Re: [foxboro] FBM230/Modbus slave too fast problem
- From: Pablo Lioi <plioi@xxxxxxxxxxx>
- To: "foxboro@xxxxxxxxxxxxx" <foxboro@xxxxxxxxxxxxx>
- Date: Fri, 13 Mar 2015 23:51:00 +0000
I would like to share my conclusions about this problem so far, as I think they
may be useful for others.
In my original post I omitted some details for the sake of simplicity, but now
I will describe my problem and findings more thoroughly.
Right now I have an FBM230 talking to a Fisher ROC using Modbus RTU over RS232,
at 9600 bps, on port 3. This communication is working ok.
I want to modify my system so that I can use Port 3 to communicate with 3
(networked) Fisher ROCs using Modbus/TCP. So I installed a Modbus Serial to
Modbus TCP Gateway (Schneider TSX ETG100), connected Port 3 of the FBM230 to
the serial port of the gateway and the Ethernet Port of the gateway to the ROC
network. My first test was to read one register from the same ROC I was
previously reading over RS-232. The ROC replies and the gateway converts the
response back to Modbus RTU and sends it to the FBM230, but the FBM230 discards
the message and retries.
I know for sure that settings, cables, dip switches, etc., are Ok, because I'm
monitoring the line with a protocol analyser. As I stated before, I've already
had problems with slaves that replied too quickly to polls from the FBM230, so
this time I made some measurements with the protocol analyser. The results are
as follows:
In my original configuration, with the FBM230 reading from the ROC over RS232
at 9600 bps, the time between the last byte of the query and the first byte of
the response message is 16 ms.
In my new configuration, with the FBM230 connected to the gateway al 9600 bps,
the time between the last byte of the query and the first byte of the response
message is 8 ms. If I change the baudrate to 38400 bps, the time goes down to 5
ms.
Conclusions so far:
- In my new configuration, the difference in response time between 9600 and
38400 bps is due to the 3.5 char times the gateway waits to mark the end of the
incoming message on the serial port. At 9600 bps, this time is 3.4 ms and at
38400, 0.8 ms. As nothing changes on the Ethernet side, I am assuming that the
time it takes for the gateway to convert the message, send it to the ROC over
Ethernet and get the response is the same in both cases (4.5 ms, more or less).
- The FBM230 modbus master driver software is doing something (flushing the
receive buffer, probably) that doesn't allow it to cope with responses that
arrive in less than, lets say, 10 ms.
- The only solution I see is to install a gateway with a configurable response
delay on the RS-232 side, or something to that effect.
Thanks to everybody
Regards
Pablo Lioi
Total Austral S.A.
Tierra del Fuego - Argentina
> From: dmarc-noreply@xxxxxxxxxxxxx
> Subject: Re: [foxboro] FBM230/Modbus slave too fast problem
> Date: Thu, 12 Mar 2015 12:15:59 -0400
> To: foxboro@xxxxxxxxxxxxx
>
> I believe if you can produce the foxboro PO or sales order they will send you
> an updated driver at no cost. It will be a computer transfer.
>
> Sent from my iPhone
>
> > On Mar 12, 2015, at 11:00 AM, Pablo Lioi <plioi@xxxxxxxxxxx> wrote:
> >
> > Scott,
> > As I stated before, my driver is version 1.1 and I am using B0700BG-C.
> > Supported device counters are Good Messages Sent, Good Messages Received
> > and Timeouts.
> > I suppose I should upgrade the driver...
> > Regards
> > Pablo Lioi
> >> From: Scott.Landry@xxxxxxxxxxxxxxxxxxxxxx
> >> To: foxboro@xxxxxxxxxxxxx
> >> Date: Thu, 12 Mar 2015 08:51:55 -0400
> >> Subject: Re: [foxboro] FBM230/Modbus slave too fast problem
> >>
> >>
> >>
> >> Pablo,
> >>
> >>
> >> Latest Modbus driver is V2.3.4 (B0700BG-J document)
> >>
> >> If you are using RS-422 or RS-485;
> >>
> >> Set the RS-232 switches for the channel such that the C5 and C6 switches
> >> are in
> >> the CLOSED position and the rest of the switches are in the OPEN position
> >> (see
> >> Figure 4-15 on page 39).
> >>
> >>
> >>
> >> Something else I do when I set a FDSI interface up is configure all the
> >> port and device counters, possible.
