From comparing reviews, it appears that the new 5A driver for T3 puts out around 220lm with the 519A 4000K, quite a bit more than the 150lm of the old driver, despite both claiming to push 0.5A on NiMH.
If you have both the new and old drivers and a NiMH to test, does the new driver seem brighter? Also, does the new driver have a higher or lower moonlight on NiMH? Are there any issues with mode spacing, etc?
Thank you in advance!
From T8 review from 1 lumen which most likely shares the same driver from T6 they state it outputs around 300lm in turbo with NIMH. Their review model has Luminus SFT-25R so for nichia 519a it will be probably around 220-250lm I think. Which it still brighter than what output at 0.5A should give (around 150lm with 100lm per watt efficiency). My own measurement for my t6 current at battery with eneloop were around 3.5A so it around 3.5A x 1.1V x 0.7(efficiency) ~ 2.6W so it should give 260lm on NIMH.
So clearly driver was improved for nimh as well.
Thank you very much for this very helpful and informative first post! I ended up with an SFT40 3000K version of the new T3 a while back (and forgot to update this thread with my observations), and my experience is indeed consistent with yours: the new driver pushes the LED noticeably harder on NiMH (in comparison against an old T3 with 519A 3000K). I don’t have a reliable way to measure such high current, sadly.
Welcome to BLF! I hope you find your time here worthwhile.
Thank you!
One thing I notices for convoy drivers that there big gap in efficiency between 30%/50% and 100% mode.
For example per T3 1lumen review for old 1.5A driver when using NIMH:
on LI-ION on 30% which has same brightness with 100% NIMH we have 0.40A for 155 LM with 94 LM per watt.
Per that D3AA driver is definitely on different level since it keeps efficiency even on high current with 1.2 element. Would be great if convoy introduce a new group with 70% as max so we have around 2A as highest current with hopefully better efficiency.
In that regard I like t7 driver which only consume 1.5A even on 100% so it scales much more linearly between 30% and 100% modes and it produces much less heat with eneloop pro.
That’s very surprising, and I haven’t noticed! Going from 91 lm/W to 50 is quite a steep drop.
What voltage values are you using to calculate lm/W? Your 30% measurement suggests 1.3V, while your 100% measurement suggests 1.1V. This is consistent with voltage sag at higher currents, just curious where you got the numbers from.
Also: I don’t know how much I can cross-reference reviews of 1lumen and ZeroAir, but the latter has a T2 review (LH351D 3500K) that indicated 136lm at 1.05A. If I were to use 1.25V, this gets me 104 lm/W, which conceivably suggests around the same driver efficiency as the 91 lm/W, up to measurement errors and emitter difference.
So it looks like Convoy’s 1AA drivers have more or less the same maximum efficiency, just that the 1.5A version, for whatever reason, is extremely inefficient on 100%.
I think the 50% group should do just fine–I doubt the human eye can tell between 50% and 70% without a side-by-side comparison.
Good to know! I’m a tailswitch fan and really wish more work is done in the community to optimize tailswitch UIs. Anduril gets periodic updates, yet tailswitch UIs have seen no development since 2016 (the introduction of biscotti/12-group), which was a decade ago.
Correct. But you are right if we pick eneloop data from lygte for 3A it should be 1.2V - 1.15V range so it means the efficiency even less that what I calculated.
I did some test since I only have multimeter and light intensity sensor on android smartphone I got following:
My T5 has 20 dgr tir and T7 60 dgr tir so obliviously T5 has more candellas.
But we see T7 is still has to pump 4.6 more current to get 3x brightness. It looks a bit better since it underclocked comparing to regular 1.5 driver.
That true but it normal to desire get most output which practical.
