Proposed Improvements to Convoy's user interface(s).

I’d settle for a way to convert the E-switch drivers to Anduril. Whether that’s with flashed MCUs, MCU swaps… I’d love the L7 with Anduril, and others as well.

If Simon was using ATTinys we’d be able to easily flash whatever firmwares we wanted.

I agree. That would be excellent. I’ve added it to the hardware changes section.

L7 with Anduril would be interesting. The tailswitch would be rendered useless except for mechanical lockout, but I’d be happy with that. It bugs me that L7 current requires two hands to operate since it’s so large and the switches are so far apart.

All sounds good TG, no complaints.

Anduril has a compile-time option to enable dual-switch.

Oh, now that’s interesting! L7 would be incredibe with that firmware then!

I’m a big fan of Convoy flashlights but my biggest problem with the convoy drivers is that they seem to be unreliable. In my experience i’d say there’s about 70% chance you’ll get one that works properly. I had to replace the 4*7135 in my s2’s multiple times because of issues like it getting stuck in the config mode or stuck with the mode memory enabled or just randomly behaviour with the mode groups.
I ordered a packet of 5 drivers and 2 out of 5 had problems with config mode.
I love how easy it is to fix the S2 if it breaks (especially with the 4*7135 driver or the 5A driver because you have a retaining ring and don’t need to solder the driver to the pill) but I really would love it if the drivers were as reliable as the hosts.

As for the 12 group driver interface, I think the battery checker mode is really nice but there are only 3 groups that have it, it would be nice if there was a shortcut for battery checker from any group.

The mode spacing for the 5A driver could be better I think, with Lh351D and SST40 I think there is too much of a difference between mode 2 and 3. It seems like it goes from around 16 lumens on mode 2 (1) to like 160 lumens on mode 3 (10). A group that has a mode with 5% brightness would be really nice as a balance between brightness and runtime for general use.

I’m thinking maybe 1, 5, 10, 35 and a shortcut for 0.1% moonlight, 100% turbo and battery checker but realistically I’d be happy if the current drivers were just consistently reliable.

I quite like 12 group/Biscotti. It is nice and simple. The most pressing change to 12 Group/Biscotti is to change the strobe from the terrible alternating frequency strobe back to to constant frequency (anywhere between 10 to 20Hz with 50/50 on/off time would be good. Also strobe HAS to be able to memorized so that it is instantly accessible from off with a single click*. This would make it the best budget light for light painters.

  • sadly so many flashlights now make it so difficult to access strobes, that 95% of flashlights are useless for those who actually need to access strobes. It is possible to make firmware so that strobes might need the (what seems to now be usual) double or triple click to access from on, but after strobe is selected, it should then me memorized. This would mean that those who don’t need to access strobes should never enter strobe mode, but those who do use can have it memorized at the next turn on.

If Convoy does go down the Anduril route, a tail e-switch S2+ style light would be great.

I have no way of confirming this suspicion, but I think most driver issues I’ve heard of have been with 7135 based drivers. So, in theory, the newer linear drivers may be more reliable? Either way, unreliable drivers are certainly an issue.

I updated the post with details about how I think a battery check should be implimented.

After moving the blinky modes to a toggle-able option rather than having duplicated mode groups, there should definitely be space for another mode group with a 5% mode instead of 10%.

How do you suggest the 0.1% moonlight shortcut should work? The way I would do that is by selecting a mode group with 0.1% as a mode, and disabling mode memory. It’s not really a shortcut, but the light would always turn on at 0.1% mode. Do you have an alternative solution? Moonlight shortcuts are usually implemented on e-switch lights where holds are a possible action. Holds aren’t an option with mechanical switches really.

I added Anduril2 to the 3X21A, but had to lose the charging. Benefit was more power as well, 4 FET's replaced by one good one:

I know, it's cheating

That’s pretty incredible. I would be extatic to have Anduril on my 4X18A, even if it meant losing built in charging.

There ya go with the piggyback again!

I'm working on another one now for someone. Added a 3rd 20 AWG ground, and on mine I used two pair of 18 AWG wires, but for the new one, I'm using 1 pair of 16 AWG's. Every time I do a mod again, always make some improvements.

