Local display of fieldpoint data

[Gurdas]

Hi!,

We have a setup wherein Fieldpoint is acquiring data from test workstations and sending them over Ethernet (or Serial) to a central server. There are 5 test stations and 1 server.

There are about 7-10 measurements per station. Each variable is being measured at 1 sample/5 secs (0.2 Hz).

Now, at each of these test stations, we also need a local display of the data being sent to the server.

1) Can Fieldpoint handle this?
2) What hardware (display etc) would answer my requirement?

We are using the standard fieldpoint modules (non-real time) and server can be either Win2000 or Win2003.

Development will be on LV PDS 6.1

Thanks in advance.

- Gurdas
gurdas@qagetech.com

[Brooks B.]

Gurdas:

Using FieldPoint (FP) to log the data at each station (7-10 measurements) and send that data back to the central server will be no problem at all. Just select the FP controller and modules to fit your measurement requirements.

Now, displaying the measurements locally will add some complexity. National Instruments does not have any display hardware for use with FP controllers. However, the simplest and cheapest solution I can think of would be a 3rd party display unit (LED or TFT style display) that can be programmed via an RS-232 or RS-485 serial port. You could select a FP controller that communicates with the server computer via Ethernet (TCP/IP) and has a serial port on it for communication with the TFT display unit. Your program would t
hen update the TFT display every 5 seconds or so. The FP-2000 series controllers have RS-232 serial ports and TCP/IP ports on them and would work well as the controller.

Hope this helps...

Good Luck!

Sincerely,

Brooks B
Applications Engineer
National Instruments

[Gurdas]

Brooks,

Thanks for your inputs.

Please note I am using FP-1601 and not a controller module because of hardware cost and also because we do not have the real time version of LV.

I have attached a schematic of the setup to make things clearer. This is only a proposed setup and if someone can show an easier (and cheaper) path, we'd happily walk it.

Further questions/points:

1) Though there are 5 test stations, we are using only ONE fieldpoint node. This would have one FP-1601 and necessary analog and digital I/O.

2) What would be the solution(s) with FP-1601 in picture instead of the controller modules.

3) Computers are pretty cheap in India. What if I use a full blown PC (P4, 128MB RAM, 20GB HDD, 15" colour monitor, Ethernet) as a di
splay option? Can my LV code running on the server push relevant data to each of these PCs?

4) Can I use and program the controller type modules (FP-20xx) without using LV-RealTime?

I look forward to help on getting the config done and thanks once again for your time and effort.

Best wishes,
Gurdas
gurdas@qagetech.com

[Brooks B.]

Gurdas:

Your schematic would work as you drew it up. You can use the FP-1601 controller to send the signals acquired back to the server CPU. That server CPU can then analyze and present the data in whatever way you wish (graphs, etc). At this point the FieldPoint/data acquisition is done and you just need to get the display to each station. A computer at each station is the best solution. You can have each computer running a VI that gets data from the server via Datasocket or even have the test station PCs load a remote-panel VI from the server PC. Multiple options here as well. Here is the dataflow I would image to be the most robust:

1) Sensors take data to FP input modules
2) FP 1601 sends data to LabVIEW code on server (via ethe
rnet)
3) Server sends processed (if needed) data to all workstation PCs over Ethernet using Datasocket or TCP/IP
4) Workstation PCs display data (each running a VI locally) by reading datasocket or TCP/IP data over ethernet.

I think this should work well for you!

Good Luck,

Sincerely,

Brooks B
Applications Engineer
National Instruments

[Gurdas]

Brooks,

Looks good. Thanks!

Now each of the 5 test stations are independant of each other. That is, they will be testing different kinds of motors, start and stop at different times. In short, there is absolutely ZERO corelation between the test stations.

Will the hardware and software (on the macro level, I am not concerned about programming details at this stage) strategy suggested handle this?

For e.g.; each station needs to inform the server about start and stop of testing so that only relevant test data is measured. Also, for the current motor type, the VI running on the server needs to load optimum parameters (to tell whether motor tested is OK or not OK) and so on.

1) Can there be 5 VIs (one for each station) running s
imultaneously on the server?
2) Can the server receive control commands from the PCs at each of these test stations (say via the digital input fieldpoint module)?

There is another issue concerning isolation which I have posted as a separate thread.
http://exchange.ni.com/servlet/ProcessRequest?RHIVEID=101&RPAGEID=135&HOID=50650000000800000053C60000&USEARCHCONTEXT_CATEGORY_0=_18_%24_17_&USEARCHCONTEXT_CATEGORY_S=0&UCATEGORY_0=_18_%24_17_&UCATEGORY_S=0

See if you can help me here too.

That is a lot of questions and I must thank you for your patience.

Best wishes,
Gurdas

[Brooks B.]

Gurdas:

Lets take a step back here...

You have 5 test stations with computers and ethernet access. You have 1 FP bank with a FP-1601 controller. You need each station to take measurements with different configurations, etc.

The BEST way to do this is run LabVIEW on each of the 5 test stations (don't need the server) and have each station do the FP reads from the IP address of the FP controller when they need to do so. The ONLY reason not to do it this way would be if you don't want to have copies of LabVIEW on each of the 5 systems. If that is the case, then you can create LabVIEW EXE files to run on each test machine and not need a full copy of LabVIEW on the test machines. There is no need for the server comput
er. Does this make sense?

Good Luck!

Sincerely,

Brooks B
Applications Engineer
National Instruments

[Gurdas]

Brooks,

Thanks and it makes all the sense!
Wonder why it did not strike me 🙂

Probably because I was not aware that the FP will be seen as a node with its distinct IP address on an ethernet network having computers (am I right?).

I'd prefer to have EXE files on each PC because we (the integrators) have LV-PDS while the client does not have any LV development environment

One small issue: I can still have the data from each test station (and its PC) stored on a central server?

Thanks again,
Gurdas
http://www.qagetech.com