monday.

-Chris
 
Fault bits? Uh-huh.

If you look at pages 93 to 97 of the Prosoft manual
http://www.prosoft-technology.com/content/download/737/6649/version/4/file/hart_protocol_manual.pdf
it looks to me like the status words are HART communications related faults and status, not faults related to field transmitter operation.

For decades, factory know-it-alls rave on about all the great diagnostics available on HART, and when I challenge them for the list and specifics about how to access those diagnostics, I hear, "I'll get back to you on that" and nothing ever happens.

I suspect HART diagnostics are communications diagnostics, just like the manual spells out, not diagnostics about the transmitter. And you need to do what you're doing to read any of it.

If this moves forward, an update on whether you find what you want would be valuable.
 
"factory know-it-all's" :) yep, I try not to be one of those.

I don't think we are going to pursue this one much further. The update times won't make sense for most of what we are doing. It also would not make sense to set this up for just the few point is would work for.

Thanks again for your input.

-Chris
 
IO-Link looks like a possibility, but I am not seeing a broad support for it outside IFM Effector.

-Chris
 

Similar Topics

I'm working on my 61131-3 knowledge for some up-coming projects, and there is an automated saw on the list that would probably be the perfect...
Replies
11
Views
3,728
Anyone reset a totalizer in flow meters using hart? I am using a if8h card?
Replies
1
Views
2,840
I'm virtualizing a water treatment plant. While the IT hardware cycle is 3-5 years, to my mind that's really not practical for this kind of...
Replies
8
Views
1,583
Hello PLCs.NET! I'm trying to auto-backup some RSNetworx files. Any idea if it's possible at all? I honestly thought that it was just an...
Replies
0
Views
1,532
We are looking to collect information on our cisco and rockwell assets. The exhaustive list that cyber security is asking for .. serial number...
Replies
0
Views
1,647
Back
Top Bottom