Jump to content

Gan Uesli Starling

  • Content Count

  • Joined

  • Last visited

Everything posted by Gan Uesli Starling

  1. We have a gage supplied by a company that shipped it with a *.exe application targeted for LVRTE 2009. I need to retarget it for 2017, but don't have the source code. The supplier had said they'd gladly supply me with a copy of the *.LV source, but they have looked and cannot find their own copy in-house. History of Need: Our global corporate mother ship's IT department, in their infinite wisdom, is mandating an upgrade from Win7 to Win10. That with yet even further constraints. They enforce a list of "approved versions" of "approved applications". And for LVRTE, they are insisting upon 2017, with 2009 being a red light. So, then, my query. Is converting an app without the source for a higher LVRTE doable at all? File is attached. If it is doable, instructions on how? Concentricity-Gage.exe
  2. There were some things I was wanting to do in my ham radio hobby for which I had no ambition to write a GUI in Perl. And so, now there is this: KY8D.net/free
  3. It's a political thing. In my own department (Test Engineering) we are fully invested with LabVIEW. The test stand in question, however, belongs to another department. And they are devoutly enamored of Allen Bradley. Cost is no issue, my company has full, all-encompassing corporate licenses for both NI and AB softwares. I am considered something of the plant-wide hardware and software guru, if only because I've been here 12 years. But this project got handed to me already with the AB constraint. Bit of an uphill slog in my future, so it would seem: new hardware & a new authoring platform.
  4. I am assigned to refurbish an airflow instrument having six pressure sensors, four temp sensors, and six on/off outputs. By preference I would use all NI hardware, but this isn't going to be allowed. I'm being pushed toward installing an Allen Bradley PLC instead. I am aware of another, much more complicated liquid flow test stand which, so I'm informed, uses LabVIEW for SCADA on a Seimens PLC. This being the case, cannot I do likewise with Allen Bradley? Can I do it entirely in LabVIEW? Or is something like that always just LabVIEW sending trigger commands and receiving data from a free-standing program written in the PLC's own native SCADA? I'll be starting from scratch, with nothing yet purchased. I can purchase whatever I choose. I have perused a couple of PDFs of Allen Bradley ControlLogix programs, and at first glance, to me they look like a major pain. Unlike LabVIEW, almost nothing shows on any one screen. Nothing at all looked to have been nested into a subroutine. I liked ladder and highway diagrams quite well enough back in the 90's when printed out on D-size vellum. Then, at least, I could stand back and see the whole thing. The same thing seen only through a tiny window that you have to scroll up and down I'm not looking forward to learning at all. Thus my hope for a LabVIEW solution, rather than purchase and learn AB's Studio 5000. I'm an old dog, and this looks like a new trick to me.
  5. There are LabVIEW Drivers located here... http://thermotron.com/pdf/_service/labview7/TH88-8200.zip ...which I've downloaded and installed. But first I must make the initial connection, which is where I still suffer confusion for where to start. Were it RS-485, I'd just set up COM3 with one of my several USB-to-Serial dongles. I have such a dongle on order for GPIB to allow connection to the IEE-488 connector, but it will take some days to arrive. Meanwhile the RJ-45 beckons, so I'd like to firstly try that. Also with regard to the forthcoming USB-to-GPIB dongle, a page at NI located here... http://www.ni.com/getting-started/set-up-hardware/instrument-control/gpib-usb ...held forth an empty promise of info relative to Windows which it says is located here... http://zone.ni.com/wv/app/doc/p/id/wv-1648/ ...but that one is a dead link.
  6. I have a Thermotron test chamber with an 8200 controller to which I'd like for LabVIEW to talk. Basically, I only just need for LabVIEW to know the current set point of the Thermotron for whatever internal program it might be currently running. Only just that. Just so LabVIEW knows what it's trying to do at any particular moment. So LabVIEW can ask and get an answer. Our 8200 controller has TCP/IP and GPIB. But I've never yet done ought with either of those, only just RS-485 up until now. So I'm not sure just where to start. Anyone out there have helpful clues?
