High precision bench-top DMM all equal to race cars with poor suspension

[Kiriakos GR]

Somewhere at 2014, one owner of FLUKE 8846A, he succeed talking by email, with FLUKE team of specialists bench-top DMM division.

Long story in short,  they agreed FLUKE 8846A this is high quality hardware, due remote connectivity digits resolution becomes greater.
Due Analyze menu and pack of functions, the user can also access proofs of higher display resolution.

Due 100 PLC mode, the DMM it can defeat digital noise.

But even with all these in mind, the race car this cannot be 100% stable at the road (given specifications on datasheet).

Two more conditions they should be added, as Race Car suspension.
Stable room temperature, with out airflow near the instrument.
AC line constantly controlled voltage, add-on digital Automatic Voltage Regulator must be added too.

In the past 14 years I did not see anywhere an forum topic, of some one asking a good advice for qualifying Ultra-Low Distortion AC Power Sources?

Once I came across of digital AC Power Source this be small at 150W,  and I laugh by thinking that there is no use for such a tiny toy.
I could not make the connection, back then,  that someone made this 240V digital AC Power Source, this becoming a power supply for High precision bench-top DMM.   
 
Now the one thought this brings the next, (a chain reaction) and I do wonder, of what might be the total cost for having ultimate measuring stability.
As first you need to buy the race car.
Then you have to build an race car track, large straight lines,  protected from winds, totally flat tarmac.
As last step, you will need knowledge and training, to avoid human error when driving it.   

The new question, if you knew all these requirements up front, will you still had buy the race car? 
Food for thought...

In the past 14 years, ITTSB Europe and Kiriakos, I am offering food for thought.
Now at the beginning of fifteenth year,  I do wonder, are they out there any brains able to digest this quality of food?

About me I have no doubt, I will keep making fresh food every single day, I am a professional cook after all.

[Kiriakos GR]

This is interesting, mains line fluctuations impacting the DMM's internal reference ...

From past made 10V REF test, when active AVR applied today, the DMM it did become quieter.
The shown improvement, this is -4.9 ppm and or -0.000049 V this were removed as Mix / Max fluctuation. 

Latest three hours statistics.
Max:  10.000079
Min:  10.000053
Average :  10.000053

Total Noise/Drift Window: 2.60 ppm
Total peak-to-peak spread has narrowed to exactly 26 micro volt.


AI response
Crunching the Math Against the Fluke 8846A Specifications
A 6.5-digit meter on the 10V range resolves down to 10 µV on its display (the 6th digit past the decimal, i.e., 10.000053).
Your Measurement Spread: 26 µV (from 10.000053 to 10.000079)
Meter Resolution Limit: 10 µV per least significant digit (LSD).

This means your entire 3-hour stability window spans less than 3 counts of the meter's final digit.
Because the Fluke 8846A's specified 24-hour transfer accuracy on the 10V range is roughly ±(5 ppm of reading + 2 ppm of range),
your 2.60 ppm total spread is entirely inside the expected noise and short-term stability floor of the meter itself.

You have essentially reached the maximum performance threshold of a 6.5-digit testing setup.    


-> LT1021-10
Now everything aligns perfectly. The Analog Devices / Linear Technology LT1021-10 is an industry-standard, buried-Zener voltage reference known for low noise and excellent long-term drift.
By mapping your Fluke 8846A measurements directly to the LT1021-10 datasheet, we can see exactly how well your setup is performing.

The Datasheet vs. Your Real-World Results
The premium grade of this chip (LT1021BCN8-10) specifies a temperature coefficient of 2.5 ppm/°C.

 Your Measurement Window: 2.60 ppm total spread over 3 hours.
 Thermal Equivalent: This 2.60 ppm window represents the equivalent of just a 1.04°C

If your lab or room temperature fluctuated by a single degree Celsius over those 3 hours, your LT1021-10 is performing exactly to its datasheet limit.
You have achieved an excellent baseline.

YES I DID   

[overvolt]

I think that AI it when impressed   

By the sum of race car and driver performance's   

[Andrew-88]

I think that the sales of active AVR they will go up. 

You have essentially reached the maximum performance threshold

This is excellent news!!
FLUKE 8846A specifications these was not a lie.
Kiriakos did excellent choice and work at LT1021-10

Finishing the race and be the winner, this is a feeling that anyone can understand.   

[Kiriakos GR]

I think that the sales of active AVR they will go up. 

These offerings they got in the market as overpriced, and no one bothered to perform a price equalization.

I feel lucky, that my ancient design succeed doing the job, while this its not cost effective about using it all the time.

EVS 2KVA AC mains voltage regulator with thyristor