My Convoy L6, lighted switch, TA FET driver plus mods

This is going to be a thread on my L6 just because it’s gone through so many changes I can’t keep track. Lol

First up is amp draw measurements with the stock driver and various resistors.

I received my new Uni-T UT210E so my current measurements will be in line with others here. I used a 6” length of 10 gauge wire on the tail cap. I measured both the Liitokala and KeepPower 5200 protected cells. Emitter was the xhp70. I only tested each for maybe 5 seconds so I could do a lot of tests without having to fully recharge the batteries each time. The hundredths place was rounded to the nearest 0 or 5. Springs are bypassed with mesh copper braid used for desoldering.

First up is the stock FX-30 driver with the 2 R082 resistors, no mods.

Liitokala
Turbo/high = 4.60 amps
Med = 2.55 amps
Low = 0.55 amps

KeepPower
Turbo/high = 5.10 amps
Med = 1.90 amps
Low = 0.45 amps

Kind of strange, I double checked it. Does anyone have a answer as to why the Liitokala were so low on turbo compared to the KP, yet the lower levels seemed normal? This is the first time I paired my Liitokala’s with an unmodded driver so the results are surprising.

Next is with an R100 resister stacked on top of the 2 R082’s.

Liitokala
Turbo/high = 7.00 amps
Med = 2.60 amps
Low = 0.55 amps

KeepPower
Turbo/high = 5.70 amps
Med = 1.90 amps
Low = 0.45 amps

Next is with no added resister, but added a wire across.

Liitokala
Turbo/high = 7.70 amps
Med = 2.70 amps
Low = 0.55 amps

KeepPower
Turbo/high = 6.10 amps
Med = 1.95 amps
Low = 0.45 amps

So 7.7 amp is the most I can get from this driver with the xhp70. My borrowed meter read 8.1, but all it’s readings were a little bit higher compared to the Uni-T meter. Also, all my readings on the next lower power level showed the Liitokala as 0.70 amp higher than the KP.

I’m going to stick with the R100 mod for the time being. Once I get my lux meter back I’ll do some comparisons between batteries to see how much brighter just a battery swap can be. I’ve also got some R082 and R120 resistors I can try, but I doubt there will be much difference over the R100. Plus it’s a pain soldering those things. :confounded: Maybe a smaller soldering tip would help.

After modding the stock driver I bought a non-Convoy FX30 driver from Kaidomain. Not the Convoy version, the original.

This has a unique interface not described on the website. The tail switch puts it into standby and the side switch only has 3 brightness levels plus a strobe. You can turn the light on and off from the side switch. Very odd. I made a quick video on YouTube showing how it works and went into more details on it.

I just installed it and was quite surprised. For those folks who want a regulated output, non direct drive, along with the ability to turn it on/off from the side switch, this is the only driver I know of. It’s mode spacing is not very good, but it’s not that big a deal to me.

I did a direct wire across the sense resistors so I’ll see how it holds up at work tonight.

Next I wanted to do a lighted side switch.

I found a source of positive power for the LEDs.

I got a good reading here, but I might have been able to use the main red wire to the led since I think this driver feeds battery power directly to the LED and the negative lead goes through the driver to control output. I never did confirm that.

The switch looks like it uses an 1206 2012/0805 sized led and the 2 bare wire pads on back go to one of each led pad. The other led pads seem to link together and connect to the blue wire. The blue wire is a ground.

I could probably add a floating resistor into the wire from that positive source to either the purple or yellow pad on the back of the switch.

It seems these leds vary from 2 to 3 volts depending on color. I found using a 15k ohm resistor worked well, but a bigger one is fine, too. It’s just a bit less bright.

I experimented with slicing the dome off the xhp70 and we’ll as trying out the xhp50.2. I also swapped to a SMO reflector and cut the bottom out then clamped down the emitter.

I did a ceiling bounce with the FX30 driver with R100 resistor added. I measured using a sliced-dome xhp70 and a dome-on xhp70. Both are N4 1A. I also compared the Liitokala against the KeepPower 5200 protected cell. I’ve already shown the Liitokala pulls more amps (7A) compared to the KeepPowers (5.7A) and here is proof they put out more light as well.

Sliced dome
KeepPower - 343
Liitokala - 377

Dome on
KeepPower - 400
Liitokala - 460

Crunching numbers:
The sliced dome looses you 14% light output on the KeepPowers and 18% on the Liitokalas. You loose lumens, but gain a more intense focus which means longer distance. Slicing the dome also helps to reduce the donut hole some.

The Liitokalas give you a 10% boost over the KeepPowers with a sliced dome and a 15% boost over the KeepPowers with the dome on. You can see the brighter output.


I cut the bottom out of the SMO reflector and fabbed some very thin copper shims to play with the emmiter to reflector distance. I also swapped to new, longer screws to clamp down the emitter for good heat transfer.


