Thanks for the TIR suggestions. I also check the Convoy page and the O-ring Simon is offering is 21mm x 1.5mm, so that must be the size I need. Could have done that before asking here. :person_facepalming: LOL
Ya know Jdub,
it would really nice to see a new Convoy soon.
Maybe USB rechargeable side switch or something. Would really MK my day. 
Is that C9 still in the works? That looked pretty impressive! https://budgetlightforum.com/t/-/38724/904
The budget market competition is advancing rapidly! (Astrolux C8, ?X, Emisar D1S)
I’m buying a black S2+ XML2 3000K T4 7A to use with diffuser as an incandescent-like floody lantern around the house at night. I’d rather have an XP-L2 3000K 7A maybe for efficiency, but I dont see any of those around.
I would mostly be using ~100 to ~300 lumens, with occasional moonlight.
1. Which driver should I buy for this use, x8 x6 x4 or x3? (edit 2, this comment says x3 has a 350lm-400lm high, so x3/4 will be fine here, esp as this use case involves bedtime with an almost-three-year-old who loves flashlights.
2. Did I read that we can request the newer firmware/UI with more options, even on the non-desert tan model? If so, how can I make this request? (edit- i see this comment answers how to request the Biscotti firmware)
From Cree PCT
Without counting losses and at 85ºC an XM-L2 T4 does (at 100% = high mode):
343.8lm @ 1.05A (3x7135) = 327 lm/A
438.2lm @ 1.40A (4x7135) = 313 lm/A
604.5lm @ 2.10A (6x7135) = 288 lm/A
749.7lm @ 2.80A (8x7135) = 268 lm/A
relative to 3x7135 (100%) the lm/A efficiencies are:
4x7135 313 lm/A -> 95%
6x7135 288 lm/A -> 88%
8x7135 268 lm/A -> 82%
In the 3/5 firmware the output levels are 5, 40 and 100%.
In biscotti the levels are 0.1, 1, 10, 35, 100% or 1, 20, 100% or 0.1, 1, 10, 50 depending on the group mode(excluding blinkies, reverse and shorter sequences).
With an x8 driver :
749.7 * 50% = 375lm (1.40A)
749.7 * 35% = 262lm (0.98A)
749.7 * 20% = 150lm (0.56A)
749.7 * 10% = 75 lm (0.28A)
With an x6 driver:
604.5 * 50% = 302lm (1.05A)
604.5 * 35% = 211lm (0.73A)
604.5 * 20% = 121lm (0.42A)
604.5 * 10% = 60 lm (0.21A)
With an x4 driver:
438.2 * 50% = 219lm (0.70A)
438.2 * 35% = 153lm (0.49A)
438.2 * 20% = 87 lm (0.28A)
438.2 * 10% = 44 lm (0.14A)
With an x3 driver:
343.8 * 50% = 172lm (0.52A)
343.8 * 35% = 120lm (0.37A)
343.8 * 20% = 69 lm (0.21A)
These values are out the emitter lumens as most flashlights are rated that way. To get real, out the front (OTF) lumens you need to multiply by 70% to 85%.
A 7135x3 or x4 driver will get about the levels you want but no ‘turbo’ and you get the most lumens per A. A 7135x8 driver will get somewhat less in the medium and low mode and a bright but hot high mode. Also you could get the 7135x6 driver and keep it in the sequence 0.1, 1, 10, 50 most of the time, you will get 60lm medium and 300lm high. If you know in advance that you will need the 100% high you can change the mode group to 1 or 2 but imho changing modes all the time is not practical.
To ask for the biscotti firmware see this thread Convoy BLF Discount? . When you contact Simon, you can ask for the updated firmware. Also you can ask for the AR glass ($1 more) and the discount for BLF members if you wish.
Thank you, that thread on the biscotti is better than what I’d found.
Edit, also re the discount, I am all about coupon codes, promo codes, sales, etc, but I don’t think I’ll ask for a $3 discount on a $20 light if I have to contact him personally to program new firmware 
Not easy to find a 3000K or 3500K stock light either, should be interesting to add to the tint collection.
That’s exactly the configuration of my S2. So far, I haven’t found any need at all to go over 50% power with it.
That’s what I always do for Convoys. However, I notice that 8x7135 drivers still tend to drop output quite a bit as the battery depletes, even though they’re supposed to be constant current. This is with a Samsung 30Q battery, so I don’t think it’s the battery’s fault. By the time the battery is about half-depleted, I’m only getting about 66% the original output (on max).
So, it seems that 8 7135 chips must drop the working voltage quite a bit. Low enough that each 7135 chip doesn’t have enough voltage to maintain its constant 350mA current.
Has anyone with a 4x7135 or 6x7135 design tested their output to see if it drops as the battery depletes? I’m guessing that fewer chips would result in more constant output. (Though lower output, of course.)
7135s are in parallel, so 8x7135 shouldn’t cause more voltage loss on their own than 4x7135. When the battery depletes to or below the voltage needed for the emitter to take 1.4A (according to its Vf and taking into account losses), a 4x7135 driver and a 8x7135 driver should allow the same current as both are not longer regulating it. When it’s higher but lower than the voltage needed for 2.8A the 8x7135 driver will allow all the current (which will be between 1.4A and 2.8A), while the 4x7135 one will limit it to 1.4A. Lower modes use PWM, so if there isn’t enough voltage, 8x7135 at 50% will be lower than 4x7135 at 100%.
