topic Re: Modbus Object Bug? in Lookout

Modbus Object Bug?

Mike@DTSI — Wed, 20 Feb 2008 21:53:46 GMT

Another problem with modbus object? Maybe so.

PLCs are Koyo DL06, same program, exactly just different modbus addresses. Sites poll every 60 seconds. I have the modbus objects copied exactly the same settings (except station address of course).

PLC1: Connection on 40889 sets a 1 on every other poll. Polls. Updates. Done. All good.
PLC2: Connection on 40889 sets a 1 on every other poll. Polls. Updates. Polls. Updates. Polls. Updates.... It doesnt stop.

Heres the funny thing. If i make a new modbus object:
PLC2TEMP: Connection on 40889 sets a 1 on every other poll. Polls. Updates. Done.

My modbus.ini settings:
[all]
UpdateOutputs=2
RefreshOutputs=1

Thanks for the help
Mike

Re: Modbus Object Bug?

Ryan.S — Mon, 25 Feb 2008 07:43:13 GMT

I do the same configuration and test on my computer. It polls every 15 sec. UpdateOutputs=2 RefreshOutputs=1, "immediately write" checked.

The below is the log file on serial port.

[01][01][00][00][00][01][FD][CA] is the read on every poll. See, if the write output is not changed, there is no write function sent out. On every poll, it just reads the value back, no write.

After I change the switch to ON, it sends out [01][05][00][00][FF][00][8C][3A] immediately. [05] is write single coil.

So, when immediately write is checked, UpdateOutputs is almost meaningless, because the output will be updated immediately.

What do you mean by "Updates. Done. All good." I think if you set polls=60 sec, it should poll every 60sec, it should not stop polling.

So could you also make the log file on serial port? With the log file, we may find something.

15:23:37.2 - Modbus1 ->
[01][01][00][00][00][01][FD][CA]
15:23:37.3 - Modbus1 <-
[01][01][01][00][51][88]
15:23:52.2 - Modbus1 ->
[01][01][00][00][00][01][FD][CA]
15:23:52.3 - Modbus1 <-
[01][01][01][00][51][88]
15:23:52.6 - Modbus1 ->
[01][01][00][00][00][01][FD][CA]
15:23:52.6 - Modbus1 <-
[01][01][01][00][51][88]
15:23:52.6 - Modbus1 ->
[01][05][00][00][FF][00][8C][3A]
15:23:52.7 - Modbus1 <-
[01][05][00][00][FF][00][8C][3A]
15:24:07.3 - Modbus1 ->
[01][01][00][00][00][01][FD][CA]
15:24:07.3 - Modbus1 <-
[01][01][01][01][90][48]
15:24:22.3 - Modbus1 ->
[01][01][00][00][00][01][FD][CA]
15:24:22.3 - Modbus1 <-

Re: Modbus Object Bug?

Mike@DTSI — Mon, 25 Feb 2008 17:24:45 GMT

The object does not stop polling. PollTime=60 secs., it polls ALL the time not every minute like it should.

It will just keep polling non-stop, the 40889 is a static 1, it is a confirmation to the plc (heartbeat time setting) So it doesnt transition.
I will try to log it.
Thanks
Mike

Re: Modbus Object Bug?

Mike@DTSI — Tue, 04 Mar 2008 17:15:57 GMT

This is not the first time i have run into it. Going to log it today.

Mike

Re: Modbus Object Bug?

Mike@DTSI — Thu, 20 Mar 2008 18:01:27 GMT

Here is the quick log of when i attach to the point. Site should poll every 3 minutes, update outputs every 2 polls. Site is MB3

Re: Modbus Object Bug?

Ryan.S — Fri, 21 Mar 2008 07:39:27 GMT

Is it the original diagnostic file? It seems scrambled.

For example,

19/03/08 14:13:59 - MB3 ->
[0C][06][03]x[00][01][C9]J
MB3 - RTS low
19/03/08 14:14:00 - MB3 <-
[9F]/h[F9]

function code 06 is write single register. As its response, if successful it will be an echo of the request, which means the same data sent back.

But I only see the scrambled or missing characters everywhere...

Re: Modbus Object Bug?

Mike@DTSI — Fri, 21 Mar 2008 16:43:34 GMT

Exactly, and the retries are set at 5, timeout 1200. It just keeps going, doesnt stop.

The communications at this site are not great (radio behind a very large hill), so garbled comms are expected. The driver just goes into a loop.

Re: Modbus Object Bug?

Ryan.S — Tue, 25 Mar 2008 10:14:37 GMT

I guess you don't check "value in HEX" when configure the serial diagnostic file. That's why we see some single characters in the file. But I notice that almost all the response to [0c] [06] are incorrect, while most of response to [0c] [10] are correct.

Could you give me an example of how you configure the connections? I see there are both function code [06] and [10] sent out, I want to have a look at the process. Just an example is ok.

Re: Modbus Object Bug?

Mike@DTSI — Tue, 25 Mar 2008 16:04:22 GMT

Here is the data member config. This is the connections on the driver.

Re: Modbus Object Bug?

Ryan.S — Wed, 26 Mar 2008 07:28:24 GMT

What does the dignostic file look like on the good modbus object? We can compare them.

From this dignostic file, after MB3 writes to address [03][3D] by function code [10], it writes to [03][78]. The response seems incorrect, and then it rewrites [03][3D].... it's a loop as you said.

[03][78] is 40889, named "CommReset" in the excel file. What does the "CommReset" do? What should happen after 40889 is set 1?

I guess that if you set the retry attempts to 0, it will not keep updating, right? If so, also make a diagnostic file on "retry attempts"=0.

Re: Modbus Object Bug?

Mike@DTSI — Thu, 27 Mar 2008 14:57:41 GMT

The other objects are working fine. CommReset is a constant 1, sent to the plc as a heartbeat type of response. The other sites have the exact same setings and use the CommReset. This site can use the commreset on MB4 just fine, but whenever a value is put on MB3 (0 or 1) it goes into loop. Retries do not effect it.

This exact setup is used on another system, different company of course, but same process and connections. No issues. Only MB3 is effected. It was deleted and rebuilt from a new object and the problem still persists. modbus.ini is not referencing seperate objects, only all of them.

Here is the modbus process,
Thanks

Mike

Re: Modbus Object Bug?

Ryan.S — Tue, 01 Apr 2008 06:05:24 GMT

Give me an Email address to access you.

I will give you something to test.

Re: Modbus Object Bug?

Mike@DTSI — Tue, 01 Apr 2008 15:38:22 GMT

mikecrabtree AT destekinc DOT com