bm_engineering

Yeah, I have thought about that Windows 10 thing also. Unfortunately, I updated all of my computers from 7 & 8.1 to 10, leaving just a couple of old XP machines. While I am not a lover of Microsoft, I must admit that Windows 10 overall has been pretty solid with regards to running all of my controls-related programs (PLC's, HMI's, etc.). And I have had no problems with Visilogic up until now. But fighting to uninstall that Beta made me think that maybe the installation was corrupt. I am just gun shy about reinstalling it on my main computer. Joe, thanks for anticipating my question to them about rolling back the database. Now I will just send the files and ask them to do it. Ben

I hadn't thought to check that at the time. I don't think the serial port was connected to anything, although it goes through a switch box to one of several PLC's (Allen-Bradley & Unitronics). It appears it was in a dead position, but I can't be sure it was always that way. I reverted back to Visilogic 9.8.65 so I can work on some other projects without messing around. Even the BETA uninstall was not easy. I had to use a third party cleaner but it got the job done. That leaves me with two projects that are the newer database that I can not open at the moment. I am going to install the Beta on another computer and play with it some more. I am also going to ask the help desk if they can modify those project files back to the previous database. I would rather wait with ".79" until it is stable. I have already alerted the affected clients. While I enjoy challenges, none of this pays the bills! Thanks to the group for the help.

Good thought. Search did not find the missing file. I just ran another test. Saved my file as "Junk test.vlp". Made a revision. Pressed the save button, and got the error message "Save Project has Failed. Try to save the project under another name. Error number (508)." Saved it as "Junk2 test.vlp". When I looked in the directory, "Junk test.vlp" was gone. The new file was there. Saved it again as "Junk3 test.vlp". No error, and both files then existed. Closed and reopened the Junk3 file and got several error windows in sequence... "Connection is aborted due to timeout or other failure". "The connection is reset by remote side". 1Invalid operation at current state9". But it then opened. When I tried to save it, I got the same warning as previously, saved it as "Junk4 test.vlp", and the Junk3 file disappeared. So the old file disappears. And previously when I saved it as the same file name, that one disappeared. This could be a corrupt file. But it all started when I installed the Beta version. And I did that because of a different file that was sent to me that was database 135, and the previous Visilogic release version was database 134 (working from memory on those numbers) so the file would not open. I have also been working on that file and it has not had any problems. But both of these files are now the newer database so I don't believe I can go back to the previous release version of Visilogic, without reverting to an older file revision and recreating the changes I have made. When I do save the file successfully it appears to run just fine in the panel (V1210), but this disappearing act is not good. I may need to contact Unitronics help desk and see what they can do.

Thanks Joe I agree 100% with you. Beta download should be our option, not the one that looks like the next release version. Strange that I worked on another file that seems fine.

I did not intend to install the Beta version yet, but I was forced into it by a client who sent me a file that the production version would not open. I have experienced a problem when saving another file as I edit it, and I am not sure where to post this. It usually works fine, but every so often, I get an error message saying it can't save it and suggesting I save it as a different name (at the moment I can't create the problem to get the exact wording). When I do that, it appears to work, but when I look in the directory there is no file. What is worse, if I point to an existing file, I get the expected warning that it exists, I say ok, and it apparently deletes it because it is gone when I look for it. I was wondering if anyone else has experienced this. I am running Windows 10 64-bit. The file is one that I created a a few versions back, and have modified along the way with no issues until now. Thanks.

Update: I repeated this using a second computer running Windows 10. The download was created on that computer, from the same Visilogic project file. The results were the same. I also tried a creating and downloading a file from a V570 project, and it worked perfectly. So my conclusions so far are that the download creator and UniDownloader seem to be working fine. The problem is not related to the computer or operating system. The only difference between success and failure appears to be the PLC. The V1210 does not work, the V570 does. While the two applications were obviously also different, the problem occurs during V1210 system file downloads, and in fact the V1210 application download works fine. I also observed that when it fails, the PLC is locked up. You can't even go into boot mode. holding the screen press does nothing. After cycling the power, then it can be restored by reloading the operating system from the SD card.

Not a dumb question at all. I needed to double check, but yes, Visilogic.exe is set to run as administrator (as I recall I had some problems running it until I did that). I routinely update the firmware in my various lab vision units (V350, 430, 570 and 1210) from this computer. I also routinely use SD cards both here and in the field, with files created on this computer. I have not used the UniDownloader very frequently. This PC runs Windows 8.1. I did not try any of this from my laptop with Windows 10, so I think I may do that. Take this computer out of the equation completely.