> >> I also have a graphic I re-use so they are easy to see (depends on the
> >> driver) some drivers
> >> have a script to build the blocks and graphic(s). Although not Modbus
> >> driver.
> >>
> >>
> >> Port Counters:
> >> 1. Number of good messages sent
> >> 2. Number of good messages received
> >> 3. Number of parity errors
> >> 4. Number of framing errors
> >> 5. Number of checksum errors
> >> 6. Number of transmit errors
> >> 7. Port Scan Time
> >> 8. Reset Counters
> >>
> >> Device Counters:
> >> 1. Number of good messages sent
> >> 2. Number of good messages received
> >> 3. Response timeout
> >> 4. Number of Modbus Device Exception responses
> >> 5. Number of Modbus gateway exception responses
> >> 6. Last Error message received from the gateway
> >> 7. Configured Options
> >> 8. Number of Over Scans
> >> 9. Over Scan time high
> >> 10. Group Scan time
> >>
> >>
> >> Regards,
> >>
> >>
> >> Scott
> >>
> >> ________________________________________
> >> From: foxboro-bounce@xxxxxxxxxxxxx [foxboro-bounce@xxxxxxxxxxxxx] On
> >> Behalf Of Landry, Scott [Scott.Landry@xxxxxxxxxxxxxxxxxxxxxx]
> >> Sent: Thursday, March 12, 2015 7:14 AM
> >> To: foxboro@xxxxxxxxxxxxx
> >> Subject: Re: [foxboro] FBM230/Modbus slave too fast problem
> >>
> >> Pablo,
> >>
> >>
> >> Try playing with the driver scan rate, slower and faster. See what
> >> difference it makes, if any.
> >>
> >> Are you using an xml device config file ?
> >>
> >>
> >> Regards,
> >>
> >>
> >> Scott
> >>
> >> ________________________________________
> >> From: foxboro-bounce@xxxxxxxxxxxxx [foxboro-bounce@xxxxxxxxxxxxx] On
> >> Behalf Of Pablo Lioi [plioi@xxxxxxxxxxx]
> >> Sent: Thursday, March 12, 2015 5:10 AM
> >> To: foxboro@xxxxxxxxxxxxx
> >> Subject: Re: [foxboro] FBM230/Modbus slave too fast problem
> >>
> >> I'm using RTU, but my communication is not a radio link. The FBM230 and
> >> the slave are inside the same cabinet.
> >>
> >>> From: dmarc-noreply@xxxxxxxxxxxxx
> >>> Subject: Re: [foxboro] FBM230/Modbus slave too fast problem
> >>> Date: Wed, 11 Mar 2015 15:18:10 -0400
> >>> To: foxboro@xxxxxxxxxxxxx
> >>>
> >>> Are you using Modbus rtu or ASCII? On radio units I have had better luck
> >>> with Modbus ASCII.
> >>>
> >>> Sent from my iPhone
> >>>
> >>>> On Mar 11, 2015, at 1:28 PM, Pablo Lioi <plioi@xxxxxxxxxxx> wrote:
> >>>>
> >>>> Hi everybody,
> >>>> I'm sending this again because for some reason carriage returns seem to
> >>>> have been lost and the post was difficult to read.
> >>>> In the past I had a serious problem with an FBM230 with Modbus protocol
> >>>> communicating with a slave using RS-232, in RTU mode.
> >>>> When the slave response to any query from the master arrived too fast,
> >>>> the FBM would ignore the response and retry. I finally worked around the
> >>>> problem delaying the response from the slave a few milliseconds using a
> >>>> setting (at the slave device) that was meant for radio communication
> >>>> (RTS to TxD delay).
> >>>> I am now facing the same problem again, with another slave device. The
> >>>> problem is that this time I have no way to delay the slave response. I
> >>>> tried using RTS/CTS handshake on the RS-232 line, but the problem
> >>>> remains. I know for sure that cable and switch settings at the TA are ok
> >>>> because I'm looking at the line with a protocol analyser and see correct
> >>>> query and response messages.I would like to know if somebody else has
> >>>> experienced this problem too.
> >>>> Regards
> >>>> Pablo Lioi
> >>>> TOTAL AUSTRAL S.A.
