DM wiping when reading from EEPROM

[wombat]

I'm interfacing with a T100M over serial coms, using multi-point protocol commands. When reading EEPROM integer data (RXI...), sometimes timeouts occur at the PLC and this appears to correspond with timer based operations on the PLC. the problem gets worse as it also causes the DM to be wiped.

I'm flying blind on this as I don't have the PLC's code; everything is based on external observations so any input would be most welcome.  I've observed that one timer appears to reset when this timeout occurs. What's even more interesting is I can only get it to do that when reading more than 16 words at a time. If it is Hex 20, (32 words), the timeouts when they occur will correspond with one of the timers restarting and the DM being wiped. If I only use Hex 10 words, timeouts are benign.

What could cause this conflict ?

Thanks

[support]

What is the software that executes the RXI to read from the PLC? Or did you simply use the TLServer's Serial Port Setup screen to read from a T100M PLC?

DM data are wiped - is it caused by the PLC being reset? Could it be due to watch dog timer reset? If you have access to the PLC please observe if the 3 status LEDs (1 green, 2 red) turns on of 1-2 seconds then goes off? If so there is a WDT reset.

[wombat]

Software to access the PLC is .NET 4, using the Serial class. All seems good. When using only 16 words, there's still occassionally timeouts, but no problems. With 32 words it works to read although it may take retries when the conflict occurs, and causes the DM wiping. In quick simple tests the 32 word usually works with out a hitch, although occassionally it will cause the problem; it has to be done in succession so as it hits the timer to cause the problem to get a good repro.

Not sure on the lights as I am accessing remotely. Is there a programatic way I can tell if it has had a WDT reset ?

Thanks

[support]

The other way is to read the result of the STATUS(1)  - but since you have no access to the PLC code then it is not possible for you to modify the program to show the STATUS(1) result.

Perhaps reading of 32 words (maximum limit) from T100MD may have taken too long that the watch dog timer triggered under certain circumtances. T100MD turns off interrupts during EEPROM read and write so that may be a contributing factor and could also lead to communication error.  

It is better to break down your EEPROM read into two separate commands from different starting addresses.

P/S: The FRAM used on Nano-10 and F-series PLCs are newer and much better technology for non-volatile memory storage since FRAM read and write at full speed and has unlimited write life.