Here’s a quick comparison of the Convoy L6 with SMO reflector and xhp70 with the dome on and with it sliced off.

The sliced dome looses you 14% to 18% of lumens, but you gain a more intense focus which means longer distance. Slicing the dome also helps to reduce the donut hole some.

The smaller emitter size with the dome removed directly translates into a smaller hotspot.

All my throw numbers are in post 8.

I just finished my lighted side switch. A complete success on the first try!

This light has the FX30 driver from Kaidomain with the weird firmware that let’s you turn the light on and off with the side switch. The only problem is telling if the rear switch is on or off. I’m always double checking it. Now I always know plus can find the switch in the dark. No more fumbling for it! Yay!

First I filed out some notches to the switch hole to allow easier driver swaps. Kudos to Kawiboy for the suggestion. Note that the top and bottom lips are very thin, but the sides are very thick. So it’s very quick and easy to file in the locations where mine are.

Then I took 2 white SMD LEDS, kudos to DBCUSTOM for the LEDs, and added them like in my diagram above.

I bridged the rear connector to get power to both LEDs. The resistor is a 15K ohm. I believe that’s equivalent to 7.5k ohm per LED. It’s bright enough, but not too bright. The amp draw went from 1.45 milliamp to 1.85 milliamp. So no big deal for me.

Here’s a video of it. I’m a bit long winded. Lol

The clear switch covers are from Convoy. Simon uses them on the clear version of the L6. He will send you some only for the cost of shipping (a couple dollars) or if you buy something he can add them to the order for free.

The LEDs I got from DBCustom. I’m pretty sure they are sized 0805/2012. They are cheap if ordered from china. EBay has lots of them.

At first I thought 1206 leds would fit, but as you can see, they are a little big for the switch pads.

Normally it doesn’t matter which switch wire (red or blue) goes to ground and which goes to the “switch signal” part of the driver because it’s a simple contact switch. When you add LEDs, the polarity of the switch wires suddenly become important because the LEDs are polarity sensitive and share a ground wire with the switch.

Since the blue switch wire is a common ground (for 3 things, the switch, the left LED and the right LED) you want to make sure it’s going to a ground source. If it’s not, you need to swap the switches red and blue wires around where they make contact on the driver.

If you switch wires were backward then that might have been what caused your problem. You might have fed 3 or so volts into the switch positive trigger when its designed to only sense ground.

Polarity of the LEDs is also important. They need to be like this.


Now if your using a different momentary switch you’ll need to map out its traces on the pcb. I just used the continuity circuit on my multimeter.

Here is the switch from the FX30 again. All the blue dots go to the blue wire. Red dots to red wire.

I then ordered a Texas Avenger FET driver from Lexel.

I just got my TA driver installed. Everything works including the lighted side switch! Yeah, I’m 2 for 2. :smiley:

It’s pulling 12 amps with the Liitokalas at 4.1 volt. Yikes! It might be higher when fully charged. Maybe I should NOT have used 18ga wires? I need to tone it down a little.

The scary thing is after about 10 or 15 seconds I see the light color start to shift towards blue/purple. I quickly turned it down in brightness. Maybe the 20mm DTP copper star is having trouble with that heat? (EDIT, it was my copper shims, not the star or LED)

I tried my KeepPower protected 5200 and got 9.6 amps at 4.2 volts. No color shift seen.

For those that want to use a lighted side switch on the TexaS Avenger FET driver in the L6, here is the basic layout.

My “power on” led resistor broke in two when I tried to put test probes on it to measure the resistance. I hope that’s not a sign that there was an issue with the solder paste. Update, I broke the resistor, paste was fine.

Since I’m running 2 leds in parallel I’ll swap that 4,700 ohm resistor for something double in size. 10k to 20k should be fine. I used a 15K resistor.

BTW, I swapped the white leds to red leds to the switch this time around just to be different. They are not as bright as the whites, but still okay.

Simple little video here for anyone interested.

I swapped the switch leds out again and made the resistor less cumbersome.

This section is about the SMO reflector and throw numbers.

I decided to see how much throw I could get. I bought a SMO reflector to experiment with. I shined it on the ceiling with the stock centering ring and then I removed the ring and added some insulating tape to the bottom of the reflector. This allowed the reflector to sit lower which changed the focus. You can see in this picture how lowering the reflector just a tiny bit reduces the size of the corona and puts that light into the hotspot which increases the lux.

I knew this required more investigating so I cut the bottom off the reflector and bought some new, longer panhead screws to bolt the emitter down.

I also experimented by slicing off the xhp70 dome. You can see that this shrinks the hot spot and increases the intensity.

I bought an inexpensive LX1330B lux meter and did measurements at 5 meters, at about 30 seconds in and with fully charged KeepPower protected 5200 batteries. I then calculated it back to 1 meter.

With the stock driver and OP reflector I got 58,000 lux or 481 meters.

