Once I got efficiency of my converter over 100% (actually about 98%). I thought: the perpetual motion machine and Nobel Prize it is good, but … I need more precision tools. The most significant error is a current measurement and current shunt is a great toy. I considered some resistors (other story) and DMMs. Keithley 6500 has 1 Ohm range, therefore has better accuracy than Keysight 34461/5 (lower range 100 Ohm).
Here comparison of specifications (10 mOhm/1 year)
Keysight 34461: 1uOhm (reading) + 4mOhm (range) = +/-4,001 mOhm.
Accuracy: +/-40%. Keysight 34465: 0,6uOhm (reading) + 4mOhm (range) = +/-4,0006 mOhm.
Accuracy: +/-40%. Keithley 6500: 0,85uOhm (reading) + 200uOhm (range) = +/-0,20085 mOhm.
Accuracy: +/-2%. Here one more comparison of specifications (1 mOhm/1 year):
Keysight 34461: 0,1uOhm (reading) + 4mOhm (range) = +/-4,001 mOhm.
Accuracy: +/-400%. Keysight 34465: 0,06uOhm (reading) + 4mOhm (range) = +/-4,0006 mOhm.
Accuracy: +/-400%. Keithley 6500: 0,085uOhm (reading) + 200uOhm (range) = +/-0,20085 mOhm.
Accuracy: +/-20%. So I got DMM6500.
I understand that the actual accuracy is much better. I will show calibrations data report of my multimeter once again.
Measurement of 10 Ohm and 1 Ohm at calibration:
1 Ohm. Spec. accuracy: +/-285uOhm. Actual error: - 5 uOhm, uncertainty: 31 uOhm.
10 Ohm. Spec. accuracy: +/-1005uOhm. Actual error: - 50 uOhm, uncertainty: 140 uOhm.
Unfortunately, any data for smaller resistance it is not provided. I decided to make experiment: measurement of low resistance - 10 Ohm, 1 Ohm, 100 mOhm, 10 mOhm and 1 mOhm.
Less than 3 months after cal. I made special cable for experiment: teflon shielded + crimped connectors. All measurements repeated many times. Durations: 0.5-2 hours. Tamb - 21-23C. Settings: NPLC - 12, repeat filter count - 100, autozero ON, power sync ON … Later I will show a difference between other settings
Resistors inaccuracy not considered (better than 0.01%). I consider the resistor like an absolutely precise - I do not know the actual resistance, and it does not matter at all.
My modest experience does not allow commenting on these data, I can’t speak about uncertainty of measurements and more difficult aspects. I understand the difficulty of measuring low resistance. I will be very grateful for any comments.
This is just data:
10 Ohm. Spec error: +/-0,0105%. Measured range: 9,999611-9,999625. Error:-0.0039%.
1 Ohm. Spec error: +/-0,0285%. Measured range: 0,9999636-0,999962. Error:-0.0038%.
100 mOhm. Spec error: +/-0,2085%. Measured range: 99.99736-99.99858. Error:-0.0027%.
10 mOhm. Spec error: +/-2,0085%. Measured range: 10,00009-10,00070. Error: +0.007%.
1 mOhm. Spec error: +/-20%. Measured range: 999,5403-999,9838. Error:-0.046%.
Conclusion:
and now I have opportunity to create my own precision shunt.