First use of lathe - P60 pill for L2T

I'm building some lights for some local police officers, and while I liked the Solarforce L2T host, hosts have come a long way from the P60 design for better durability and heatsinking. (actually, durability is just fine I'm sure, but with the whole copper wrapping and the spring for continuity, it just can't trust it) I lent an officer a Convoy M1 and L2T to test out for a few weeks, and he really liked the L2T. It has better grip, nice anodizing, and a more accessible rear switch. He had me sold.

I spent so much time thinking about custom pill options. I really would be fine with just using the P60 pill as-is, but using a slice of a copper pipe instead of the spring like many have done here, and then wrapping the reflector in copper. But I ended up going with something really basic, which should hopefully have good thermal performance.

All I did was machine a pill so that it's really snug in the L2T, almost press-fit, and it bottoms out in the shelf so it should provide good heatsinking and a good resistance free electrical path. I then machined down the threaded end of the reflector so that it rests on the pill at the right focus.

The head is then used to hold everything in. I did the math, but I still think I got really lucky in machining everything at the right height so that there's not much of a gap between the head and the shelf on the body.

Oh yeah, the driver is the standard 105C (2.8A) with two 380ma 7135's added, and a 17mm to 17mm press-fit brass adapter ring. It's using the STAR_off_time firmware as you can see the capacitor soldered on to star 4. Basic two mode Turbo and Low, no memory with a ~250ms off-time mode reset. This way the officers can use it as a momentary light to "blip" the light on people/search areas without having it switch off of max output. You have to be pretty deliberate about it if you want to access the low mode.

The turbo is obviously set at a PWM level of 255, but starts ramping down immediately one PWM level every 250ms until it hits 140 (so about 30 seconds). This is to manage the heat. And honestly, I couldn't even tell that it was ramping or had ramped down, so I had to hook it up to an ammeter.

The major downside to this is that each pill has to be custom machined for each specific host, due to slight tolerance variations. The whole point of the P60 was being able to buy new modules to drop-in, but these are being built as complete lights.

This light was my first prototype. Next up I want to turn down the reflector body so that a split piece of copper pipe can wrap around it and fit snugly in the body. I also want to polish the inside of the L2T and the outside of the pill. I will have to polish the pill while it is on the lathe though, as I need to test fit the host while it's still on there (3-jaw chuck that has a pretty bad runout).

Hopefully reading all of that wasn't a waste of your time, as this is such a simple build compared to what a lot of people are doing on here :)

- Jon

Nice!

Man can you make a Convoy S2 pill w/ threading that is an all-in-one pill/spacer for use with a 3*XP-G2 noctigon and TIR?

Gotta learn how to cut threads sooner or later :slight_smile:

Awesome build man…

Looks like very nice work. I have one question… Care to share your code with the turbo ramp down instead of one step down?

This is really the only logic change off of my standard STAR program, other than changing the WDT from 500ms to 250ms. MODE_TURBO_LOW is defined at the top as 140.

ISR(WDT_vect) {

#ifdef MODE_TURBO
// Turbo mode is always at end
if (mode_idx == (mode_cnt - 1) && PWM_LVL > MODE_TURBO_LOW) {
// Step down
PWM_LVL = PWM_LVL - 1;
}
#endif

}

Welcome to the never ending spiral of depravity. I don't believe there was any luck at all in your machining and love the mod. One of my first mods was very similar to your own which a member here has. Don't keep us in suspense. Where's the list for future mods and flashlights to be made?

Now that’s cool…a turbo ramp down, as soon as it hits turbo it’s automatically ticking down to “high” rather than having a timeout…most cool!

L>M>H>TURBO>>>>>>>>>>>>>>>>>>>>>>>>>H>TURBO>>>>>>>>>>>>>>>>>>>>>>>>>H SWEET

I digress…back to the sweet sweet chip making :smiley:

Nice build, and nice ramp down mechanism which gives both maximum brightness and good heat management without the harsh stepdown.
I also have a c8 in which I made the resistormod for the nextmodememory and I used a so low value that it is basically a one mode light, but with the ability for the other 4 modes, this makes it really noobsafe. I need to tap on the switch really fast to switch modes.