> >>>> Tierra del Fuego - Argentina
> >>>>
> >>>>
> >>>>
> >>>> _______________________________________________________________________
> >>>> This mailing list is neither sponsored nor endorsed by Invensys Process
> >>>> Systems (formerly The Foxboro Company). Use the info you obtain here at
> >>>> your own risks. Read http://www.thecassandraproject.org/disclaimer.html
> >>>>
> >>>> foxboro mailing list: //www.freelists.org/list/foxboro
> >>>> to subscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
> >>>> to unsubscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave
> >>>
> >>>
> >>> _______________________________________________________________________
> >>> This mailing list is neither sponsored nor endorsed by Invensys Process
> >>> Systems (formerly The Foxboro Company). Use the info you obtain here at
> >>> your own risks. Read http://www.thecassandraproject.org/disclaimer.html
> >>>
> >>> foxboro mailing list: //www.freelists.org/list/foxboro
> >>> to subscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
> >>> to unsubscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave
> >>
> >>
> >>
> >> _______________________________________________________________________
> >> This mailing list is neither sponsored nor endorsed by Invensys Process
> >> Systems (formerly The Foxboro Company). Use the info you obtain here at
> >> your own risks. Read http://www.thecassandraproject.org/disclaimer.html
> >>
> >> foxboro mailing list: //www.freelists.org/list/foxboro
> >> to subscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
> >> to unsubscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave
> >>
> >>
> >> ______________________________________________________________________
> >> This email has been scanned by the Symantec Email Security.cloud service.
> >> ______________________________________________________________________
> >>
> >> *** Confidentiality Notice: This e-mail, including any associated or
> >> attached files, is intended solely for the individual or entity to which
> >> it is addressed. This e-mail is confidential and may well also be legally
> >> privileged. If you have received it in error, you are on notice of its
> >> status. Please notify the sender immediately by reply e-mail and then
> >> delete this message from your system. Please do not copy it or use it for
> >> any purposes, or disclose its contents to any other person.
> >>
> >>
> >> _______________________________________________________________________
> >> This mailing list is neither sponsored nor endorsed by Invensys Process
> >> Systems (formerly The Foxboro Company). Use the info you obtain here at
> >> your own risks. Read http://www.thecassandraproject.org/disclaimer.html
> >>
> >> foxboro mailing list: //www.freelists.org/list/foxboro
> >> to subscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
> >> to unsubscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave
> >>
> >>
> >> ______________________________________________________________________
> >> This email has been scanned by the Symantec Email Security.cloud service.
> >> ______________________________________________________________________
> >>
> >> *** Confidentiality Notice: This e-mail, including any associated or
> >> attached files, is intended solely for the individual or entity to which
> >> it is addressed. This e-mail is confidential and may well also be legally
> >> privileged. If you have received it in error, you are on notice of its
> >> status. Please notify the sender immediately by reply e-mail and then
> >> delete this message from your system. Please do not copy it or use it for
> >> any purposes, or disclose its contents to any other person.
> >>
> >>
> >> _______________________________________________________________________
> >> This mailing list is neither sponsored nor endorsed by Invensys Process
> >> Systems (formerly The Foxboro Company). Use the info you obtain here at
> >> your own risks. Read http://www.thecassandraproject.org/disclaimer.html
> >>
> >> foxboro mailing list: //www.freelists.org/list/foxboro
> >> to subscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
> >> to unsubscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave
> >
> >
> >
> > _______________________________________________________________________
> > This mailing list is neither sponsored nor endorsed by Invensys Process
> > Systems (formerly The Foxboro Company). Use the info you obtain here at
> > your own risks. Read http://www.thecassandraproject.org/disclaimer.html
> >
> > foxboro mailing list: //www.freelists.org/list/foxboro
> > to subscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
> > to unsubscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave
> >
>
>
> _______________________________________________________________________
> This mailing list is neither sponsored nor endorsed by Invensys Process
> Systems (formerly The Foxboro Company). Use the info you obtain here at
> your own risks. Read http://www.thecassandraproject.org/disclaimer.html
>
> foxboro mailing list: //www.freelists.org/list/foxboro
> to subscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
> to unsubscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave
>
_______________________________________________________________________
This mailing list is neither sponsored nor endorsed by Invensys Process
Systems (formerly The Foxboro Company). Use the info you obtain here at
your own risks. Read http://www.thecassandraproject.org/disclaimer.html
foxboro mailing list: //www.freelists.org/list/foxboro
to subscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
to unsubscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave
Other related posts:
