[volt-nuts] HP 3458A DC current accuracy (part2)

[Frank Stellmach]

..

but it does not explain completely the mediocre specification of 
resistance and current ranges, that's true.

Especially the 1µA range, measured directly by the internal 40kOhm 
standard resistor (TC 1.3ppm/K, and "precise" to 2ppm within 24h ) 
should be on the same level of accuracy in the specification, which is 
not the case at all.

Also, HP has not specified a transfer accuracy for the reistance ranges, 
although the instrument is capable for stable measurements if the 
environmental temperature is kept constant... Flukes version / 
modification, the 3458HFL specifies resistance transfer accuracy.

Well, to understand the instrument further, you may download the 
schematics (CLIP), and you will see, that the 100µA shunt resistor has a 
T.C. of 10ppm/k only. Therefore, it is not that stable over temperature 
(and time), so the limits have to be wider.

HP could have done much better, either with better resistors for all 
ranges, but also with the specification, or better said, with the 
validation of DC resistance and current.
  Perhaps the DCI specifications had been estimated very generously, 
saving time, instead of trying to validate to the lowest possible 
inaccuracy.


My understanding of the instrument is, that its main purpose was for 
military use (because of the broad env. temp. range) and for DCV and ACV 
mainly.

The other modes were an add on, and less attention was paid for them, 
and it was not the intention to make it a "metrological" instrument in 
first instance.

Other 8.5 digits instruments beat the 3458A by far on those ranges..

Frank