The LTC6655 datasheet shows smooth temperature drift curves. I ordered the EVM DC2095A-B and measured the drift by doing a temperature ramp and keeping sampling the reference output. There is some random noise shown up and it may result from my measurement instrumentation.
I am just wondering how is the measurement carried out for the curves shown in datasheets? Is it also doing a temperature ramp, keep sampling and then filter the high freq noise in math?
Yup, those are connection of data of several points. Its not necessarily every point in the waveform is measured, but these multiple points should be enough to predict the points inside two…
What measurement device are you currently using? How do you ramp your temperature that fine? Or maybe how many samples did you take for the whole data?
Normally the way we test it is to measure it at a number of specific temperature points.
Thanks for the quick reply.
I am using the 34401A DMM sampling at 10 Hz. The EVM is put in an oil bath and it takes about 30 mins to ramp from 0 to 80C.
So the smooth curve shown in LTC6655 datasheet is the connection of several points?
Yup, those are connection of data of several points. Its not necessarily every point in the waveform is measured, but these multiple points should be enough to predict the points inside two measured points.
Another thing to consider here, is the how long the oil bath stabilized at certain point of temperature. Will you be able to check the temperature of the oil bath along with the data?