[volt-nuts] Checking an LCR meter

[Dr. David Kirkby (Kirkby Microwave Ltd)]

I have bought an HP 4284A precision LCR meter. This is an old model with a
basic accuracy of 0.05% and covers 20 Hz to 1 MHz.

Converting the specifications into determining the uncertainty of a
measurement is nontrivial, but I think it reasonable to assume the
uncertainty will always be >0.05%.

Surprisingly the current precision LCR meter from Kesight, the E4980A (20
Hz to 2 MHz) offers the same basic accuracy. So while fairly old, the 4284A
doesn't seem to be miles behind the current crop  LCR meters from the top
manufacturers.

The recommended calibration period on the 4284A is 6 months, which would
get rather expensive - on the current E4980A the calibration period is a
more respectable 12 months.

I am looking for suggestions on how I can get "reasonable" confidence in
the instrument at "reasonable" cost, without returning it to Keysight every
6 months.

I have a 3457A DVM, but mot much else in the way of precision low frequency
equipment.

It has 4 BNC connectors for Kelvin probes.

I suspect that getting precision resistors and keeping them for a house
standard might be worthwhile,  but are looking for suggestions on the best
approach.

I will send it to Keysight once when it arrives to ensure that there are no
faults on it, but I don't currently feel I can justify getting it
calibrated every 6 months.

Maybe I can make some stable "standards", then measure them soon after the
LCR meter calibrated been calibrated and periodically measure their values.
Any suggestions about how to approach that?

Dave