Then I added an R082 resistor and got 65,000 lux or 510 meters.

Swapping in a stock SMO reflector gave 81,250 lux or 570 meters

I then tried a new xhp70 of the same N4 1C bin, but I got slightly better results.

OP - 71,500 or 534 meters. A 10% increase.
SMO - 86,250 or 587 meters. A 6% increase.

Next I cut the bottom out of the SMO reflector so I could adjust the focus better.
I got 92,500 lux or 608 meters.

Next I sliced the dome off.
I got 115,000 lux or 678 meters

I pretty much stopped there. Even though my measurements are lower than others, I did pretty much double my lux from 58,000 to 115,000.

Matt at Adventure Sport Flashlights did most of the same mods, but his measurements were all higher. He ended up doubling his lux as well, so I think we are about even.

Here’s his video on it.

One potential reason Matt got higher numbers may have been due to batteries. After all these lux measurements I bought some Liitokala 5000’s to compare to my original Keeppower protected 5200’s. My battery tests show the Liitokala cells always drew more amps than the KeepPower (7 amps compared to 5.7 amps). This extra amp draw results in higher lumen output. Maybe Matt was running better batteries than me? We did use the same model lux meters, as well.

My thrower version of the L6 with R100 added to FX30 driver and SMO reflector. This is before the FET driver. The stock light could never reach those warehouses. With the FET installed, it’s even better than these pics! Lol

Here’s a few shots of the xhp50.2

I tried it at 2 slightly different reflector heights. They are pretty similar. The 50.2 with dome on seems to match the size of the xhp70 with sliced dome. I’m guessing the 50.2 with a sliced dome would shrink the hotspot size again.

With the camera settings turned down you can see the dark center.

I probably should have taken a few lux meter measurements while I had the 50.2 installed, but I didn’t. I ended up buying a new xhp70 to put back in and I adjusted the reflector depth to give a bigger corona creating a smoother hotspot while reducing throw. I don’t need max throw, just a good compromise between throw and a nice looking hotspot.

Nice comprehensive write up Jason. :+1:

Thanks.
It’s way too much for me to remember so it’s mostly quotes from many other posts I’ve done. So the verbage might sound funny at times, but I’ll eventually go through and tweak it.

Good stuff, nice it has it’s own thread :+1:

Wow! Nice job on this! Just updated my Convoy L6 with the latest NarsilM v1.0 and all is working well. Ramping is doing good Pulls 18A, 186 kcd, intense hot spot, no black hole with SMO reflector, running the XHP70.2 P2 3B, 8,090 lumens @30 secs. This is with dome, but hoping to try sliced dome next.

Last night with 2 LK 5000's at 4.00V, it still gets 15A on a clamp meter. Last night I added solder blobs on the LK's. Before I was using a JAXMAN spring adapter between the cells, but the spring was bypassed with a short 24 AWG wire.

Just tried setting moon level to PWM value of 1 via the UI and it's wild - the XHP70.2 literally barely glows - noticeable but not measurable - amazing. You can tell it's on because the yellow of the phosphor turns slightly whiter than the phosphor on the LED base.

Looks like you are having fun with this light :partying_face:

Cool write-up for others who are modding this light. :beer:

Last night I charged up the solder blob topped LK 5000's to 4.20V and re-tested. Actually I use only one that's solder blob'ed since there's a driver spring, bypassed of course.

Amps at the tail: 17.3A, lumens: 9,210 @start, 8,160 @30 secs, and 193 kcd @5m (879 meters)

Parasitic drain is 62 uA (no indicator LED on this L6), so pretty happy with that low reading. No need to turn off the power switch.

Thanks to steel_1024 for the XHP70.2 P2 3B!

It's an impressive thrower, but way too much heat to handle, but that's what thermal or timed step downs are for.

I use Narsil v1.2 so I can have the lvp, but you can’t have that and thermal monitoring at the same time.

Maybe v1.2 has a timed stepdown?

That’s no big deal as I use my hand to monitor temperature. For normal use I adjust to slightly below turbo output. You still get bright output, but much less heat and battery load. That’s much more practical. Then as the voltage runs down I go back to full output.

I save the full output with fresh cells for testing or showing off. Lol

NarsilM has LVP and thermal step down for 2S setups. v1.2 had a poorer ramping table as well. Bunch of improvements in NarsilM.

Not sure I can afford the time to look back any more at old revisions, unfortunately.

The stock driver is a buck driver, tailcap amps are basically a useless measurement for buck and boost drivers.

Buck drivers keep a constant output current but the input current will vary with voltage. Higher input voltage = lower input amps and vice versa. So in this case the lower the amps, the better the battery.

Cool write-up Jason! When I want to mod my L6 I will find my inspiration here. The lighted switch will certainly be on the list! (and XHP70.2 and SMO reflector).

he got v1.2 with my improoved ramping table