Short answer, when the battery depletes enough, 4x7135 and 8x7135 should have the same output if the emitter is the same.
Makes sense. But, I assume that the 8x7135 must hit that minimum required voltage level pretty quick, and is not able to deliver the full 2.8A to the LED. Because, it doesn’t remain at full output for very long on a fresh cell.
Based on your explanation, the 4x7135 driver should maintain its full output for quite awhile. Because, I don’t think the 8x7135 drops to under 50% output until the battery is depleted to well under half capacity.
So, I think a 4x7135 driver design is good if you want well-regulated flat output. 8x7135 is good for much higher output, but at the cost of losing flat output.
You could also just use 8x7135 or 6x7135 capped at 50%. It maintains very good flat-regulated output, presumably like full blast on 4x7135 or 3x7135 respectively.
Well, according to ggf’s explanation, the 8x7135 at 50% will have less output than the 4x7135 at full output, for at least part of the run-time. Since PWM is used, the 8x7135 will drop out of regulation at a higher battery voltage than the 4x7135. Until the 4x7135 also drops out of regulation, it will maintain its full brightness.
Yes, the 8x7135 at 50% will have a flatter output curve at 50% than at 100%, but that is because the run-time is twice as long.
I’m starting to think a FET+1 driver, like in the BLF-A6, is a better way to go. Flat output on lower modes (using a single 7135 chip), and higher output (but unregulated) on high modes.
I wish the Convoy’s 8x7135 setup would allow for better regulation. Perhaps, if instead of using all 8 chips on all modes, it could use fewer chips on the lower modes. For example, use only 4 of the 7135 chips when running at 50%. That way, you’d get the benefit of flat regulation on lower modes, but still be able to get high output off all 8 chips when your battery is fresh.
I'm not sure about my Convoy S2+ Desert Sand w/ Biscotti firmware:
(I get these actual tailcap current measurements, using a non-calibrated Mastech MS8229 multimeter, using a freshly charged Samsung INR18650-30Q battery (battery voltage around 4.15+v)
on Convoy S2+ Desert Sand 7135x6 (3A tint) XP-L Hi (Biscotti):
0.008A (0.1%)
0.059A (1%)
0.34A (10%)
1.19A (35%)
2.11A (100%)
*my Convoy S2+ Biscotti driver appears to have a 'running too fast' driver (bike strobe = flashes 115 times per 60 seconds); it might have been affecting the results...
If I measure my Convoy C8 7135x8 Clear (1A tint) XP-L Hi (Biscotti)
0.005A (0.1%)
0.046-0.050A (1%)
0.29-0.30A (10%)
1.16-1.20A (35%)
2.77-2.79A (100%)
(Convoy C8 Biscotti driver: bike strobe = flashes 55 times per 60 seconds)
For some reason, the Convoy S2+ 7135x6 lower modes (0.1%, 1%, 10%, 35%) are about equal current to the Convoy C8 7135x8 lower modes. Not sure what causes it. Only the 100% mode shows the correct current (7135x6 = 2.1A and 7135x8 = 2.8A)
It would appear that at the non-100% modes, the S2+ 7135x6 might not be lasting any longer than the C8 7135x8 since it's consuming basically similar current?
For measurements below 0.01A, I'm not sure if my multi-meter is accurate enough though.
For comparison (same battery, same multimeter):
my Astrolux S1 (3D tint, XP-L HD):
0.003A (moon)
0.016A (low)
0.14A (med 1)
0.43A (med 2)
0.83A (high 1)
1.86A (high 2)
4.05A (turbo)
The Astrolux S1's moon mode seems to be a bit dimmer (I like the S1's moon mode better than the S2+ or C8).
FYI, I finally got an 18350 tube for my DT… with the Aspire cell, it comes in at about 91g. It’s cute.
Ahah, nice A shorty one is in tha house now
Curious, the DT is heavier than the regular S2+ . My guess: the metal switch and the “extra” length on the tail, right? But, it is still a difference, 16grams 5.n … grams!
I hope you enjoy that shorty one! Mine hasn’t left the small version since I got the tube :sunglasses:
FYI, my DT S2+ (8x7135) in stock form (18650 tube) weighs 79g, while my regular blue S2+ (4x7135) weighs 80g, without battery of course. Both have metal switches.
I’ve just weighted my 2 regular S2+ and the S2+ DT and there are 2/3 grams differences both in the head and in the tail. Individually I can’t find “differences” but when mounted, the DT head has 2/3 grams more, as its tail!
One of the black S2+ is 77gr, another 80 gr (it has retaining ring), and the DT S2+ is 84gr. All have a magnet installed and all in 18650 shape.
About your differences, I’ll make a joke, but…maybe that “black metal” is not that heavy after all :smiling_imp:
Can it be that or another thing, like the solder on the driver, a retaining ring, or maybe the plastic or metal “spacer” in the tail (between the switch and the button).
Is Simon’s Convoy Aliexpress store still the only place to get an S2+ Desert Tan with Biscotti firmware?
It appears so. The ones on GearBest and BangGood still use the old firmware.
The clear C8 and the BD06 appear to be the only Convoys available elsewhere with Biscotti.