Reflector hole is 7mm, and it seems like intl-outdoor sells 7mm gasket with white flat. yes those leds are connected in series. only thing i haven’t measured is space between reflector base and dct board, since i will need to solder some wires there to connect leds. i guess 22-20awg wire should be fine, will see.

hopefully experiment will be successful, i just want to narrow the beam and decrease the glare for inc traffic. previously i was modding BL70s with xhp35hi with aspheric lens, but beam intensity wasn’t enough so bought these lights instead.

anyway, thanks for the help guys.

I have also posted this in the L2/L6 modding thread, but this might be just the more appropriate place to ask :wink:

So i took the gamble and just ordered the L21a to build a nice thrower
Together with some high drain 21700 Shockli cells and the famous Led4Power 3030 MCPCB with OSRAM KW CSLNM1.TG and another OSRAM KW CSLPM1.TG

I am planning to keep the original driver, but still doubting which LED to use, @6A the 1mm might be just overdriven and the 2mm might be just underdriven.

What are your opinions?

I would go for the 1mm and remove the spring bypass.

If you take my test of the 2mm2 White Flat as guidance, at 6A you are at 120% of the maximum rated current and 93% of absolute maximum output, is that what you call underdriven? :open_mouth:

The 1mm2 gives less output and a bit more throw. The driver is regulated at 6A? You may want to look at the convoy store, they have a similar driver regulated at 5A if I remember well.

Ok Jos, you convinced me :wink:
I’ll try the 2mm, otherwise i can still swap in the 1mm to see the difference and burn the S*@T out of it

Since when do we play safe?

Good luck with your build :wink:

The 5000mA driver from Convoy gave me 4570mA on its highest setting while powering the 1mm white flat, so i guess it is not quite 5A driver which is perfect for the 1mm led.

Nice info, thanks! I have a few of these 5A drivers underway for this purpose so I’m happy now :slight_smile:

That is similar behaviour to that of the 17mm / 22mm driver for SST40 (weird firmware, max 8A). One of my ∅22mm units stock delivered ≈7.63A, and another ∅17mm one (with swapped sense resistor for 5A, R010 + R050 + R200) was delivering 4.75 - 4.8A according to my precision PSU's amperimeter. In essence, 95% of rated maximum values.

The latter is paired with an CSLNM1.TG (1mm²) and I am happy with it.

BobbyMK if you want to push that a little bit more the math is easy. V = I × R = 4.57A × 10mΩ = 45.7mV sense. Stack an R100 on top of the stock R010 and you should get 10% extra maximum current, 5.027A.


Yes but 6A is perfect for the 4040 CULNM1 ;)

Stupid question, but is that just the same led on a bigger die?

  1. No such thing (edit: "no such thing as a stupid question" is what I meant here, I re-read this later and decided to clarify)
  2. According to Led4Power earlier in the thread, essentially yes. There may be slight binning differences (flux and/or Vf) between them, or at least between the batches he got.
  3. Bonus points: Boost HX is just the 2mm2 white flat, on the larger die.

Thanks for your reply, maybe i should have asked earlier, not after ordering :person_facepalming:
That 4040 CULNM might be just what i was searching for

I see two drivers on Simon his AE for SST40, one 20mm which should deliver max 6A and one 17mm which delivers 5A max.

Are you mentioning the 17mm driver?

(The L21a standard uses a 20mm driver, otherwise you need an additional retaining ring adapter for a 17mm driver)

There's more:

  1. ∅22mm, 4-modes, max 6A P-channel MOSFET linear driver.
  2. ∅20mm, 4-modes, max 6A P-channel MOSFET linear driver.
  3. ∅17mm, 4-modes, max 5A P-channel MOSFET linear driver.

I did not tried those.

Tried these: 17mm / 22mm dual ramping plus special mode max 8A N-channel MOSFET linear driver

Cheers :-)

Thanks for the info, its good to know that these are capable of more current by changing/stacking sense resistors.

I’m talking about this driver.

I use this driver in B158 and paired with the CSLNM.1 it has 370kcd throw, last time i measured without driver and NCR B battery to limit current it was about 380kcd. I think this is the sweet spot for this led.

I wouldn't go overboard with it. As you increase current the dissipated power at the MOSFET and driver increases too. Bear this in mind.


When will the boost hx led be avaible?

No idea

That is a Sofirn C8S host with Simon's 22mm SST-40 (max 8A) linear driver. Driver filed down to a tad less than ∅21mm and modified with an additional (slightly contact worn) R050 sense resistor. Measured 8.1A+ max current with power supply.

I am getting 132-133Kcd at 30s, 137-138Kcd peak, at close to 5 meters of distance using ceilingbounce app. Measured the same way I did with M1.

I measured ≈150Kcd for this build weeks ago, with an old reflector I swapped due to some minor damage. Driver was not modified at that point (I measured ≈7.63A back then).

Smartphone is a Lenovo Zuk Z2 Pro (Z2121 with custom Oreo 8.1 ROM). Do not know how good or accurate it is measuring illuminance, albeit I know it measures less than a Moto G 2013.

But heck, the intensity is way off I think. It should be getting at least 180Kcd… should be. The reflector's opening is ∅9mm, and is a little bit taller than the usual Convoy reflector (≈32.2mm).

Geez! :(