Wouldn’t it be easier to just mill down the inside of each L2T to a specific measure instead of making every pill to another diameter…

I’m planning to add this into my standard off-time code layer tonight, one question tho how do you change the WDT?!

I wonder if you could just re-do the head and the pill as one piece. Keep the threads the same for the bezel and for the body. Then have reflector or optic held in place by the glass and bezel.

I’ve been working on some p60 drop ins for my L2Ms, but it never occurred to me to do it like this. Quite brilliant, I’ll be giving it a try for sure.

That sounds really complex, unless I'm understanding it wrong.

[quote=Cereal_killer]I'm planning to add this into my standard off-time code layer tonight, one question tho how do you change the WDT?! [/quote]

Download the Attiny13A datasheet and search for WDTCR.

For me, I changed...

WDTCR = (1<<WDTIE) | (1<<WDP2) | (1<<WDP0); // Enable interrupt every 500ms

to

WDTCR = (1<<WDTIE) | (1<<WDP2); // Enable interrupt every 250ms

Ha, yeah, it really would be, but I don't really have a way of mounting the L2T in the chuck. The spindle bore is too small, and I would probably need a 4-jaw for the four sides of the body even if it could fit. I now know why ChicagoX went to a much larger lathe, other than doing Maglite work.

It has begun, duh duh DUH! Going by your fantastic first effort, I see many awesome creations coming from you in the future :smiley:

Good job building the LEO lights,
here in Holland, the LEO’s are also doing research for more tactical lights and designed their own
There is a thread over here at CPF

I also have a few questions about the fw,

As i understand you have only these 2 modes defined ?

So your LOW is 140 ? or you have other low mode defined ?

This behavior of going from low, med to high to turbo is something i saw in the original star1.1
Most lights normally i see going from MED to TURBO, then back to a HIGH with the turbo timeout.

Is it possible to do this in this firmware somehow ?

ie a real LOW which i would set to around 50 , and a TURBO which throttles back to 140 ?
So you go from 255 to 50 by clicking the switch ?

And as you say it doesn’t do memory, why use the off-time firmware and have to solder the capacitor ?
As for mode switching, i ve never seen a LEO use its light for that short periods (less than 250ms)?

I have a low and turbo, but also have the lowest level to ramp down to from turbo defined.

As for off-time, it's good to know that when the officer pulls it off his belt, it will always turn on in turbo. If he accidentally activated the switch earlier, with on-time it will come on in low.

Well, a huge miscalculation on my part for this project. I am machining out 16mm pockets for the MCPCB thinking the emitters should always be centered on them. Well, I got a batch of 20 Noctigons and half of them are off-center, by quite a lot. Looks like the stamping machine that cuts them out was a little off. Since I don't use the reflector to center the LED (it floats above it), I'm going to have to have to machine a larger pocket and then make a centering tool that will be used to glue the MCPCB in place.

Damn, maybe I should have went with IS copper boards instead, although those might not be completely centered either :(

Our first batch of 16mm Sinkpads were the same way... some were perfect but some were a good bit off, depending on which part of the sheet they came from. They got the scaling wrong on one or more of the layers and aligned them at the top left, by the time you got down to the bottom right they were way off. Still usable but like you say, really only practical if you're using the reflector to center them.

that sucks. Any way you can glue the stars to the end of an arbor and lightly turn them so that they’re centered. Admittedly you’d then have to machine each pill to the exact diameter of the newly centered star, but at least they would be centered. See, this is what happens when you’re using a precision tool instead of your teeth and fingernails - the precision of other stuff suddenly becomes important!

That still happens for sinkpads. It is a pain in the rear to get a panel like that.

Or when the solder mask is off and you inadvertently solder to the main copper disc when reflowing emitters. Arrgh!