if you think about raising driver current specifically for this host + led combination, I think, 3A would be the best setting (keeping in mind, that this host is smaller than L6 and L21A).
Anything change with the XHP35 to run at higher amps? From this old test here, it peaks at ~2.5 amps: https://budgetlightforum.com/t/-/43473, and at 3.5 amps output starts turning blue.
Clearance will help for a bigger coil, and depends. For example on the original BLF GT buck driver DEL designed, he had designed in more amp capability, think ~5 amps or so, if I recall.Of course there was plenty of room on that driver in that light.
Can someone to tell me what is the driver diameter of new S21B. It is 22mm? From Simon store I saw that XHP50.2 version used the new 3A boost driver which is 22mm. Also now we also have very good buck driver for 3V in 22mm so we have nice S21B host for new switching drivers from Convoy, if it used 22mm drivers.
To be more precise, since output data was collected each 0.25A, we need to analyze previous and subsequent data:
2.25A: 2500 lemmings
2.50A: 2570.
2.75A: 2580.
3.00A: 2530.
Peak was somewhere between 2.5 and 2.75A. Since 2.25A shows a lower value than 3A, we can infer that the peak was closer to 2.75 than to 2.5A, probably at ≈2.66A.
Test data shows that at 2A output was already above 92% (2380/2580) of the measured peak, and 96.9% at 2.25A. In my opinion, 2.1A is already hot enough (close to 94%), imho. There's simply not enough XHP35 test data, just (!) one test. For a driver which must be paired with all sorts of XHP35 emitters, I think it is wiser not to go overboard. Not all emitters peak the same, check SST-20 tests for example: 6500K (peaked at ≈6.75A), 4000K (peaked and died at 7A?), 3000K (peaked at ≈6.15A).
Before the Chinese New Year, we have to stock up for many customers, and the employees are also unstable. After about 3-7 days, I will update the product link.