Confirmed - went from 2200 lm in kit spec (kit 26650 and everything out of the box) to 2820 lm with bypass (2x 22awg twisted ends, soldered to spring).
40T was getting 2635; up to 2965 w that bypass.
Anyone took a picture of the driver spring? I’m not sure where to solder.
I took the spring out before soldering it; just gotta “compress” it sideways at the bottom to do so.
And yes that’s two wires 
Thanks a lot!
But I meant the Spring on the driver PCB 
Oh! Oops!
I’d probably do something similar - gotta desolder the wires to the MCPCB first to get the driver out though; they’re quite short.
Ok, you only bridged the tailcap spring and that is already very successful. Good to know, I’ll try this 
I solder the bypasses inside the spring - if the solder joint breaks, you don't want that wire touching something it shouldn't on the driver side.
Though the outer bypass has advantages - the spring can completely collapse down if a battery fits tight.
On the driver side I’d do my best to do it on the inside (would spend more time on it than I did), but for the tailcap side, it’s all one slab of metal so I figured it was fine for the outside.
It’s unfortunate that “3500” ended up being 2200, though.
Yes, this is my concern. I have a 20700 for the SP33. Maybe I need to go back to the 26650. Or leaving the spring on the driver side untouched.
2200 is much lower than mine before the bypasses. By out of the box, do you mean the battery out of the box as well, not fully charged?
Sorry, no - fully charged, but the kit battery. Have tested in both a lumen tube and sphere, both calibrated with lights from Maukka.
Also some crazy thermals.
Is X in minutes or seconds? My lumens measurements are in "Tom E" calibrated lumens which is 12% higher than Maukka cal'ed. Just subtract 12% from my #'s. I didn't get their kit w/battery though.
My "Tom E" units were originally based on manufacturer ratings, but leaning towards the manufacturers that are higher end and consistent. It's rare to see any manufacturer come close to Maukka rated lumens. I try to give the manufacturers benefit of the doubt by still using "Tom E" ratings rather than Maukka cal ratings, but comparisons on BLF here is better to use Maukka ratings. For me it's easy though - 12%, that's it. I did have the Maukka cal lights to calibrate my PVC lightbox by.
I’ve gotta say… I did a couple updates to my lumen tube and set a new conversion factor (using my TO50R as a trusted source, alongside Maukka’s confirmed measurements of the TO50R) and I’m now showing 2200 lumens with my SP33 V3 as well 
Ah, sorry. X is minutes; just grabbed that from my google sheet.
This is a closeup of the first 20 minutes on High (my labelling is garbage):
My Maukka cal #'s are 12% lower - simple conversion, calibrated w/Maukka lights.
Still on a red Dolilada 5500 mAh 26650, I got 3270 at 30 secs. Subtract 12% and you get: 2870. Could be the stock cell is really that bad, or some differences in the lights.
It can be good (enough) though - I got 10A in a tailcap amp test with the Sofirn cell, and when I bypassed the spring in it, I jumped up to 2800lm. With a 40T I was only getting 11.25A…
/shrug
I’ll check it one last time tomorrow night because I’m kinda done with it 
Yikes, yeah, that 11.25 amps on a 40T is much lower that what I got (14.7 amps using a clamp meter, heavy gauge wire). That explains the lumens.
Thankyou for more detail
Shows he got a faulty light I reckon. I get just over 3200 like you Tom. Real shame it just happened to be the light up for review on 1lumen
FYI: Lan at Sofirn just informed me the 5k version of the SP33v3 is (finally) BACK IN STOCK!!! I have one ordered since last August.