Wow. Your mismatched set means weāll have at least one other member with a mismatched SS/Cu setā¦ but it could birdnest further quite easily. Its going to get interesting.
I have read that many times already, and it just didnāt sink inā¦ looks like mixing was unintentional, but reported and expected.
I am so envious of the members that had the steady resolve to examine the shipping options. I did not, and according to tracking, for the last two days, my set still sits in Shenzhen, still āpreparing for shipment,ā even though it already shipped that far. My guess is the tracking is not quite accurate.
Youāre welcome.
Initially I started the video and couldnāt get it to go into config mode, I was clicking too fast.
In the same way that the reverse clicky is a fairly long press and hold, because of that we canāt click too fast to get the configuration mode, the off time cap isnāt registering the delay at faster speeds.
Weāve had to adjust to this kind of difference with every sample set, they keep making some kind of change that alters the timing and this time, the final time, itās not real obvious what they did. Iāve been using this firmware of TKās for quite some time now, Iām used to a half second delay so these are really different. Oh well, theyāre deliberate anyway.
Just remember that it does not have to be the exact amount. Just click as many times as you want and however you want. 20-25ā¦ Once it passes the set point you will be in config. A light touch at the edge of the switch is responsive.
But seriously. If anyone has a better idea how to scroll through the menu options with only one āpixelā and a reboot switch, lemme know. It can be a bit tricky, considering that every button press literally reboots the light. So, achieving any sort of continuity of logic flow across button presses is ā¦ awkward.
Okayā¦ Since people asked, I made an estimate of how many lumens to expect from each level in each mode group on the production lights.
Note that these are (mostly) calculated, not measured. But they should be fairly close:
1400 lm
5, 1400
5, 293, 1400
5, 135, 543, 1400
5, 79, 293, 688, 1400
5, 54, 178, 419, 813, 1400
5, 47, 135, 293, 543, 858, 1400
5, 35, 99, 213, 365, 613, 953, 1400
ā¦ and muggle mode:
5, 163, 688
Moon on mine measures around 0.30 to 0.35 lm, and the tailcap measures around 0.03 lm.
ā¦
Note for my own reference: lumens estimated with: level_calc.py 64 1 10 1400 y 2 0.2 140
(I measured low at ~5.3 lm, then measured a few more modes, and found that this curve fit best except for the lowest mode)
It looks like youāve got a hardware problem. Itās likely that the OTC is shorted or something, so that it always thinks you did a long press no matter how fast you tap the button. Do you have the ability to open it up and look at the driver? Youāll probably need to open the bezel, un-solder the leads from the emitter star, then remove the driver from the back. (Iām assuming they didnāt use long enough wire to access the driver while itās still attached)
It may be something you can fix with a quick tap of a soldering iron, or it may require a new driver. Hard to say without actually seeing it.
Oh, this reminds meā¦ something worth mentioning:
Muggle mode does not have the full config menu. The only config option in muggle mode is āexit muggle modeā.
So, if you enter config mode while muggle is activated, it will blink once, buzz, then return to regular output. Click during the buzz to get out of muggle mode.
Nope, the production units ended up one firmware revision before that landed.
I prefer group 9 also, the ādjozz specialā group, but itās not there. Because of this, Iāll likely be reflashing all of mine even if I have to un-solder a few parts to get flashing to work.
If drivers are sold by themselves later, like, if another batch gets made, Iām guessing theyāll probably have the newer code on them.
The diff shows the exact details, but the short version is that the two closest levels at the time were:
7135 at 249/255, FET at 0/255
7135 at 255/255, FET at 1/255
None of them settled exactly on:
7135 at 255/255, FET at 0/255
The Nov2015 revision changed this to ensure one level would fall exactly on full 7135 with no FET, and then assigned that level in a few places.
Visually though, itād be extremely difficult to tell the difference. The PWM runs at 15.6 kHz, too fast to see unless you look very closely with a thin fast-moving white card with the output set between ~5 and ~120 lumens. Above 140 lumens, the emitter never actually shuts offā¦ so the PWM toggles between bright and brighter. This is nearly impossible to see without specialized equipment.
Ok, please keep this as a thread, I am neither a EE or a firmware programmer, but it would be like watching a rocket being built before launch. Fascinating, and while I will not understand everything going on, it will be amazing to watch (read in this case). It would be worth the time, even if my role is just a monkey watching this collaborative ballet of talentā¦.
Just my thoughtsā¦
ToyKeeper, this UI is fantastic! Thank you sincerely for your labors on this projectā¦