8.4 EEPROM Write Cycles

By specification the EEPROM memory used in this range allows 100,000 changes (although typically much more). If this write cycle count is exceeded the instrument will display an E2.Er message. It will become unusable and must be returned for repair. For the main control setpoint only this figure is increased to one million write cycles.

 

In order to provide the user with advanced warning of a potential problem a warning alarm is generated if a parameter write cycle approaches a threshold (section 4.4.2).

 

The following sections give examples of parameters which could cause this limit to be exceeded over a period of time.

 

Setpoint Ramping

Although writes to the main control setpoint have been increased to one million, continuous changing of setpoint via digital communications – for example a ramping value – could still be the cause of EEPROM wear.

 

One solution, given in the section above ‘Host/Client (Broadcast) Communications’, is to select “Remote Setpoint” in the Variables list and write values to Modbus address 26 (hex 001A).

 

An approximately 5 second timeout is applied to writes to Modbus address 26 so that if values are not received within this period, a remote fail alarm will be generated (section 4.3.13) – this can also trigger a problem with EEPROM wear - see ‘Alarms and other Status Changes’ below.

 

To avoid this problem write to the Target Setpoint at address 02, but note that any value written to this parameter will not be retained over a power fail. In order to access the Target setpoint it is also necessary to enable the remote Setpoint (iTools STATUS list address 276).

 

It is critically important to select the remote setpoint if updating the setpoint on a regular basis otherwise the setpoint change will be saved to non-volatile memory and EEPROM wear will result.

 

Alarms and other Status Changes

Alarm status is saved in non-volatile memory and this includes status alarms such as sensor break, loop break, remote fail and individual alarm and alarm latching status. Every transition into and out of an alarm condition triggers an EEPROM write. Thus, if there is any fast toggling of an alarm status, EEPROM wear can result within the expected lifetime of an instrument.

 

An example of this is where event alarms are used to provide an on/off control loop. 800 series instruments should, on no account, be used in this manner since the toggling of the output will rapidly use up the 100,000 writes. The On/Off control in the PID algorithm should be used instead. However, any situation where alarm states can change rapidly should be avoided.

 

Mode and Timer Changes

Rapid changes to instrument mode (Auto/Manual) or the Timer operation can cause EEPROM wear because the status (run/hold/reset) or the segment number are stored in EEPROM on each transition.

 

In normal use where segments or timer sequences are relatively long, it is unlikely that problems will be seen. However, in some applications where a sequence is run frequently, EEPROM wear will occur. An example of this is where a digital input is used in an application to trigger a timer sequence and the operation is performed as fast as possible by an operator, EEPROM wear occurred after a few years.

 

Digital Inputs

Care should be taken with any rapid cycling digital inputs. Typically a digital input triggering timer or mode changes (as above) should be carefully considered so that they do not switch more than 100,000 times during the expected lifetime of the instrument.