  7. On my Front Panel I have mostly buttons. Buttons and a single Input box for typing serial number. That single Input box feels like a stumbling block to users who habitually TAB their way from widget to widget. They stumble because, after typing the serial number into the box, they can't TAB their way out. Instead (of course, to my way of thinking) the Input box accepts the TAB as a character (rather than interpret it as a command). It bothers some (unreasonably, to my way of thinking) that they should have to press Enter, rather than TAB, to exit that box. What they want is to press TAB and have the data be entered, plus take them to the next widget (a Button). So my question here is, can I make that happen for them? Can I make TAB perform like Enter while inside an Input box? Or should I just insist they get used to Input boxes requiring Enter?
  8. SOLVED! My issues are now resolved. How we did was: Via Windows Control Panel, first removed all NI Software. Said software having been previously installed from an in-house source (corporate IT hosted out-of-state server). Removal of which took many hours. Install LabVIEW 2013 afresh from CDROM, electing “Evaluation only…” as the license option. Update the NI Package Manager when so prompted. Opted for only URGENT and LabVIEW 2013 related updates. Ran, ex-post-facto, the in-house "vlmclient.exe" to associate our in-house license. That went WAY more smoothly than (as our own corporate IT had set up) installing LabVIEW from "setup.exe" files hosted through our intranet on corporate servers out-of-state. Post this fix, on running LabVIEW, all seems to work as it should. Nothing comes up missing. We get no permissions errors. All is well for plural users. I assume that our in-house license server took as, upon starting, we do not get a pop-up about how many days left in the evaluation period.
  9. Okay, so I tried those and it still didn't help. A full-reinstall, this time carefully choosing the machine-based (aka "all users") version from our company's (far away) network server. Still the same issues, no matter what. Other users could open LabVIEW and even write their own tiny LabVIEW test routines and have them work ... but only so long as their VI did not include the Three Button Dialog. For other users different from me, the Three Button Dialog (and a few others) were missing from the pallet. Only for others, never for me. So I decided nest to "Remove all NI Software" via the Control Panel before attempting yet another re-install. And that removal took many hours. I began it at 3PM and by 5PM it was only barely beginning ... on a removal! I came back at 7PM and it was just over half. I went to dinner and came back well after 8PM and the removal thermometer was only up to two-thirds. By morning the removal was done. But now I am shy of the whole company-network installation process. Know that all of our prior installs of LabVIEW via the company server (in a state far away) have taken all night. And sometimes we'd come in to find the install had failed. Has anyone a similar experience? Or does anyone know of a way to do LabVIEW installs via CDROM and then, ex-post-facto, make a configuration change so as to get user-rights from their corporate license by setting a configuration file? Anything which won't take all night? Thanks, Gan
  10. Our company has a company-wide network license server for LabVIEW. I have a machine where I installed LabVIEW and everything works dandy there. But nobody else can use it at all. I've already been through plural rounds of local IT shooting in the dark: try this, try that, try something else, do the first thing again. It's all been no use. So does anyone know where I can delve into my own WORKING connection for LabVIEW to its networked license server to learn what it is that my own account does right? A register someplace, a configuration menu, a hidden *.ini file somewhere? Anything like that to tell me how my own, working LabVIEW gets along fine. Then I'd have some positive info for knowing how to fix the license connection for others. Attached is the error message I get when trying to fire up LabVIEW from a different account on the same machine. It finds all the VIs which I wrote just fine. What's failing to load, so I surmise, is some other, more basic part of LabVIEW itself. TIA, Gan Uesli Starling, KY8D Test Enineer
  11. Version 1.0.0


