ESI SR1010 LTC - 1973 label arithmetic and EU calibration options

[Muza]

I got an ESI SR1010 LTC transfer standard with an old (1973) calibration label which has a line that makes me wonder if I have the math right.

Resistor |	Individual deviation (ppm) |	Cumulative sum (ppm) |	Mean deviation (ppm) |	Nearest whole ppm |	Label value (ppm) |
R1	-1	-1	-1.00	-1	-1
R2	-3	-4	-2.00	-2	-2
R3	-6	-10	-3.33	-3	-3
R4	0	-10	-2.50	-3	-3
R5	-8	-18	-3.60	-4	-3
R6	-6	-24	-4.00	-4	-4
R7	-6	-30	-4.29	-4	-4
R8	0	-30	-3.75	-4	-4
R9	-5	-35	-3.89	-4	-4
R10	-5	-40	-4.00	-4	-4
R11	-5	-45	-4.09	-4	-4
R12	0	-45	-3.75	-4	-4

I know it is difficult to cross check averages from the rounded numbers, but no matter how we check, one of the numbers for R5 (individual or cumulative deviation) seems wrong. For the Individual -8ppm to make sense, its cumulative deviation should be around -4ppm, not -3ppm. Alternatively, for the Cumulative -3ppm to make sense, its Individual deviation should be -7ppm or -6ppm, not -8ppm 

I wonder if it's just an error in handwriting or if I'm missing something.

Additionally, I would like it calibrated with the same uncertainty, so that I can reprint the calibration table with a fresh date and see its long term drift.
But, I gotta admit, finding a 1ppm calibration lab in Denmark for a quote was not easy. A few labs I contacted seem to offer sub-1ppm only for oil filled resistors. Where do other EU-voltnuts and ohmnuts get their gear calibrated?

[Dr. Frank]

I guess, that is a simple typo only, if the deviation of the individual resistor is a +/- 1ppm.
Don't mess around that very small error.
For a complete 100Ω to 1Ω transfer, this error shoud be around <100ppm for the Hamiltonian resistive effect.
So it would not affect your 100:1 at all, I guess.
Maybe, is it possible for you to measure, whether these resistive values are still in line with that old, 50 table?


Another question would be, are you able yo measure the Resistor which is kept across the whole 1Ω block, to compensate for the switching effect, please?

Thank you.
Frank

[Muza]

Sure.
I ordered shorting bars using Copper (101) from PCBWAY yesterday and will do some measurements when they arrive.
Both of my meters (Keithley DMM7510 and Hioki RM3545A) have recessed banana sockets so i'm still searching for some good copper connectors. TBH, i want to avoid sticking Q-tips into the instrument sockets if i can avoid it but I've seen it work well.

[Dr. Frank]

Sorry for my typos in my text. I've used another small screen and a small keyboard.

I'm very interested in this resistivity, as I'm building myself exactly this SR1010 for the 100:1 Ohm reference. And I was not yet able to calculate this resistor on my own.

Many thanks again
Frank

[trobbins]

Given the age, I'd anticipate that each of the 10R steps has deviated significantly.   I posted a summary table of my SR10xx boxes in post #73 : https://www.eevblog.com/forum/metrology/esi-resitance-standard-sr1010-and-standard-resistor-sr104/msg6331564/#msg6331564.  My SR1010 LTC 10R step now has a ppm spread of 45ppm, from an initial 6ppm spread back in 1985.

I'd recommend first trying to measure the spread of the steps - with likely the cheapest method being to use a similar vintage ESI 240 bridge - that may temper your enthusiasm for paying for a cal to actual ohms once you've seen how far they can spread.