Skip to Content

Register Login

Industry Online Support

Technical Forum

Navigation

CPU sometimes have memory error after power cycle.

Created by: xBigGx at: 12/26/2011 10:32 PM (25 Replies)

Rating (2)

Thanks 8

Result pages:: |; 1 | 2 | 3; |

26 Entries

Entries per page: 10 | 20 | all

12/26/2011 10:32 PM	Rate (0)
xBigGx Posts: 68 Rating: (0)	Hi, I have a propblem with a S7 1200 (6ES7214-1AE30-0XB0, 1214C DC/DC/DC) CPU. (Firmware: V2.1) CPU is running for a few months now in a waste water facility. There are some troubles with the installation causing the power to go down few times a week. It happened a few times the CPU did not go to RUN anymore after a power cycle and has to be set to factory default to correct some internal memory error. (see attachement) Then we have to reload the hardware configuration and software to get him up and running again. Could it be that writing continiously in retained var's could cause this ? Programmed a totalizer in a Cyclic OB (100ms), the result is written in a DB withretained vars. Or is this impossible and could it be hardware faillure ? I can power cycle the damn thing for 50 times and nothing happens and then suddenly its dead ... Attachment Diagnostics.zip (603 Downloads)
xBigGx Posts: 68 Rating: (0)
Suggestion To thank Quote Answer

12/27/2011 3:11 AM	Rate (0)
Marko Bursic Platinum Member Joined: 9/8/2009 Last visit: 7/17/2023 Posts: 1410 Rating: (154)	I think you should ask support on Siemens mySupport, it is clear that has hanged by itself when powered down, 650ms after, probably you're right that it couldn't write all retain data. If you have large instance block, uncheck retain if you don't need to retain values. Retain only those which you really need, and send report to experts. 4 of 16; Event ID: 16# 02:4046 CPU error: Inconsistent configuration object (not user relevant: AID=16#00000000, ORID=16#8A0E0001) Function aborted, new startup inhibit set: .. - parameter assignment error pending Incoming event 12/21/2011 10:44:59.751 AM ----------------------------------------------------------------------------------------------------------- 5 of 16; Event ID: 16# 02:4001 CPU info: Power on memory card type: No memory card CPU changes from NO POWER to STOP (initialization) mode Outgoing event 12/21/2011 10:44:59.098 AM

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

12/27/2011 2:25 PM	Rate (0)
0_Einstein Diamond Expert Joined: 7/7/2010 Last visit: 1/24/2025 Posts: 15558 Rating: (2468)	Try some of these: Install a UPS to powerthe 24VDC power supply providing power to the PLC. Verify the common is properly bonded to earth ground for the power supply and that the PLC ground and common have no voltage potential between them. Put in a separate power supply for the PLC, using same ground as existing power supply. Move your totalizer to a different OB set for 100 ms rather than your main OB. Update to latest firmware if you have not done so already. If you are not using V11 SP2, maybe something was fixed in that service pack addressing this problem.
	science guy
Suggestion To thank Quote Answer
This contribution was helpful to 1 thankful Users xBigGx

12/27/2011 3:15 PM	Rate (0)
xBigGx Posts: 68 Rating: (0)	Marko Bursic If you have large instance block, uncheck retain if you don't need to retain values. Retain only those which you really need, and send report to experts. I made a list from the beginning to keep the retained vars minimal. 3 REAL values, written every 100ms.(Totalizers) 3 REAL values, 1 BOOL value containing control parameters only written when changed by user. (never written) 6 BOOL values containing edge memory bits. Manual states they have to be stored in a DB, maybe I can uncheck the retain option on these because i do suspect they are written every program cycle ? (Pulse_buffer_(1-6)) [IMG]http://users.telenet.be/bigg/siemens/PSEDGE.jpg[/IMG] I'm aware a UPS would solve this problem :-) Common is not connected to the earth ground, but is this in the wiring manual??? Totalizer is in a different 100ms OB. CPU firmware is up to date to my knowledge. Not using SP2 yet, could try this. Thx for the reply's, Regards, Gunther.
xBigGx Posts: 68 Rating: (0)
Suggestion To thank Quote Answer

12/27/2011 5:25 PM	Rate (0)
0_Einstein Diamond Expert Joined: 7/7/2010 Last visit: 1/24/2025 Posts: 15558 Rating: (2468)	xBigGx ... I'm aware a UPS would solve this problem :-) Common is not connected to the earth ground, but is this in the wiring manual??? Totalizer is in a different 100ms OB. CPU firmware is up to date to my knowledge. Not using SP2 yet, could try this. Thx for the reply's, Regards, Gunther. I'm not suggest a UPS is the final solution. However, if you have a DC power supply with maybe marginal capacity or bad brown-out handling or short time current overload capacity, a UPS can help you figure out if there is noise in the cabinet. I've seen a large AC load create memory corruption when the PLC is installed too near that device (like an AC drive). It was a simple panel layout mistake that was easily remedied once identified. Use the UPS to see if it's a simple brown-out issue (meaning a larger power supply might help). Where I work, having DC power supply commons electrically connected to panel ground works quite well to keep noise off the common line. I have worked on very high frequency equipment where nothing was allowed to connect to ground except where required to prevent shock and protect equipment. Finding the root cause can help prevent it from coming back. Sometimes just adding a separate PLC power supply fixes the problem, but you do not know what really fixed it. Measure the potential of your PLC common to your ground. It should be very very small mV. When you get it working, please post your results.
	science guy
Suggestion To thank Quote Answer

