[volt-nuts] 3458a RAM batteries - longer life if unit is continually powered?

Frank Stellmach frank.stellmach at freenet.de
Thu Mar 27 12:33:17 EDT 2014


Hi,

the DALLAS datasheet specifies 10 years data retention. Quite obviously, 
that's the unpowered time.

That time includes self discharge and RAM sustaining currents.

self discharge and RAM supply current will be of the same order, i.e. < 
1µA, I assume from experience.
And I assume also, that those currents will vary over samples, i.e. over 
samples of battery and RAM.

Recently, there was the case, where the CAL RAM was still alive after 24 
years (instrument vintage 1989, same DALLAS date code!), but with 
unknown powering history.

Additionally, a typical 10 years of lifetime is given also for those 
batteries.

Therefore, in practise there will be no pronounced difference between 
powered / unpowered state.
On the other hand, DALLAS implemented an electronic seal, so that the 
RAM inside the package is initially powered only after first power up. 
But there's no parameter in the datasheet, specifying data retention 
according to date code or to initial power up.


The waste of energy just to extend the batteries lifetime, does not pay off.
On the contrary, the MTBF time of the instrument will be consumed, the 
display will get bad, and here in Germany, this energy consumption would 
cost a fortune (meanwhile: 25+ Cents / kWh, crazy).


Yes, as long the instrument is powered, the CAL RAM will not lose its 
content.. but what about unexpected power outages?


If you save the CAL RAM and repogram a fresh one after failure, there's 
no need to send it to agilent.


No, not any RAM error needs recalibration.
Those error messages are not documented.

So I've scanned the text fragments in the firmware file:

RAM TEST 1
RAM TEST 2

probably refer to either to the configuration RAMs (2x32kB), andthe CAL 
RAM (8kB).

Eror messages are:
NONVOLATILE RAM HiGH /LOW
CAL RAM FAILURE

and known ones from other users:

RAM TEST 1 HIGH / LOW

So I assume, that a real failure of the cal constants only apply on 
message: CAL RAM FAILURE and perhaps RAM TEST 2 ...., as RAM TEST 1 
should refer to the configuration RAMs only, HIGH to the high byte RAM, 
LOW to the low byte RAM.

Errors of the cal RAM will manifest in additional ACAL or CAL failures.


Frank


More information about the volt-nuts mailing list