hi everybody! Could be a good idea to ovenize a cerdip MAX6350 buried-zener by a to-220 bjt as heater glued with pt100 or pt1000 RTD onto voltage reference? temp drift and hysteresis could going better?
This has come up before in this forum more times than I can count on my fingers. The general idea goes like this:
"If I take (this) inexpensive monolithic reference IC and put into a temperature controlled environment, won't that be as good as a Fluke 732?"
The short ...
Poor man, ovenized or rosted, maybe doesn't buy tantalum-nitride resistor networks, I am thinking to a
5Vdc reference to use with ltc2400 or LTC2410.
LTZ1000 is not for poor man, LM399 needs expansive resistor to
lower at 5V. Maybe I can use LM399 with LTC1043 to divide by 2 but I don't know if temp drift can
raise or not. I can buy MAX6350 in cerdip case, it has a buried zener (like LM399, LT1021, AD586) and
laser wafer trimming of high stability thin-film resistors inside the same ceramic package to lower Vz.
It remains the problem of noise, 1.2µVp-p for LTZ1000 vs 3µVp-p for MAX6350.
I don't well understand you if I can avoid any hysteresis ovenizing cerdip MAX6350...
Temperature hysteresis for LTZ1000 and LM399 are different from Temperature hysteresis of MAX6350?
Of course I know it is impossible to reach LTZ1000 standard by a cheap reference.
Tnx a lot DiligentMinds.