    Test stand had a touch screen. User wanted radio buttons instead of a pull-down menu. Wanted tests to run on pressing the button and pass/fail indication thereof. So I did that for him. Here is the stand-alone proof-of-concept that I worked up for it. Arrays are used so Boolean texts can change to suit different models of unit under test. If a Boolean text is blank, then that button will go invisible meanwhile, there being as many (up to 7) or as few (down to one) as may need be. I should mention, in case it's not obvious, that when in actual use, the three arrays are hidden inside of a sub-VI and not (as shown here) visible to users as part of the front panel.
  12. View File Example Custom Radio Buttons Test stand had a touch screen. User wanted radio buttons instead of a pull-down menu. Wanted tests to run on pressing the button and pass/fail indication thereof. So I did that for him. Here is the stand-alone proof-of-concept that I worked up for it. Arrays are used so Boolean texts can change to suit different models of unit under test. If a Boolean text is blank, then that button will go invisible meanwhile, there being as many (up to 7) or as few (down to one) as may need be. I should mention, in case it's not obvious, that when in actual use, the three arrays are hidden inside of a sub-VI and not (as shown here) visible to users as part of the front panel. Submitter Gan Uesli Starling Submitted 02/24/2017 Category *Uncertified* LabVIEW Version  
  13. So I have a multi-tab GUI. But for some reason, on load before run, it doesn't display the tab that was showing on save. It shows a tab which I don't want to be showing instead. On run, it starts with the right tab. Then saving and closing with that tab showing, on re-open (before run) it still shows a tab I don't want. How can I influence that? On top is the tab I want to show on open before run. At bottom is the tab that shows instead. You see my problem...
  14. Okay. That was indeed the issue. I figured out starting the tasks outside the loop and that helped a LOT. Thanks.
  15. Yes, I am starting and stopping the tasks every time on each iteration of the loop. Such is the examples in the books I have seen. So I've always just done that. Should I instead start them outside the loop, then carry them through round-and-round, stopping only outside the loop? Not really sure what that would look like.
  16. I have a LabVIEW program, which when operating with simulated input data is blazing fast during mock-up. As soon as I hooked into a real cDAQ on Windows, however, the update speed fell off to 4 seconds. That's awful slow. I don't really have a whole lot of channels. I have read tasks for Volts of maybe 5 channels, write tasks for both On./Off for maybe 10 channels (simultaneous) and write tasks for 4-20mA of only 5 channels. How may I improve on that? Is it maybe a priority issue in Windows or to do with USB? What? How to improve? Thanks in advance.
  17. So I'm calibrating a NI-9205 as voltage on a range of 2-10 VDC (the differential of 4-20mA through 500R). And one channel is behaving oddly. It clips at 42% of its 50 PSI scale, but only while measuring via scale, not while measuring volts. I'm setting pressure via a dead weight standard, pressing air into a Specter sensor which outputs 4-20mA. I'm measuring it as V across 500R because I need to filter it with a Sallen-Key active filter (a retro-fit to eliminate pump ripple from out of the signal). So, anyhow, Ch 0 on the NI-9205 is fine in all ranges from 2 to 10 VDC while measuring volts. Let me turn on the scale, however, and then it's nice and linear right up until 21 PSI where it clips. The proper voltage is there. I am double checking it with a Fluke DMM. I can add weights on the dead weight standard right up to 50 PSI, and while set to show Volts and not scale, then NI Max shows the right volts. At each PSI level, I can switch back and forth between Volts and scale on NI Max and both are good right up to 20 PSI. Above that, however, and even though Volts keeps on climbing, scale clips and holds a t21.345388 PSI. I have tried remaking the scale. I have tried that same channel with other scales. Those other scales too (all for ranges of 2-10 VDC but different PSI ranges) all clip at 40% of their full scale on this channel only. I have swapped out the NI-9205 and the new one does just the same. Any ideas?
