// How long should he run it continuously to recover? Before sending it for
calibration?
// How long should he run it continuously after changing the oven temperature?
David,
the real question is, if the LTZ1000A is affected at all by this timely hysteresis.
SN-18 gives no hint and no help, just lets you leap in the dark.
So I proposed to use a known good external reference (stable over time, near zero T.C.) for comparison. It's not requested by SN-18, that's correct, but if you have one, you can decide within 24h, simply by monitoring the 3458A over that time.
Prerequisite is a stable environment (+/-0.1°C over 24h), otherwise you'll only see the different T.C.s of the 3458A.
The drift of the U180 A/D hybrid should also be visible by such a measurement, but you can as well do that w/o an external reference.
But any temperature change will also influence CAL? 72 parameter, as this parameter is used for ACAL-compensation of any drifts of U180.
Maximum T.C. for U 180 is on the order of 0.35ppm/K, as given in the datasheet, i.e. 0.5ppm/K overall (w/o ACAL) minus 0.15ppm/K for the LTZ (with ACAL)
My U180 has about -0.24ppm/K, as in the picture which I'll append. I've measured this parameter at different temperatures, especially also at 25°C, making an ACAL DCV directly after power on.
You also see, that my instrument / U180 seems to be quite stable over time, during the season with constant temperature in the basement.
Sorry Illya, but your 0.01ppm/day criterion can not be determined practically, because it's hardly possible to extract such a low timely drift from thermal fluctuations .. this would require a stable room temperature on the order of < 1/35°C over a whole week, or so.
A more practical approach is to observe CAL? 72 and TEMP?, after each ACAL DCV over one week, and then trying to separate the temperature effect from a possible gross timely effect.
Everything else, if it's an instrument which is OK only (like mine), or even a very good performer, has to be extracted from a longer observation time.
Frank
// How long should he run it continuously to recover? Before sending it for
calibration?
// How long should he run it continuously after changing the oven temperature?
David,
the real question is, if the LTZ1000A is affected at all by this timely hysteresis.
SN-18 gives no hint and no help, just lets you leap in the dark.
So I proposed to use a known good external reference (stable over time, near zero T.C.) for comparison. It's not requested by SN-18, that's correct, but if you have one, you can decide within 24h, simply by monitoring the 3458A over that time.
Prerequisite is a stable environment (+/-0.1°C over 24h), otherwise you'll only see the different T.C.s of the 3458A.
The drift of the U180 A/D hybrid should also be visible by such a measurement, but you can as well do that w/o an external reference.
But any temperature change will also influence CAL? 72 parameter, as this parameter is used for ACAL-compensation of any drifts of U180.
Maximum T.C. for U 180 is on the order of 0.35ppm/K, as given in the datasheet, i.e. 0.5ppm/K overall (w/o ACAL) minus 0.15ppm/K for the LTZ (with ACAL)
My U180 has about -0.24ppm/K, as in the picture which I'll append. I've measured this parameter at different temperatures, especially also at 25°C, making an ACAL DCV directly after power on.
You also see, that my instrument / U180 seems to be quite stable over time, during the season with constant temperature in the basement.
Sorry Illya, but your 0.01ppm/day criterion can not be determined practically, because it's hardly possible to extract such a low timely drift from thermal fluctuations .. this would require a stable room temperature on the order of < 1/35°C over a whole week, or so.
A more practical approach is to observe CAL? 72 and TEMP?, after each ACAL DCV over one week, and then trying to separate the temperature effect from a possible gross timely effect.
Everything else, if it's an instrument which is OK only (like mine), or even a very good performer, has to be extracted from a longer observation time.
Frank
