Hello BabyDoc,
I ran a test that you may be interested in.
I had 4 NiMh cells and wanted to choose 2 of them to use. They had been sitting around for awhile, so I wanted to pick 2 that were reasonably matched.
I had 2 cells that tested at 60% on the ZTS. I took those cells and ran a discharge on the C9000 to see what they had left in them. One came in at 1053 mAh, and the other came in at 1185 mAh.
While it was not the best match, it also wasn't all that bad either.
The procedure we came up with for CR123 cells is to test them 3 times. If all 3 readings are the same, you go with that. If not, you continue to test until you get 3 of the same readings in a row. If this takes lots and lots of tests, there is an "issue" with the cell, and you should reserve it for a single cell application.
The ZTS is repeatable, but not precise. The actual % given does not directly apply to any cell that I have tested so far, but if you do a test, let the cell rest, then come back an do another test, the ZTS will give you the same value.
We have tried to come up with various other test methodologies to see if we could get beter accuracy, but there are such a variety of cells that one test method does not seem to apply to all of them.
When I am using used cells in a multi CR123 cell application, I match them on the ZTS and go with that. So far, so good...
Tom