Posted: 11/21/2012 7:20:12 PM EDT
|
I'm looking to get started with LabVIEW. I have a substantial collection of very decent test and measurement equipment that I want to integrate into a semi-automated test system via LabVIEW. I'm trying to get started in it with the tutorials but right now all I see is trees. I can't really make out the woods yet. I want to be able to, for example, enter a test frequency and hit "receiver tests" and have the system automatically control a signal generator, SINAD meter, distortion meter, etc. and record the results, and then switch over to comparable transmitter tests, making power output, modulation, and spectral purity measurements, all running in a very simple interface and with complete logging of selected parameters. Right now I've only just begun and even something as simple as getting a reading from one of my power meters seems to be kind of complicated. I haven't gotten that far yet. If you're experienced with LabVIEW, please feel free to jump in and offer advice, hints, tips, etc. Equipment-wise, I'm very well set up. My control computer is a Rohde & Schwarz PSL3 industrial controller, with embedded GPIB and plenty of assorted ports. I'm running LabVIEW 2012 in the trial mode for now. If I can make the system run like I want it to, it's worth buying. CJ |
| Not an expert by any means but I can get around in it. I used it to run my senior project at school but really disliked it. Had some bad experiences with the NI support people too. Had the LabVIEW rep come through our lab and asked him some questions that he either couldn't or didn't want to answer. Just my $.02 |
|
17 year vetran.
Yes, LabVIEW is a great test program. You are tackling a midLevel project (you know this already). 
I learned by going through the Tutorials ( I then went to Austin for the more advanced classes). They used to be in the manual and were well organized. Now they are included in the software. 1. Spend 2-3 weeks going through the tutorials. 2. Spend time with each instrument, one at a time. Learn how each one wants to be talked to (THE MOST DIFFICULT PART OF THIS PROJECT). Let me know when you can make a (virtual) LED flash on the screen at a 1 sec interval. |
|
I'm testing conventional and digital portable and mobile two way radios operating in the VHF (136 to 174 MHz) and UHF (450 to 470 MHz) bands, FM modulated, or for the digital radios, using 4FSK, C4FM, and CQPSK modes. They all carry voice traffic. My list of test equipment to bring to the task is respectable. Two Rohde & Schwarz FSEA 30 spectrum analyzers equipped with tracking generators and the flexible digital modulation analysis option B7, allowing me to look at virtually any digital modulation format and analyze it quickly. Two Agilent E4418B RF power meters with sensors and 30 dB pads for measuring power outputs up to 200 watts Agilent E4406A transmitter test set, which is a specialized spectrum analyzer with additional capabilities, which mostly don't apply to what I'm doing, but the ability to run at a RBW of 0.1 Hz is very nice. I have three of these units but only need one in this system. Agilent E4431B ESG series signal generator which allows me to generate nearly any digital modulation format and pattern that I wish. Anything from BPSK to QAM-256, in addition AM, FM, and PM. Rohde & Schwarz SMY 02 AM/FM/PM signal generator (At one time I owned ten of these but I'm down to three.  )Rohde & Schwarz AMIQ 04 digital modulation generator Rohde & Schwarz PSL3 industrial controller (this is the PC with embedded GPIB, which will control the system and also hosts the WinIQSim software that controls the AMIQ generator.) And, finally, the shop's General Dynamics R2670B service monitor, which would be all that MOST radio shops would ever need for bench test equipment, practically, but I'm engaged in a considerably deeper level of measurement and verification than that. I use this service monitor for routine checkouts, though. When a custom sends radios to me for yearly PM or repair, I don't usually drag out the big guns. But when they're having communications problems that aren't obvious, then it's time to go to war with a quarter million buck's worth of equipment. I have the pleasure of telling the local General Dynamics rep that the new R8000 digital service monitor he loaned to my shop for a few days, bearing a calibration date in June 2012, is significantly out of calibration in digital NXDN test mode, and I can prove it. It's definitely off by about 8 percent in modulation accuracy, and the format spec allows only a 3 percent variance. The task of integrating this system and generating equipment performance reports should be somewhat demanding but well worth it when it's all done. You're probably wondering, where in the hell did I get such a load of hideously expensive test equipment and why do I own all of it personally? (Except the R2670B.) Because I was in the right place at the right time. A local company that makes cellular infrastructure equipment downsized and had surplus test gear to sell. I bought out a great deal of it. Not all at once, but I'd buy a few pieces, sell them for fat profit, and use the profit to buy more. I kept the best equipment for myself. Most of my sales were through ebay. Let's just say that it was a very profitable year for me. I would very much like to have a warehouse full of test equipment to sell like that all the time.  |
|
CMJ,
If you have any C/C++ experience then you will be able to use LabWindows CVi/ easier than LabVIEW. It is another NI product that with TeatStand is about the same price as LabVIEW. TestStand allows you to do repetitive test and reporting fairly easy. LabWindows CVi has the GUI like LV but the coding is C++(ishH. |
|
Funny you should mention that. When that company downsized and got rid of a bunch of stuff, I got a lot of software they were just throwing out. Including a copy of Teststand with the yellow certificate of ownership and serial number. And Matlab. With serial. The company even wrote out a letter transferring "ownership" of these things to me. Since they no longer require most of their license seats, I've been "given" those license seats. I grabbed every bit of software they tossed out that I could lay my hands on. But reportedly, a lot went into the garbage can that I missed.  But I don't have a shred of C or C++ experience. However, I do have MS Visual C++ and Visual Studio, also from the same round of toss-outs. I realize that for me to get to where I want to be, I will have to LEARN. The question now is to define what I need to learn in order to achieve my goals in the most efficient manner, learning process included. CJ |
|
I found out something very interesting about the current version of Labview, version 2012. You can download it and run it in demo mode. If you then set the clock back on your PC, AS FAR AS YOU WANT TO, the software simply calculates the days remaining until the expiration date, and tells you that you have that much time remaining on the demo. "10,000 days remaining" if you set your date back far enough. So, if you're a dastardly bastard, set your PC's date 10 years from now, install Labview, and then reset the date to the right date. You'll have 10 years to play with the software. This is comical. I thought EVERY company had wised up to the clock reset trick on a limited time demo package? CJ |
|
I'm the resident Labview expert at my company and I've somehow managed to write some pretty impressive software using it. Mostly I use it to control test equipment like you described and record data. Best advice is just to start playing with it and going through tutorials and sample programs. You will have to learn the GPIB commands to send to all the equipment to make it do what you want, and then you will have to write the programs to send the commands and run the interface. Start small and create small VIs that do one or two things and have an error in and error out. Wire the small VIs together in a larger VI and connect the error lines so that you can do things in the order you want, otherwise everything on the block diagram executes at the same time and things get messed up.
Be sure to look at the producer/consumer using queues example. That was the biggest eye opener for me when I took a class and it really changed the way I design programs. |