Network-published variables and cFP targets

[CAndersen]

I have created a dozen or so network-published variables and bound them to different targets on our cFP 2120.  Will it be possible to use a VI to query these network variables in order to access and work with the data from the cFP?  By this I mean do I have to have a separate VI running to populate these network variables before my queries will be successful?  Or is the network variable always automatically populated with the correct and current data as long as it has been bound to the proper source?

[RavensFan]

You have the shared variables bound to particular channels of the compact Fieldpoint?  If so, then those shared variables will automatically have the most recent data from that channel.  No need to add any code.

Now if the shared variable is located on the cFP, but not bound to any of the channels, then you would need a VI on the cFP that reads the channel and writes it to the shared variable.

[CAndersen]

Thanks!  Yes, I believe the shared variables are bound to particular channels on the cFP.  I created a variable library in my project and then selected the particular cFP channels in the "Bind to Source" area while creating the variables.

One follow up question... I have several input modules on this cFP (1 AI-100 and 2 AI-111s and others).  The AI-100 has eight channels with a per channel max update rate of 360Hz.  Each AI-111 has sixteen channels with a per channel max update rate of about 1Hz.  I suppose it may depend on many things, but would you expect to have problems sampling all of these inputs simultaneously at their maximum update rates from a VI running on a PC on the same subnet?  I basically want to be able to log data from any or all of these channels onto a hard drive on the subnet (since the cFP doesn't have the storage capacity I need for this logging).  I've been thinking of setting up the datalogging code in one (or more, depending on the particular test setup) VI that would run on a single dedicated PC on the subnet. 

Is there a better way to do this?  There's no way to hook a hard drive directly to the cFP, is there?  I just want to be certain that I won't sacrifice my effective sample rate because of network lag on the ethernet line and I expect that I'll need many megabytes of storage over the span of any particular test. 

[DiscoBall]

Hi,

We generally don't provide network throughput benmarking data because network traffic conditions vary widely between customers. However I do suggest reading through KnowledgeBase 1O3CJ7US Sampling Speed of FieldPoint Modules and KnowledgeBase 335J66FL: How Can I Optimize the Update Rate of a FieldPoint Ethernet Network Controll... (I will be updating to fix broken links eventually...). You can get some Network Shared Variable benchmarks from NI Developer Zone Tutorial: Using the LabVIEW Shared Variable. I am unsure how often a bound network variable is updated and how that would affect throughput.

You could setup a benchmarking test such that you know the waveform that is coming into the AI-100 and see if how much of the waveform you are able to reconstruct. This should give a general idea if your effective sampling rate is good enough for your application.

How much storage space are you needing? Are you using LabVIEW Real-Time and embedding VIs on your cFP-2120? Perhaps another product/approach is better suited for your application...what is your application in general?

Message Edited by DiscoBall on 10-14-2009 06:33 PM

[CAndersen]

Thanks for the help!  The hardware and software package we have was originally selected in order to provide data acquisition and control for a fairly flexible test bench.  We were sampling and logging pressures, flows, temperatures, and using PID loops for other pressures and flows.  We have the RT and DSC modules with LV 8.5. 

That project was shelved a while back and now we have a different project and would like to resurrect the hardware to help out.  This project, at this point, consists simply of datalogging one or more tests simultaneously.  Each test may have ten or so analog inputs, some of which are coming into the AI-100 and need faster sampling and others are coming into the AI-111 and don't need the quicker sampling.  We might have up to five tests running at the same time for a total of maybe 40 inputs.  I'd guess the overall maximum desired throughput will probably be on the order of around 1000-1300 Hz. 

As far as storage space goes, it would be easiest if I could simply wire up a 1TB hard disk drive to the system and just use it as a test archive, as well.  If we had to minimize the size, I expect that anything less than about 500 MB would be too small to use easily since some of our tests are going to be sampling for months on end, 24 hours a day. 

It's been a few years since I had the LabVIEW part of my brain lubed up and running, so I'm trying to remember how to embed my VI on the cFP.  I seem to remember doing that in the past, but can't remember the specific advantages or exactly how to do it. 

I'd guess there probably are other products that are better suited for this particular application, but we don't have a budget for new hardware/software and I'd love to make this work with what we've got.

I checked through the links you attached.  Do you know if there are active links to some of those throughput utilities for the cFP 2120? 

Thanks again for the help!

Cameron Andersen

Project Engineer

Smith Lift

[DiscoBall]

I am still looking for those spreadsheet files. Using the functionality of DSC is the way to go for your application since there is no way to connect a hard drive of that size to the CFP 2120.

I will post here if I can find those spreadsheets, however I doubt they still apply since you are using a cFP 2120 and not a FP 1600 or FP 1000....