Yea I do think the 5A driver is more reliable, I only had one of them fail, I had my 5000k LH351D on mode 3 (10% brightness) and the beam started to turn angry blue and it got super hot so I had turn it off, not sure what went wrong but the lower brightness modes stopped working completely. I replaced driver and Led and it’s fine now though.
I just like the mode spacing and runtime of the 4*7135 drivers but I’d pick the 5A driver for reliability and nice output.

Well I think it’s probably good if the light always comes on in moonlight, just need to find a good balance between the mode spacing and having too many modes to tap through. Most of the groups are too similar in terms of mode spacing. Adding a toggleable option to be able to choose whether you want ascending or descending mode order would further reduce the amount of groups needed, pretty much like the Reylight driver.

I like the idea of being able to disable/enable a true low moonlight, like the way you can disable it in Reylights or the way you can disable mode memory in the current convoy drivers.

So for example an S2+ with 5A driver and SST40 (take these number with a large grain of salt)

Mode 1 (0.1%) 1.8 Lumens

Mode 2 (1%) 18 Lumens

Mode 3 (10%) 180 Lumens

Mode 4 (35%) 630 Lumens

Mode 5 (100%) 1800 Lumens

Modes 1 and 2 are fine.
There’s a big difference between 2 and 3.
Most of the time Mode 5 will look very similar to mode 4 because it will step down due to heat.

So I think adding a mode with 5% would give a nice brightness of around 90 lumens maybe in this example. And 20% could give around 360 lumens maybe.

I don’t like how the current drivers have 20% separate from any of the groups that have moonlight or 10% so all of my convoys just have the default group 1 selected, the other groups are just too similar with minor differences.

Also for your suggestion of 10 taps for battery indicator, I think that would be good as long as you don’t accidently go into config mode and yea i’d love too see accurate voltage read outs rather than the current indicator.

Agreed. I think having several groups starting at moonlight and ending with 100% would be good. Probably five total, ranging from two to 6 modes, with an even multiplier between the modes so the spacing looks consistent. For example, the 4-mode group would look like 1 - 10 - 100 - 1000 lumens, where the multiplier is 10x.

Being able to reverse the mode order is a great idea. That would eliminate a few duplicate mode groups and it would also allow for more total configurations.

That strikes me as crossing the line from adding complexity for the sake of improvement, to adding complexity for the sake of complexity. I don’t see a need for moonlight mode to be toggle-able because I think most users who would not want moonlight would be served just fine by a mode group that doesn’t have it.

Keep in mine we perceive brightness logarithmically. Jumping from 1.8lm to 18lm is visually the same increase as jumping from 18lm to 180lm. Out of those particular lumen numbers you provided as an example, mode 4 seems out of place to me because it breaks the otherwise consistent 10x lumen increase multiplier between modes.

Yeah, I would not want the 20 taps for programming to be changed, so 10 taps for battery check should not get anywhere near the programming mode.

It is not accurate to say “we perceive brightness linearly.” We perceive it logarithmically. 2 lumens does not appear twice as bright as 1 lumen. But you are comparing relative increases in terms of percentage and saying equal % increases will be perceived equally.

You are absolutely right. That was a typo and has been corrected.

That’s great if the number of groups can be reduced and added more variation for users. Like if you preferred a different multiplier for the mode spacing. I don’t know if it’s possible to make the brightness of the mode groups this specific (I have rounded up and down some of the numbers) but it would allow for a lot more customization of brightness and personally I think the x3 multiplier could be really nice.

0.1, 0.4, 3, 17, 100% (x6 multiplier)

0.1, 0.8, 4, 20, 100% (x5 multiplier)

0.4, 1.5, 6, 25, 100% (x4 multiplier)

1.2, 4, 11, 33, 100% (x3 multiplier)

0.4, 1.2, 4, 11, 33% (x3 multiplier)

There needs to be a 35% mode as that is where the S2+ tends to thermo-regulate.

That depends heavily on the emitter used. A super efficient emitter like SST40 is going to sustain a higher drive current than a lower efficiency emitter like an SST20 4000K. Since sustained output varies based on emitter efficiency, I don’t think that should be a deciding factor when determining output levels.

Actually they both regulate at approx. 35% in the S2+ but have different drivers suited for the emitter’s max output - 5A and 2.8A respectively.