Jump to content

Ausman

MVP 2023
  • Posts

    2,563
  • Joined

  • Last visited

  • Days Won

    174

Everything posted by Ausman

  1. I'm not exactly sure on what you're asking, but the software way of doing this is using SD Card Explorer, which is part of the suite found here: https://downloads.unitronicsplc.com/Sites/plc/Visilogic/Download Software Utilities/Unitronics-SD-Card-Suite-2.0.70.zip cheers, Aus
  2. So...just a follow up on this odd behaviour. I discovered a topic already existed on the forum here: https://forum.unitronics.com/topic/1747-visilogic-94-invalid-bands-collection-index/ There is mention of deleting the folder in C:\Users\YOUR_USER_NAME\AppData\Local\VirtualStore\Program Files (x86)\Unitronics, but in my case the only entries in this folder relate to Remote Access. So I tried the initial suggestion which relates to the .lyt file which is in C:\Program Files (x86)\Unitronics\Unitronics VisiLogic_C2\Main\Layout. (Note the _C2 added for the instance in question) Added old to the name for safe keeping and copied the same file over from one of other instances. I didn't use the one Cara has a link to as the size mismatched significantly. The faulty instance booted up ok, but needed direction to the project that I was initially using. Once there everything is fine. I haven't had any other errors appear (so far.... fingers crossed) like others mentioned in the earlier topic. Of special interest is what I discovered about the .lyt file upon closer inspection. Opening the good file in Notepad++ brings up multiple lines of code up to 1672, but opening the dud one and it only has ONE line of code that is all NULS. So somewhere in proceedings the file has obviously been significantly corrupted. I can't really explain this, I don't believe I bit the byte gods that day (readers can now laugh at my little pun), and I don't recall any other odd things going on. Both drives in the puter in question are good name, young SSDs and smart shows no errors or reallocations have taken place since new. As already mentioned, caching is disabled as well. Something to ponder, but not too long for, unless it happens again. cheers, Aus
  3. All fair enough, but doesn't your explanation mean that it is on for a number of scans matched to the length of button push? Which is what Joel is doing .......... And that was why I suggested the possibility of repeats each scan messing something up. cheers, Aus
  4. Sorry for delay, weather events knocked systems out. Further explanation. I don't use unilogic, but in visilogic such a button press utilising a set and reset based on the button push would keep the bit on for quite a few scans, instead of just one. The action you are trying to initiate might then perhaps be unable to operate properly because it is being called in consecutive scans, rather than the single call it needs. cheers, Aus
  5. Thanks Flex, I already know this, but I am going to do it progressively as previously described to try to narrow it down a bit. That's why I first said "easily fixed". I have the same system of virgin copies. Just annoying that something so stable suddenly throws a wobbly for no apparent reason. cheers, Aus
  6. Thanks, but the project opens fine in the other instances I have. In the faulty one it doesn't get to run the window at all. But it does leave Visi as a running program that I have to force close to try again. The "This application is already running" alert. The real puzzle is how a fully closed program was affected by a sleep instance. I don't have write caching enabled. Visilogic was well and truly closed, done ages before the sleep. A full drive scan of both drives involved shows no errors. Regarding doing a repair, I can't do it because the instance is the "2nd one" and the "1st" is where the install/repair tries to go. I can't choose destination during the process. What I am going to do (for interest's sake) is spend a little time replacing particular folders in the faulty instance until the error goes, and then try to narrow it down progressively. Unless someone can pinpoint the exact file and trigger for me. The one thing I have found with a detailed look is that Windoze has at some stage once again turned on a host of "features" that I disable. It is possible that one of these damn programs has upset something during the sleep. I reiterate that everyone should have a look at what task scheduler does in the background. You will likely be a bit amazed. It is plain as day that if a huge number of people take steps to disable a method, then M$oft will work their butts off to do it another way, under another name. I'm not paranoid, just careful, and try to limit my "internet exposure". But it is a bit hard when an O/S I am forced to use, and have paid for, is so aware of everything I do and tells their data miners all about it, even though I constantly try to stop such things. cheers, Aus
  7. Never seen this one before. It appeared during loading Visilogic after a wake the puter from a sleep, which I don't do too often. It's only in one of my instances, so should be easily fixed, but it would be nice to know what caused it and what particular file is corrupted somehow. I'm guessing it's got nothing to do with a number of musicians playing outside their contracts....! cheers, Aus
