Skip to Content

Register Login

Industry Online Support

Technical Forum

Navigation

CFC/D7-SYS V8.0 UPGRADE - The Hidden DANGER - BEWARE!!

Created by: LA_COLD at: 11/21/2013 11:20 PM (13 Replies)

Rating (3)

Thanks 5

Result pages:: |; 1 | 2; |

14 Entries

Entries per page: 10 | all

11/21/2013 11:20 PM	Rate (0)
LA_COLD Posts: 24 Rating: (8)	We have recently upgraded our CFC & D7-SYS to the latest version (V8.0) so that we could run on Windows 7. All went well and FM458 came right back up and ran well. However, our process had a number of issues and it took weeks to find out what happened. History: Once our programming station was upgraded to the latest SIMATC software, we opened theold Step 7 Project. We then opened a chart inCFC and it wanted to convert the old charts from V7.0 to V8.0. We said yes, andconversion finished. Then the latest D7-Sys V8.0 Libraries had to be imported for CFC before it would compile. We imported the libraries using the "Block Types" option from the CFC menu. All seemed to go fine, FM458 Downloaded andstarted to run again.We had some process issues since the upgrade, and just recently we found out that during the import of the FBSLIB the "MUL" (Multiply) block initializes all "0.0" inputs to "1.0". All the old code that Siemens commissioned and left in our processors "disabled" by multiply the result by 0.0, got re-enabled when the new libraries were imported!!!!!! I have repeated this many times to prove it. This blog is mainly a warning to EVERYONE considering upgrading. Siemens will rewrite your code in the process. If you have any "MUL" blocks with 0.0 on any of their inputs,connect them to a NOP with 0.0 on its input or delete them since they're not doing anything. Andyes I had to look at everyone ofthem all through the CFC cross Reference tool. Also, if anyone knows of any other blocks that do this please let me know so I can fix those too. Thanks,
LA_COLD Posts: 24 Rating: (8)
Suggestion To thank Quote Answer

11/22/2013 6:07 AM	Rate (1)
KSN Platinum Expert Joined: 8/25/2006 Last visit: 11/14/2024 Posts: 2211 Rating: (280)	Hello. It looks like you have upgraded from D7-SYS version lower than V7.0, because the default value (1 instead of 0) at MUL inputs was changed in D7-SYS V7.0. This is described in 'D7-SYS - Release History' (Start-Simatic-Information-English), did you read this file? Function blocks MUL, MUL_I and MUL_D Default value at input X has been changed from 0 to 1.

Suggestion To thank Quote Answer
This contribution was helpful to 1 thankful Users LA_COLD

11/22/2013 1:17 PM	Rate (0)
LA_COLD Posts: 24 Rating: (8)	KSN, Thank you for pointing this out, but the file only says exactlywhat you have pasted.Maybe it should also say "THIS WILL CHANGE YOUR EXISTING CODE DURING AN UPDATE!!" We did go from D7_SYS V6.2 to V8.0. The "D7-SYS important Notes" in the same directory also says: The Default Value for MUL_x blocks at generic inputs has been changed from 0 to 1 (up from V7.0). I guess this would be okay if a simple library update did not cause an existing block to change state. On a new insertion, this is OK, on existing code, this has beena disaster. It seems like you have some experience with this update, is there anything else that may cause us issues? Thank You for taking the time to comment.
LA_COLD Posts: 24 Rating: (8)
Suggestion To thank Quote Answer
This contribution was helpful to 1 thankful Users KSN

11/22/2013 2:10 PM	Rate (1)
KSN Platinum Expert Joined: 8/25/2006 Last visit: 11/14/2024 Posts: 2211 Rating: (280)	LA_COLD Maybe it should also say "THIS WILL CHANGE YOUR EXISTING CODE DURING AN UPDATE!!" Hello. May be you are right, this will be a good warning for those who wants to upgrade. LA_COLD I guess this would be okay if a simple library update did not cause an existing block to change state. On a new insertion, this is OK, on existing code, this has beena disaster. I can't imagine why you need to use MUL-FB with 0 at its input, when you know that its output will be 0, and if its output is connected to some other blocks, it seems it is easier to enter 0 at inputs of other blocks or to connect output of NOP with 0 at its input. LA_COLD It seems like you have some experience with this update, is there anything else that may cause us issues? No, I only always read the 'Release History' document.

