IR Thermometer

powertool

Member
Join Date
Jul 2010
Location
Quebec
Posts
6
Hi, people I have a simple problem but I am stumped as to what can be done. I have an RF network with several Modbus slaves and one Modbus master(the radio tx itself)on an RS485 two wire network. My problem is that slave address 1 to 4 work at 9600 8n2 and the slave at address 5 is fixed at 9600 8n1. When I use the first 4 slaves alone everything works fine. Of course when I introduce the fifth slave device I get framing errors seeing that the first 4 work on a 11 bit frame and the last device works on a 10 bit frame. The first 4 devices can be configured to work at any baud rate , even,odd,or no parity but the hic is that if you use no parity you must add a stop bit hence 11 bits. This long story brings me to my question, is there some way that I can have both 11 bit and 10 bit frames on the same bus? If not is there some device that would do the conversion on the fly?
o_O
 
Don't see anything on the B&B site that would work with RS485. You might call them - they are the experts on this.
If B&B can't help, I think you could do it with 2 Analog Clicks.
The third port is RS485. First Click is RS485 in, RS232 out to the second Click. Second Click is RS232 in, RS485 out.
 
I avoid no parity just for the reason you've encountered. This is the classic Modbus "no parity" problem.

The Modbus spec says that no parity shall have 2 stop bits to maintain the frame size, but try it is not at unusual to find devices that can not do 2 stops with no parity.

But Modbus is an open standard with nearly everything 'self-certified'.

My solution is to use either even or odd parity for all devices, because there's no dispute that 8 bit words have only 1 stop bit with even or odd parity. Why not?
 
I avoid no parity just for the reason you've encountered. This is the classic Modbus "no parity" problem.

The Modbus spec says that no parity shall have 2 stop bits to maintain the frame size, but try it is not at unusual to find devices that can not do 2 stops with no parity.

But Modbus is an open standard with nearly everything 'self-certified'.

My solution is to use either even or odd parity for all devices, because there's no dispute that 8 bit words have only 1 stop bit with even or odd parity. Why not?
Thanks, DanW, you seem to understand the same thing as me. The unfortunate thing is that the fifth device has a fixed interface of RS485,9600,8n1 and can not be changed so I am stuck with the no parity.
 
I was under the impression that most RS-485 isolators are wired-in:wired-out, without any conversion ability.

But it's a big world and maybe B&B has a module to address this issue.

If you find a 'frame converter' module that can handle input of 'no parity one stop bit' to 'no parity two stop bits', would you let us know?
 
If adding a second RS-485 port to your master is not feasible, here's one possible work around:

Red Lion Data Station Plus (LE) might be useable as a store and forward converter.

It has a native RS-485 port. Set it up as the slave port on the existing 8-N-2 link.

Have the Red Lion's internal Modbus master read its slave port, and then write the entire data message, unaltered, through its master port (an RS-485 option card), to the field slave device.

The separate RS-485 option card can have its comm parameters set up independently of the native RS-485 port, at 8-N-1.

Likewise the internal Master has to read the response from the field slave and write the response to the internal registers on its slave side, and then enable the slave response to reply to the Main master.

It isn't clear to me if the slave side response can be disabled while the Master side makes it request, gets the response from the field device, and writes data to the internal slave registers. But if the communications has been thought through, then there's a likelihood of that being the case.
 
If adding a second RS-485 port to your master is not feasible, here's one possible work around:

Red Lion Data Station Plus (LE) might be useable as a store and forward converter.

It has a native RS-485 port. Set it up as the slave port on the existing 8-N-2 link.

Have the Red Lion's internal Modbus master read its slave port, and then write the entire data message, unaltered, through its master port (an RS-485 option card), to the field slave device.

The separate RS-485 option card can have its comm parameters set up independently of the native RS-485 port, at 8-N-1.

Likewise the internal Master has to read the response from the field slave and write the response to the internal registers on its slave side, and then enable the slave response to reply to the Main master.

It isn't clear to me if the slave side response can be disabled while the Master side makes it request, gets the response from the field device, and writes data to the internal slave registers. But if the communications has been thought through, then there's a likelihood of that being the case.

Well that could have worked but I would end up have one master(my rf xmitter) facing another master on the same bus and I am not sure that would work seeing there are also slaves on the same bus. Well I ended up contacting the mfg. of the temp sensor an they agreed on reprogramming the firmware to agree with the MdBus specification. My project will be delayed but all around I think that it will be the best thing to do. Thank you to all for the ongoing support and suggestions. I believe the problem is resolved.

Thank you again.
 
That's great that the vendor will make a firmware change for you.

Would you mind posting who (which company) that is (or PM'ing me)? That's so unusual to take steps to comply with the spec.

Were you to try the Red Lion deal, your main Modbus Master would never see the Red Lion Master, your Main master would only see the Modbus slave port on the Redlion. The ports can be segregated.
 
I had a PM asking about the Modbus Spec/Standard requirement for 2 stop bits with no parity. Here it is:

MODBUS APPLICATION PROTOCOL SPECIFICATION V1.1b

k1qoo4.jpg

Note that even parity IS REQUIRED as a default. But who pays attention to the standard?
 
Thanks for the link but my fifth device does not have analog, just RS485
Didn't mean to throw you with Analog. There's two types of Clicks. The higher end model has analog and a RS485 port. The 485 port is the important part.
Take a closer look. At first it looks like you're wasting a couple of PLC's, because all you're doing is having one talk to the other, but it's a low cost solution.
One 24DC power supply and two PLC's. Total will be about $300.00. Software is free. Program one to input via RS485, send it out to the RS232 port. Program other unit to do the opposite. Configure each RS485 port to anything you want.
I haven't looked at all the details, but I'm sure it could be done. Quick and cheap.
Two PLC's C0-02DD1-D
24DC PSB24-060-P
Program Cable D2-DSCBL
Crossover Cable ZL-RJ12-CBL-2
 
Last edited:
That's great that the vendor will make a firmware change for you.

Would you mind posting who (which company) that is (or PM'ing me)? That's so unusual to take steps to comply with the spec.

Were you to try the Red Lion deal, your main Modbus Master would never see the Red Lion Master, your Main master would only see the Modbus slave port on the Redlion. The ports can be segregated.

DanW,thanks for the heads up on the Red Lion setup . I guess I was confused with the details. The company that is supplying the IR temperature sensor is Calex and can be found at http://www.calex.co.uk/. The sensors(PB301MT) are of high quality and easy to use. The customer support ,although in the UK is very good. The project that I am working on is to remove a long(600 feet) cable that links mv output from four load cells on a 40,000 lbs crane to the weight scale electronics. The temperature feature was asked for at the last minute. The project was brought about because the 600 foot cable broke frequently and cost $8K to replace. Now the 4 loadcells go to 4 Laurel converters plus the temperature sensor and all is sent wireless through Modbus RTU to our central DCS system. Well I would like to thank everyone for the input to this subject but with the cooperation of Calex I think the problem is now solved.
 

Similar Topics

I need to take temperature measurments of the bearings of some vibrating screens. In the past we used common type K thermocouples but they have a...
Replies
5
Views
2,066
Looking at monitoring temperatures with the following IR thermometer: http://www.omega.com/Manuals/manualpdf/M3946.pdf My question is, how to I...
Replies
25
Views
9,824
change temperature reading from ferinheit to celcius and celcius to ferinheit. the thumbwheel will represent the ferinheit reading the LED display...
Replies
20
Views
5,217
M
We have two 11MW rated compressor drive motors with resistance thermometers embedded in the windings to measure windings temperature. These are in...
Replies
2
Views
3,143
Back
Top Bottom