  8. So...putting a hypen in it is the trick. Who'd have thunk that will now make it acceptable. Does that now mean that we have a new word "tard" which actually means advanced/quicken? I'd be pretty gruntled at that one! cheers, Aus
  9. A topic on the forum got me thinking again on the PC sillyness around. We're heading to not being able to use Male or Female as a name for blokes or girlies. So I ponder the discussions that are going to go on at my refrigeration supply shop.... "Hi, I'd like a piece of brass with a 1/2" pointy-out thread on one end and a goey-in thread on the other end. The pointy-out end has to have 45° angles that match up to another piece of brass that will have a goey-in thread and angle matching the 45° of the other goey-out bit. The other bit of that piece will run through to a pointy-out guide stub that will have holes drilled through it around the pointy-out end." "Huh???" goes the counter staff. 🤔 Instead of "I'd like a 1/2" MFL x FFL union with frost nut please". "Sure, Aus". Seriously, in Orstralia this is going to create so much fun and laughter in all the traditionally "blokey" trades that always use such stuff. And believe you me, the growing number of girls that are taking up such trades will be in on the laughter as well. Just ridiculous. The worst bit is that someone somewhere gets paid enormous sums to come up with such silly concepts that somehow filter through to laws. And incidentally, regarding the original "retarded" word that started this for me, I squizzed at a few online dictionaries and there was absolutely no mention of retarded being in common use for engines or indeed anything that uses precise timing for operations. As in "retarded/advanced" timing. Therefore, wokeness in the organisation has completely focussed on the "derogatory" aspect completely ignoring another common usage. I guess they're bowing to pressure to eliminate the word altogether by stealth. Good thing I've got old paper printed dictionaries to prove such things to "influenced" youngies who can't think for themselves. cheers, Aus
  10. A one scan set? Or the number of scans between the physical actions? This might be the issue, in that you're asking over many scans for an action that should only be asked once.
  11. Hi Joel, so the obvious question is what model, what firmware and what version of Unilogic are you now using? Different from your first two? If your previous program worked fine, don't change anything unless it is absolutely necessary. cheers, Aus
  12. Thanks for the thanks! If the old system does come out, don't dispose of it. Some members here might like it for historical value. cheers, Aus
  13. I don't use Unistream and in reading your posts you are way above my pay grade! However, one simplistic thought...... zipping the files and putting them on a specific users' read only folder on a NAS or similar would likely do it. Whether you can do this automatically on the plc I don't know, but you could certainly do it externally. But any sort of data is open to manipulation by determined users.....the same way machine users endlessly try to get around safety systems to make their job "easier". 😞 cheers, Aus.
  14. Hi Tozo, my initial thought is to check file permissions. If you are using C drive for your interactions, Windoze has added all sorts of odd restrictions to C access over the last few years, and perhaps this odd behaviour is a result of such restrictions. Check the restrictions present on the new files, use them from somewhere else etc. But it also may be a program quirk that others will chime in on......... cheers, Aus
  15. +1 to all the other suggestions. So it is asked every second, but will go to 9 seconds if there is an error, thus mucking up the entire procedure. You MUST allow for all possible timing TOTALS in arranging your calls. Your calling methods and intervals can include using SBs to check whether calls were done ok. Even then I'd allow a few scans to eliminate potential buffer issues. I do 1 retry at most, but many here don't even do a retry...if the thing is failing it will do it again and be noticeable, and needs to be fixed. It all comes back to how critical comms speed is to machine operation. For my stuff I do fixed speed calling to at most perhaps 20 devices in a repeating loop, with the interval between each call slightly longer than the total possible time that will occur if there is an error on the previous call. cheers, Aus
  16. Sooo annoying, but so easy to do and miss in checks. My method after years of doing this is an excel sheet that lists all the I/O and other connection types, locations, cable colour and ID number, and is changed to background of green by screen, or physical mark on a printed version, on final check before power up. Including checking connection direction of clamping diodes if not using back to back types. And...even then I've been occasionally caught out by unlabelled terminals where you're looking at a printed description on the body some way from the terminal, particularly in cramped locations. Count terminals along from the end but still make an error....poop. Double checking is a must. cheers, Aus