I have been unable to do a full system and application update on a V1210. I have created several different .UDC files for testing, and regardless of whether it tries to update the binary, the boot, or the O/S, it fails with an error message during the first step of erasing the flash memory. On the HMI, the "status" shows "erasing" with no percent completion. The application file alone updates fine, but not with any one or more of the others. I have run the downloader as administrator. Everything updates properly from the SD card or from Visilogic. Versions are as follows... Visilogic 9.8.31 (build 0); O/S 4.3(05); Boot 2.2(29); Bin Lib 0-3.10(03); Unidownloader 5.0.25 Does anyone have any thoughts on what might be happening? And in a related question, do I need to do this or will the application file update the O/S? My concern is if they update the application onto an older operating system in the panel which is incompatible. The job is in Korea and I am in Chicago, so obviously I need to get this right so the customer can do it without problems. Thanks in advance

I just tried this but I get an error message "Application Error-if this problem continues, contact your distributor". The original works, the one in the copy folder does not. This is Visilogic version 9.8.2 build 0, on Windows 8.1. Any thoughts?

I need to log to SD as many as 256 values (96 will be more common) to each csv record, fairly infrequently. Each value could be 6 or 7 characters, plus date and time. In earlier versions of this application we only had 20-30 values and I used the create CSV and write CSV function blocks which worked well. However, with the larger amount of data now, defining the create CSV function block for that many MI's is a lot of work (which I might need to do), and it seems like a data table doesn't allow enough columns for that big a record. Does anyone have any other suggestions?

If you still have not solved this, disconnect all wiring from the analog output and check the output from the PLC. If you are getting the expected voltage, then you have a wiring problem externally, possibly a short to ground. If you do not get voltage out then the problem is something internal, such as wrong configuration, or you have a bad analog output. I would second the earlier suggestion to move to a different output address and see what happens.

There appears to be a potentially serious bug in Visilogic 9.7 that we encountered this week. Graphics are being corrupted in the clone files that we created and loaded from SD cards. We found various strange things such as objects that were not displaying, and in one case an analog gauge with no background but a pointer that sort of worked but left trails. We have seen this on a V1040 and a V350 so it is not hardware dependent. We verified this on several different computers, etc.The same programs work perfectly with a direct transfer from the PC. However, if you are creating clone files, either with or without the operating system, you better check them. Version 9.5 works fine for us and is our interim solution. Version 9.7 has problems. We did not test version 9.6. We have contacted Unitronics about this and are awaiting their response.

Here are some further test results of multiple coil requests: Simulator (master) to V130, 350 and 570. Limited to 7 coils. TImeouts occur if higher. Simulator (master) to HMI panel brand X. No problems requesting up to the simulator limit of 100 coils. Simulator (master) to PLC brand Y. No problems requesting up to the simulator limit of 100 coils. V570 (master) to V350. No problems (R.C #1, vector length 1900) HMI panel brand X (master) to V570. No problems with 20 tags. However, I have no control over how it acquires them. I believe that because they are sequential it would request multiple coils, but I have not confirmed that. So the only issue appears to be the simulator (ModbusView TCP from Ocean Controls) combined with the Unitronics Vxxx devices. I have no idea why. This is more of a curiosity exercise since I typically pack bits into registers, rather than using coils. I just stumbled onto it running a quick test of some V130 code, to be sure it was communicating.

Thanks. I was wondering if it might be the simulator, although I have never seen that before. I will try another master device and see what happens.

So what does that mean for clients who previously used Visi series, but want to move into the Unistream? Do we need to recreate the program, or is there a way to migrate it? it would seem to be to your benefit to have this ability, which would allow users to move into the newer system.

I was just playing with the V130 Modbus TCP slave communication. I set everything up and connected through my simulator. Reading holding registers worked fine. However, when I switched to reading coils, I got timeouts. With a bit of playing, I discovered that if I asked for 7 or fewer addresses it worked fine. At 8 or more I got timeouts. However, I could shift the address range up by 7 and ask for 7 more and it worked fine. It is not the addresses, just the number of multiple coils requested (one byte?). This is far below the Modbus spec for multiple addresses. I got the same result on a V350. I have not found this documented anywhere, but it appears to be the intended operation. I don't use the Modbus coils very often, so I may have just never noticed it. Have others experienced this limit?

