re-repeating the experiment

Following up to the blog from yesterday:

https://wb9kzy.blogspot.com/2023/09/repeating-experiment.html

First, tested 5 1n4148 diodes, these are regular silicon PN diodes, tested with the Metex DVM diode test, the LM317 sourcing 5.2 mA and then 10.5 mA:

number       1.1 mA     5.2 mA      10.5 mA
1            600        .694        .734
2            599        .693        .730
3            600        .695        .732
4            604        .700        .741
5            600        .694        .731


Then tested 10 HP 5082-2800 diodes (which are comparable to the 1n5711 used by M0NTV):

number       1.1 mA     5.2 mA      10.5 mA
1            371        .564        .746
2            366        .538        .699
3            371        .565        .745
4            371        .563        .741
5            368        .553        .721
6            371        .562        .738
7            370        .560        .738
8            372        .566        .743
9            370        .564        .745
10           372        .567        .745

Still kinda inconclusive for me, not sure why the 1n5711 is favored over the 1n4148.  Also not seeing a super huge disparity in 1.1 mA versus 10.5 mA  readings.  

I'm kinda wondering about trying curve tracing, especially the Octopus design which seems pretty simple:

Best Regards,
Chuck, WB9KZY
http://wb9kzy.com/ham.htm