  17. 1). Using the twisted pair from an ethernet line.......Is it separated completely and just sitting there as a twisted pair in open space? Or is using a twisted pair alongside the other 3 pairs? I have found through experience that you MUST use dedicated 485 shielded twisted, as either single pair for data only, or 2 pair for data and power. 2). Allowing the cable limitation. I can't quickly see how often you are polling. I think this is likely your main issue. How many retries are you doing? If anything faults on the line, it will do a retry and if this all overlaps into the next ask, the system will fail. It looks like you are polling every second, but your timeout is 2. 3). Speed. I would only be going with 9600 on such a length. Don't forget 485 can't move a lot of data quickly. Ensure you don't have ID clashes, change everything to the same speed and settings, and see how it goes. Perhaps in trying different speeds you didn't change everything correctly. If you do get stable conditions on 9600, you can then try to go up incremently to find an optimum. Baby steps first...teach it to eat, burp and fart. 4). Ensure that all the separated systems are deriving their control power from the same phase, and share common earthing. Ensure the (new) cable shield (once fitted) is correctly earthed. A few things to check/try. cheers, Aus
  18. But I think what Zebioo is saying is that although the above detail says max 24, the previous Visi version in use didn't bring it up as a compile error, but the new one does. Perhaps the restriction IS "artificial", but if that is what the makers have specified as the max, then that is what the programmer should implement. cheers, Aus
  19. If you're really patient you could try changing the hardware code +1 each time, but not include numbers already in use. And take aim at more likely numbers. I'd also have a "virgin" copy of the program to always start again with, as having a wrong number might permanently upset something. Doing this should get you there in the end, and actually might not take that long. You will perhaps need a soft brick wall for you to bang your head against whilst doing it, and some good coffee. And another but.....but only if the code is actually able to be recognised by the main program, which may not be there anyway. I guess you've already looked, but is there any way to reference the existing codes to some other accessible part of the software? As Joe is vaguely hinting at, perhaps you need to stop all attempts and put in something newer. But if you have a few of these things, and they are working fine, then I do see why persevering is worth it. One of my favourite sayings is that if it ain't broke, don't fix it. I have some DOS programmed PLCs still running fine, many many years after first install. As a matter of historical value, how big is the software? Perhaps it should go up here on the forum for posterity? Also out of curiosity, does the program start itself (via a bat perhaps) into cmd, or do you have to do it yourself. The reason for the ask is that perhaps it wouldn't hurt for it to be run as admin. Without knowing the old capabilities, perhaps an upload PLC to PC is lockable on this system. And yes......Unitronics, like many other firms, does name communication direction descriptions differently. As in, you download something from the web to your browser/PC. But download in Visilogic is send it to the PLC from the PC. So the description applies from the PLC's point of view. cheers, Aus
  20. It might be worth trying to see if the filename, machine model no, or manufacturer name is the password. I've seen this very occasionally on some machines. Seems dumb but in some ways is clever. Is info mode accessible and not passworded? If you can get into Info, you can find the project's name to then try. How complex is the machine? If it is not too sophisticated you are likely much better off finding a local programmer to look at what you need and redo the program for you, with tweaks you know would be useful after having it for a while. If very complex I'd still be trying to source the original project somehow. cheers, Aus
  21. Look at point 2) in this topic: https://forum.unitronics.com/topic/6229-ongoing-24-topics-almost-worth-pinning/ I'm not sure if the EXF shares the issue, but it is a solution. Others can chime in if there is something better. cheers, Aus
  22. Have a read of this topic. There is a vague chance that the way the software you have is arranged in a similar fashion, and the topic's answers might offer a solution for you to try. https://forum.unitronics.com/topic/3632-visiologic-ver-9760-build-0-v350-35-tu24/#comment-36546 cheers, Aus
  23. You are likely in big trouble. I'm assuming from what you are saying that with a new battery the 1210 is running it's program, but variable parameters crucial to operations, that you can't enter/edit via screens have been lost. My first suggestion would be to search across all computers in your organisation to see if you have a *.vlp file anywhere. This is the program in the 1210 and may have been originally supplied to you, but I doubt it. Also, is there an SD card in the 1210? If so, what files are on it? Can't even find any old employees of the company who would know? Other firms in your line of business using the same machine might know?
  24. All of this information and a lot more is in the Help files.
×
×
  • Create New...