This just happened to me, exactly the same way. Every link in every screen object that did not already have a value in it got the same address (mine was MB1020) It was a result of doing a find and replace to multiple MB operands. I sent it to tech support, and they did clean it up for me. However, they never responded to my question of what happened and how to prevent it again. I have since had some additional, but far less catastrophic, problems with operand find and replace. In one case, an MI got replaced by an MB, definitely NOT what I had specified. I believe there is a major issue with the Visilogic find and replace function.

I found a work-around for images based on your idea, with one modification. Use a binary image switch, based on the "hide" signal. Put the visible image in one of the conditions, and put an .emf image of a solid color in the other. I found that it doesn't even matter what color it is, as long as it is solid! Check the transparent background box. Voila. The solid color is apparently seen as the "background" and made transparent so it becomes the "hidden" condition. The screen background graphic is displaying as it should underneath it. Seems to be working great so far. Not so lucky with text. Still playing with that.

That is good to hear. I am looking forward to playing with one. 2 questions: 1) Does that mean that software development is basically stalled at today's level for the Vision series? 2) How easy is it to migrate a project from Vision to Unistream?

Thanks for that info. I also discovered why the text won't hide. You can't hide text with a transparent background! Unfortunately, my client provided the background graphic and it has color variations and a white "swirl" running through it. So my choices are solid colored boxes behind the objects, client changes the graphic, or your suggestion regarding a list by pointer and try to grab the correct background area. This is not going to be fun. While I understand exactly what you meant by "expected behavior" (by design of the software & operating system), that really depends on your viewpoint. My client expects that those objects will have transparent backgrounds and will disappear, leaving their graphic visible, and without clunky solid boxes. My expectation as a programmer is that if I make something "hidden", I see the next layer under it. Not a hole showing two layers down! And why can't I hide text with a transparent background? As a note to Unitronics, competitive HMI's do not have these restrictions, as well as some others such as the ability to stack multiple dynamic objects on top of each other, a wider variety of color options for 3-D buttons, etc. Hopefully you will consider these things in future upgrades!

Here is a further detail. This screen uses a custom background graphic, so the background color is normally not visible. The binary image object is set to transparent, and not displaying on background color. It displays properly when visible. However, when it is hidden, a box is visible displaying the screen background color. It is as if the object had "burned a hole" through the background graphic.

I have run into an issue on a V1040 project, where screen objects are not hiding correctly. I have confirmed that the "hide" bit is set. Binary image objects do hide, but they are leaving a visible box even though they are set to transparent background. Text objects, more than one, are simply not hiding at all. I just updated the unit to the latest operating system (3.7.00) with no effect. I do this a lot on the V570 with no problem at all. Any thoughts would be appreciated.

I have just been noticing this recently using Version 9.5.0 build 0 on windows XP, on a project with a V1040 connected by USB . Turning off the PLC or disconnecting the cable while monitoring results in a fatal error message, and then the program crashes. It happens every time. Even worse, the Visilogic.exe process remains running and cannot be manually stopped, so you can't restart Visilogic. On several trials, it cleared itself after about a minute. However, on one occasion I could not restart the program and I could not stop the running process. Rebooting the computer was the only solution. Attached graphics show the error messages. I don't have earlier program versions running so I can't verify whether this happened on them and I just never noticed, however that is easy enough to determine. I also don't know if it happens with serial or Ethernet as well. Regardless, this is a bit of an extreme reaction. As a suggestion, a far more elegant approach would be a warning message, and then just dropping out of on line test mode. This is what happens if you try to initiate it with no connection. Just a warning message.

OK, I have learned to live with the lack of a usable undo function, and the fairly primitive cut and paste capabilities on the HMI screens. However, today I went to send a screen shot to a client, and luckily I looked at the BMP file before I sent it. There, stretched across the center of the screen, was the Windows "file already exists, do you want to replace it ?" dialog box. Really? Why not capture the screen when the export button is first pressed? The mouse pointer will be off of the screen, and there won't be any dialog boxes. Sorry to vent, but seriously, as good as VisiLogic is, it could be much better if it integrated more Windows functionality into it. Other programming software does, so the ability is there. Paste objects anywhere on the screen, without moving the original. Undo every action, especially an accidental movement of a screen object (something that virtually every Windows program can do). Etc. These seem like small things, but they are time wasters that add up, and time is money when you are programming.

I have been playing with a handful of micro SD cards that I had lying around. They were all formatted using SD Card Suite. I tried them in the V350, and also in SD adapters in the V570. My very limited results to date are that 4GB Kingston and 2GB Transcend cards worked great. However, the 2GB Sandisk cards did not work. The PLC simply does not recognize the card. They work fine in a PC card reader. Does anyone have either a formal recommendation or informal information about which cards work well?