Suggestion To thank Quote Answer

11/22/2013 2:17 PM	Rate (0)
LA_COLD Posts: 24 Rating: (8)	Quote: I can't imagine why you need to use MUL-FB with 0 at its input, when you know that its output will be 0, and if its output is connected to some other blocks, it seems it is easier to enter 0 at inputs of other blocks or to connect output of NOP with 0 at its input. KSN I agree with you 100%. Multiply by zero is useless, but this is the code that Siemens left us after they commissioned the project. I have cleaned up a lot of their logic, like this, but must have missed some. Its hard to go through it all. Thanks for your help.
LA_COLD Posts: 24 Rating: (8)
Suggestion To thank Quote Answer

11/22/2013 2:48 PM	Rate (0)
KSN Platinum Expert Joined: 8/25/2006 Last visit: 11/14/2024 Posts: 2211 Rating: (280)	LA_COLD this is the code that Siemens left us after they commissioned the project. I have cleaned up a lot of their logic, like this, but must have missed some. Its hard to go through it all. It seems this is old project. Do you know the 'Chart Reference Data' via which you could quickly find, e.g. all MUL blocks in a simple way? Regarding changes at MUL inputs after upgrading I can say onlyRTFM. I think after upgrading libraries you anyway should check that the program is working properly. LA_COLD I have repeated this many times to prove it. And don't be so angry, I think you are familiar with D7-SYS and other Siemens software not first year.
	Last edited by: KSN at: 11/22/2013 3:11 PM
Suggestion To thank Quote Answer

11/22/2013 4:04 PM	Rate (0)
LA_COLD Posts: 24 Rating: (8)	I agree with what you say, and I did RTFM but, "Default State from 0 to 1" does not exactly translate to: "Any Multiply block already installed in your program with a Mathematically Valid Input of 0 will be changed to 1, please check all blocks ahead of time for inputs of 0". Just saying . . . Since the block update process can take every other valid input: 1.0, 0.01, 3.14, etc. and maintain it on the input connector, why can't it do it with 0.0 also? It's not a new insertion with default values, it's an update that should maintain current values. I'll stop now.
LA_COLD Posts: 24 Rating: (8)
Suggestion To thank Quote Answer

11/22/2013 4:23 PM	Rate (0)
KSN Platinum Expert Joined: 8/25/2006 Last visit: 11/14/2024 Posts: 2211 Rating: (280)	LA_COLD block update process can take every other valid input: 1.0, 0.01, 3.14, etc. and maintain it on the input connector, why can't it do it with 0.0 also? I also thought about what will happen with the MUL's during upgrading that have non-zero at their inputs, but I could check this only at work on monday. Do you mean that non-zero inputs retain its values after updating?

Suggestion To thank Quote Answer
This contribution was helpful to 1 thankful Users ggbg

11/22/2013 4:41 PM	Rate (0)
LA_COLD Posts: 24 Rating: (8)	KSN LA_COLD block update process can take every other valid input: 1.0, 0.01, 3.14, etc. and maintain it on the input connector, why can't it do it with 0.0 also? I also thought about what will happen with the MUL's during upgrading that have non-zero at their inputs, but I could check this only at work on monday. Do you mean that non-zero inputs retain its values after updating? Yes. Non-zero values do live through the update. I am working on another update now, and it appears that this issue does not exist if you update from D7-SYS V7.1 to V8.0. I've only seen it going from 6.2 to 8.0
LA_COLD Posts: 24 Rating: (8)
Suggestion To thank Quote Answer
This contribution was helpful to 1 thankful Users KSN

11/22/2013 4:59 PM	Rate (0)
KSN Platinum Expert Joined: 8/25/2006 Last visit: 11/14/2024 Posts: 2211 Rating: (280)	LA_COLD I am working on another update now, and it appears that this issue does not exist if you update from D7-SYS V7.1 to V8.0. I've only seen it going from 6.2 to 8.0 It is also interesting for me, haven't check this yet, I'll try this type of update at monday. Anyway we should always take in mind that there could be some changes in D7-FB's during updating libraries.

Suggestion To thank Quote Answer

|

Share this page:

Share this page on...

Follow us on