I'm looking for multiyear data from precision voltage references from initial 
power on which includes  time-date, temperature and voltage.  Relative humidity 
would also be nice if it is available.

I am a retired PhD level oil industry research scientist.  I spent a number of 
years heavily involved in the mechanics of materials.  I also spent several 
years working on sparse L1 pursuit (aka basis pursuit) solutions to inverse 
problems involving the 1D heat equation as it describes 1D fluid flow in porous 
media which is important to understanding the behavior of shallow reservoirs 
after fracking.

The aging curve for references appears to be of that general form so I'd like 
to take a crack at seeing how accurately I can forecast the aging drift.  Prior 
experience and testing solutions of the heat equation indicate that one can 
make good predictions for a period equal to the length of the prior history.

So if anyone has multiyear data and is willing to share the first half of the 
series with me I'll see what I can do.  It would be helpful though, to have the 
full series for a few references with long histories so I can test myself 
before submitting the results for general scrutiny.

In considering integrated circuit device construction, it seems likely that 
both the aging drift and the voltage hysteresis caused by thermal excursions 
are related to the different coefficients of expansion of gold, aluminum and 
silicon at the wirebond.  In particular, I suspect thermoviscoelastoplastic 
deformation as the mechanism.   However, I have not seen a photomicrograph of 
an LTZ1000 reference die showing the bonding details.  If anyone knows of such 
I'd be grateful for a link.  

For reference, Feynman demonstrated thermoviscoelastic deformation at the 
Challenger hearing in his famous C clamp and ice water experiment.  In the case 
of a voltage reference one must also consider plastic deformation to be able to 
account for a permanent change in the value.

Thanks,
Reg

_______________________________________________
volt-nuts mailing list -- [email protected]
To unsubscribe, go to 
http://lists.febo.com/mailman/listinfo/volt-nuts_lists.febo.com
and follow the instructions there.

Reply via email to