12/27/2011 8:23 PM	Rate (0)
xBigGx Posts: 68 Rating: (0)	Thx for the suggestions Huggy, but i think i can be sure that this is not a power issue. The PLC has his own (Siemens) DC convertor and I have had these 'crashes' during configuration also. (other location, other power source.) Regards, Gunther.
xBigGx Posts: 68 Rating: (0)
Suggestion To thank Quote Answer

12/27/2011 8:38 PM	Rate (0)
0_Einstein Diamond Expert Joined: 7/7/2010 Last visit: 1/24/2025 Posts: 15558 Rating: (2468)	So if I read your last response correctly, this same problem has occurred on your desk as well as other installations. Since the PLC has its own separate power source, is that power supply properly grounded? Memory corruption seems odd after power cycle as I've not seen it with the S7-1200 even after 100's of power cycles over the past 1 1/2 years of using them in several locations. My peers use the S7-1200 PLC and they likewise have not seen this problem. Maybe a picture of the terminals or schematic for the power to the device might help highlight a problem. It's such a serious problem for your system but many others have not had this except in certain situations. For those that have the problem, it seems to happen often. For those who do not have the problem, it seems it never happens or is so rare we do not recall it. My intuition says the most likely culprit is ungrounded 24V DC power supply.
	science guy
Suggestion To thank Quote Answer

12/27/2011 8:39 PM	Rate (0)
xBigGx Posts: 68 Rating: (0)	HOWERLIN I have seen this problem in forum long time before, but now anyone know the correct solution to avoid it? I wonder if I could continue use S7-1200, even now have SP2 and V2.2. In fact, SP2 is faster and more mature, but I still got crash everyday. So you have SP2 and CPU V2.2 and have crashes every day? You power down the CPU every day ? Regards, Gunther.
xBigGx Posts: 68 Rating: (0)
Suggestion To thank Quote Answer

12/27/2011 8:49 PM	Rate (0)
0_Einstein Diamond Expert Joined: 7/7/2010 Last visit: 1/24/2025 Posts: 15558 Rating: (2468)	xBigGx HOWERLIN I have seen this problem in forum long time before, but now anyone know the correct solution to avoid it? I wonder if I could continue use S7-1200, even now have SP2 and V2.2. In fact, SP2 is faster and more mature, but I still got crash everyday. So you have SP2 and CPU V2.2 and have crashes every day? You power down the CPU every day ? Regards, Gunther. I'm using V11 SP2 with CPU FW V2.2 and find that downloading without stopping the PLC is not 100% yet. I changed something in a DB and had to redownload. If I just change logic changes, it seems to allow that without stopping the PLC. I see crashes as well, but it's the software, not the plc. I restarted my XP VM and the crashes went away.
	science guy
Suggestion To thank Quote Answer

12/27/2011 8:55 PM	Rate (0)
xBigGx Posts: 68 Rating: (0)	huggy_d1 So if I read your last response correctly, this same problem has occurred on your desk as well as other installations. Since the PLC has its own separate power source, is that power supply properly grounded? Memory corruption seems odd after power cycle as I've not seen it with the S7-1200 even after 100's of power cycles over the past 1 1/2 years of using them in several locations. My peers use the S7-1200 PLC and they likewise have not seen this problem. Maybe a picture of the terminals or schematic for the power to the device might help highlight a problem. It's such a serious problem for your system but many others have not had this except in certain situations. For those that have the problem, it seems to happen often. For those who do not have the problem, it seems it never happens or is so rare we do not recall it. My intuition says the most likely culprit is ungrounded 24V DC power supply. Yes, the problem did occure on a desk while configuration was in progress. Back then I contacted my supplier and they recomended CPU firmware upgrade to V2.1 to solve this issue. The Power suply is properly grounded. I have also other projects with identical CPU's, power sources and build the complete system myself. Only 1 CPU has this problem. The other did NEVER crash after power cycle. The common on the DC side is not grounded, but thats the same on al the CPU's I have running. The only difference I can pinpoint is the software. This is the only CPU with a 100ms OB that writes in retained REAL var's every 100ms.
xBigGx Posts: 68 Rating: (0)
Suggestion To thank Quote Answer

|

Share this page:

Share this page on...

Follow us on