The 732A has an actual output of 10.00000471 when compared to a 732B sent from Fluke Calibration. Fluke assigns output values of their 732Bs based on their superconducting, cryogenic Josephson Effect standard, which is accurate to 18 decimal places. The error of the 732A is taken into account. We only track the error on the 732A. No turning of the pots. That was done 04/14. The maximum drift of a working 732A is 0.6ppm/month. That was measured by a 48 hour test, and the 732A is within spec., and it's within calibration. So, the uncertainty is 0.6ppm*10months=6ppm max, but probably less. Given that the 732A has had no observable change since last calibration, it is probably less than 6ppm uncertainty.
.... claimed stability of 0.15ppm/°C
that is the mean s=10 from 16-26C.
... 1.5ppm/1kHrs in operation
the long-term drift is stated as can be expected, which is a bit softer than will be.
Hello Awesome14,
these additional information are to my opinion much more reputable.
I think, it is not necessary to have such a long and tiring text on your ebay offer, which made the impression on me of a market crier.
I would recommend, that you bring the specification of your device in a more familiar form, only using the important parameters with typical and maximum limits. Latter ones are really missing, or misleading in your current specification and description.
Then it will be obvious, that your product really delivers the (quite good, but not stellar) performance you intended, and what everybody who is trained in electronics / metrology can expect from such a crude setup, and also without the need to refer to any of these 'miraculous' or whatever features or treatments.
It's now obvious, that there are no such, and the buyer only gets what everybody can see in the pictures.
Your device and the trimming of output and T.C. is really very similar to the Geller Lab device, and I think it therefore may perform similar, despite your design.
Additional questions:
- How often do you calibrate your 732A, and do you have a history?
- What denotes "s=10"?
Do you mean, that you measured the averaged T.C. by taking 10 output measurements between 16..26°C, and then really trimmed to 0.15ppm/K?
- Did you ever measure these 1.5ppm/1000hrs on one or on a set of devices, over a longer period of time?
OK, let's see, what we can summarize up to now from your information:
uncertainty of Fluke 732B compared to NIST / SI: 0.3ppm, 30 days / 1.6ppm, 1yr. (which one?)
uncertainty of 732A output, compared to 732B: 0.1ppm transfer, using HP3458A
stability of 732A: 0.6ppm/ mo. => 6ppm after 10 months (maximum value)
typical uncertainty maybe < 1ppm, if enough history is available, to predict the timely drift of the 732A
initial uncertainty of D-105 compared to 732A: 0.1ppm transfer accuracy, using HP3458A
Therefore, if I sum up all these uncertainties, optimistically taking the best 732B uncertainty, 30d.:
uncertainty of D-105 to NIST: 1.5ppm typical ,
6.5 ppm maximumTherefore, your claim of +/-2ppm (accuracy? uncertainty?) is reasonable, when you'd declare that as a typical value.
Here's room for improvement, if you use the history.
T.C. = 0.15ppm/K , maybe average T.C. over 16..26°C (box method), max. 2.5ppm/K (from REF102C datasheet)
=> no precise information about trimming or measuring the T.C.
realistic again is the <1ppm/K max., which Geller labs specified, because the design is similar.
Geller uses high grade, low T.C. resistors, D-105 uses inferior thin film resistors.
long term stability:
1.5ppm/1000hrs. typical, but not proven yet
more realistic is typically 20ppm/1000hrs. ; typ. 5ppm/1000hrs after 168hrs. op. stabilization
Well, if you could provide some more details about your T.C. trimming and maybe long term monitoring,
this could really make up a quite good reference.
Frank