USB Datalogging

I have been asked to create a very basic control system that allows rudimentary datalogging. My problem is that the request for datalogging is that an end user will be able to walk up to the system, plug in their laptop via USB, and start collecting "streaming" data in Excel. Any thoughts on how to make this work?

Thanks,

Marc

What type of control system are you talking to?
What type of software is available to the end user?

Walk up, plug in, start software.

http://www.measurementcomputing.com/

Not sure it would make a difference what control system it was in the long run. The issue may be if you can convert the plc's communication to work with USB, for some may have to permanenly connect that companies USB adapter.

The laptop would probably need the plc's software and an OPC setup for collecting the data.

At least that is how I would "attack" the situation, look at converting the plc port to work with USB then see how I could setup OPC (or similar) for collecting the data.

I did something similar to this once using Kepware http://www.kepware.com/ which has a free demo version that works for couple hours at a time, may be all you need in a case like this.

Thanks everyone for the quick replies.


cbeck77: Everything is unspecified at the moment - we're very early on in the project. I just never ran up against this "Plug in by USB and start logging" requirement before and was wondering if anyone else had approached it. Typically, we use Rockwell hardware / software.

Alaric: Yes - that's the basic functionality that I'm looking for. I've seen similar DAQ packages from LabView, Omega, and Windmill. What I don't know is how robust these packages are. Have you worked with the Measurement Computing solution? One potential issue with these solutions is (I think), that the interface to Excel is only with batch exporting. Something like TalTech's WinWedge software is an example of "realtime" data logging in Excel that would be desireable in this application. Another potential issue is scalability - this specific application is very basic so this approach ought to work fine, but if the prototype looks promising and people keep adding I/O . . .

Ron: That may be the approach we need, but in the case of something like Alaric's solution, there may not even be a PLC (these are simple USB devices that allow polling of real world I/O from the client software installed on the end user's laptop). Your approach may be more scalable.

Thanks again,

Marc

It'd be cheapest to build your own. You could create a simple scp using a microcontroller and use the analog output from the rack. Leave the scaling constant in the app, or make it adjustable depending upon what you were logging.
You could have it dump the data to a thumb drive every month let the plant guy grab the thumb drive and download the data to their laptop. At the same time you could make the actual trending values to go another output for real time trending.

If you're using allen bradley you could set up a topic in linx and copy dde to an excel sheet that the customer could access at the processor, or remotely; either from somewhere on the network or through a telephone (if you set up channel 0 for use with a modem).



good luck and have fun!

I've used a couple of Measurment Computing's USB data loggers. You can set up the freebie software that comes with it doesn't write into an open excel spread sheet at a predetermined update rate. It can log data relatively fast, but if you are logging data and writing to Excel at the same time at a fast update rate your laptop can get so busy you can't get the sofware to respond to anything else. Anyting slower than 1 second updates and you should be fine unless you have a really slow laptop.

I'm currently using 4 of the measurement computing DAQ boards daisy chained together...I've got a hydraulic test stand they are currently implemented on. I'm sampling data every 1ms for 3 seconds. I am outputing this data to excel and have 8 channels of data. It's a fairly large undertaking and I really miss ladder logic. The analog and data acquistion is great, but the digital I/O is just a pain in the arse to program.
I'm using softwire and visual studio 2006.


If you are looking for something simple, the supplied software with the measurement computing stuff is pretty robust - I use the scope quite often to measure discharge time to calculate flow on our ancient test stand..

Hope this helps!
Any questions on it, feel free to PM me.. I'll do my best to answer them!