Roche F6 hacking

Post # 106

Do they really have the good version of F6?
They say it is without memory and in the pictures the “low power warning” lens can be seen.

Where can I buy this strap separately?

I only care about the voltage indicator window thing because the fancy replacement driver has the provision for two onboard LEDs, and with a easily reflashed controller and two LEDs you could make it do some very interesting things. Well, someone who understands how to code for these things could, though I am not that someone. I only do the nuts & bolts stuff.

All those sites are still using old stock photos from the very first 'good' version of the F6, the pictures are in no way related to the current-production light you will be sent. It's impossible to know what you will get from any of those guys. The pictures never change, the listed specs never change, if it looks kinda similar on the outside that's good enough for them to pretend it's still the same light.

Have you installed that custom driver yet? How much ma output are you expecting? I’m going to remove all the resistors on the one version driver tonight and replace with solder and see what happens.

I haven't measured this one directly but it's the same parts and firmware as the other BLF-DD drivers, and those do 6.5-7A to a single XML2, and over 10A into a triple XPG2 (using good cells, 25R or VTC5). No reason for this one to be doing anything different. I have 3 F6s completed using the new driver, 1x dedomed XML2/8* TIR, 1x XPG2/stock reflector, and one triple 219B.

Oh wow, have you fired up any of these new creations of the F6? 6-7 amps is plenty of fun.

Of course.

The triple gets hot at about the same rate as if the LEDs were replaced with a solid piece of copper creating a direct short... I suppose I could make an extension handle that screws into the tripod mount on the back, that might help prevent blisters. :p

Any idea how well the F6 fares underwater? I’m still looking for a reasonably compact 1x18650 light with e-switch and a flashable driver for use in a wet environment… er, actually, part of a water fountain.

I tried to get a ZY-T29, but it’s apparently discontinued. Still looking for other hosts to try. The F6 always looked nice aside from being a bit cramped for modding, but if there is a replacement driver available it might be a good option.

I bet they get hot fast now. Is it just me or is there a strange green glow from the middle F6?

Did you miss my reply in post #78? http://www.fasttech.com/products/1674603

That's the dedomed one, it's the original that came in one of the Banggood F6s, so the actual bin/tint it started as is unknown.

… yes. :frowning:

*puts on dunce hat and sits in corner*

I’d lose my head if it wasn’t attached.

Heh :)

There is one trick to those, do not attempt to remove the switch PCB, it's glued in (looks like that Fujik crap). Should already be sealed up well though.

I removed all the resistors and replaced them all with solder on the driver pictured above. I see about a 15% increase in output when comparing side by side with an unmodified version. I cant figure out where to test current on them.

I think I have decided on how to make the indicator LEDs actually useful...

With the stock drivers the indicator doesn't do anything until it turns red, at which point it's already low. It doesn't give you any info you couldn't get some other way like the light being noticeably dim. That's not very useful!

What I'm thinking is to have at least one LED on at all times when the light is on. Say LED1 on until 3.6v, both LEDs on for some as-yet-undetermined third color from 3.6 to 3.3v, and LED2 on below 3.3v. The firmware still has the normal warning flash/rampdown to deal with what happens at the very end. I think it'd be nice to be able to check the indicators and get information about what the actual voltage is at any point.

Any other ideas/suggestions?

Holy crap the stock spring is horrible!

I just compared a battery tube still using the stock spring to one upgraded with the copper disc/new spring, using the same cell & head (25R, F6DD driver, single XPG2 S2 1A). Not dark in here, baseline reading was 4.7 lux.

stock tail spring: 15.2 lux

upgraded spring: 20.1 lux

No changes other than swapping battery tubes. That's shocking!

edit: the triple 219B version is even worse... 24.1 lux vs. 39.9 lux

Is that the purple board goodness right there?

Sure, same board I used before I just finally got the indicators hooked up.

That one has the standard STAR momentary on it, and obviously the code doesn't do anything with the internal pulldown resistors on PB0/PB4. LEDs in that pic are being powered by phantom voltage on those two pins unintentionally.

I think the resistors will be something like 470R for the green, and 220R for red. I first tried both of them with 470R, and the green was WAY brighter than red. The red has a much lower Vf than green, but the green is much brighter (both actual output and apparent - your eyes are more sensitive to the green). To get them sorta balanced when they are both on in the voltage range from 3.6 (where the red one turns on to make yellow) to 3.3 (where the green turns off to go from yellow to red), the red needs a higher drive current. They're totally safe and within specs with those resistors even with pins 3&5 jumpered straight to Vcc (or the PWM at pin 6) and input voltage all the way up at 4.2v. Since red will only be used from 3.6-3v it could get away with an even lower resistor, around 150 or something.

Wow. The stock springs are stealing some of the potential of the driver. The tail spring looks like a simple solid steel spring.