  18. Hmm... The Waveform Charts can display as "stacked plots". There is a checkbox for that in the front panel as an option. It is very like a digital oscilloscope, yes? And in my Yokogawa DL850 scope (and others) you can divide the screen into sections (or plots, stacks, whatever may be the most correct term). And in each of those sections (on the DL750 and others) you can have more than one trace. So, for instance, I could be viewing Temperature in one section, with all my temp data there, Pressure in another with my transducer data there, and Flow in another with all my coriolis meter outputs there. Similar types of data, in similar ranges, all nice and tidy. All on one device (for a scope). I was hoping for the Waveform chart to emulate this behavior. But if need be, I can call separate Waveform charts, one for each type of data. It just seemed less tidy, plus a bit more bother for having to deal with plural references (versus one) to thread the controls down through sub-sub-sub-routines via cluster. Sorry this is taking up so much of your time. It seemed a simple concept to start. Perhaps I was wrong.
  19. Thank you for that. However, it seems I explained poorly. So I beg pardon. I should have said, upper and lower "stack", each with 3 lines, both stacks on the same chart. This as opposed to needing two separate charts, thus to keep the two simultaneous. Two, because with auto-scale in Y on the same non-stacked chart, if PSI is in the hundreds while PPH is in the teens, the PPH will be sorely squished down. Your otherwise very fine example is showing six lines in a non-stacked chart. Playing around I somehow managed, quite accidentally, a 2-stack plot with one line above and two lines below. But I've no idea how that occurred and have not been able to reproduce it.
  20. So, to paint a picture... In the TOP plot, three lines corresponding to Pressure in PSI: The ACTUAL pressure as a solid white line, read from a transducer. Above that (hopefully) a dashed red line corresponding to Do-Not-Exceed value for the present circumstance (a constant). Below that (hopefully) a dashed red line corresponding to Fall-Not-Below value (another constant). In the BOTTOM plot, three lines corresponding to Mass Flow in PPH, and that with it's own set of three lines, like for above. Now, the two plots are simultaneous, and of significantly different magnitudes, so that one would end up squished small if on the same Y axis. Hence two plots. Now also, the waveform chart would be zeroed out between changes of circumstance. That is to say, each circumstance would be one step in a staircase of values across the range, some very small, others quite large. Thus the tolerance for max & min, plus the value itself, would be squished down from auto-scaling once higher values passed through the view. So I break up the display into circumstances of one stair step each, with each lasting a minute or so.
  21. Granted that I could do this very thing with two single-stack charts of 3 traces each. I could do that... However... What I'd very much prefer is to have 1 chart with 2 stacks and 3 traces on each. So my question is...Can I do that? And if so, how can I do that? I'll tell you what this is for... I have a test stand measuring pressure and flow. I want to show the current value for each in a separate, vertically stacked plot, with also the current max and min limits as dotted red lines. The settings are progressive steps, each with a different max and min for both those values. I want the operator to not have to think, only just look at the cart, and to know in an instant if it is performing amiss. Also an aid to troubleshooting. Save me a lot of headaches and interruptions down the road.
  22. Very informative and to the point. Thank you. I switched from \\Foo\ to \\ plus dotted quad, and it works fine now.
  23. Attached is my VI for writing a tiny HTML file to a intranet network directory. This is on Windows for writing to a file path like \\foo\bar\wtf.html (fictional!). But when the network is slow (quite often) the file write will fail. I've got a patch in there so survive the error and just wait till next time. But what really I need is a way to tell the write to hang in there longer and not give up just because there's no reply yet. My temporary work-around is to open a Windows file browser to that path, the twiddle my thumbs until the file list populates. Then, the path being established, I may safely start the parent VI of which this VI is a sub. Any ideas? Write_HTML.vi
  24. Can someone example a VI for obtaining the current X & Y screen position in pixels for the mouse? This is for a little widget I'd like to make to assist me in judging sizes of objects from a photograph. I'm going with LabVIEW instead of Perl, etc, just to save myself from having to mess with writing a GUI. So the idea would be this, In a photograph I know the size of one item and want to derive the approximate size of others by comparing their relative lengths (ignoring parallax) and just doing the trig. I have that working already but only in combination with GIMP while eyeball reading and manually punching in each separate X & Y, which is a pain. By contrast I could do it in Perl, but then I'd have to write a GUI, which in Perl I've not done for years.
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.