Next gen replaceable light engines?

LED Flashlights – General Info
1

I really like the P60 concept. Buy a quality host and keep upgrading it over the years. Simple, toolless, works for regular users.
But P60 has aged. Lights became too powerful and P60s poor thermal transfer can’t compete. The top-notch light engine makers tried to compensate and had some success - but not enough to compete.
Furthermore there was another revolution - e-switches.

A light engine that would thread into the host would greatly improve thermal transfer.
If it got an e-switch connector it could support fancier UIs.

Overall it wouldn’t win a fair competition against a fully-integrated light. But it could come close enough to compete and the promise of longevity would win some users over.

Do you think there’s a space for next generation replaceable light engines in the flashlight market?

2

Kaidomain had replaceable drop-ins for their C8 and C8.2, but they vanished from the store.

3

I never saw the point of P60 drop-ins.
It is far less hassle to just have another whole light(s) built in a different configuration.

4

Hmm, good question Agro.

I hear you and I feel similarly about the the P60/D26 drop-in system, it has served well as a model replaceable platform. Thermal management can be an issue when driven at higher currents and UI options are limited. I also agree that the beauty of the P60 drop-in system are as you mentioned “Simple, toolless, works for regular users.” But if the P60 system were to get screw-in heads and electronic switches for fancier UI’s would it move away from its core value as a simple toolless platform?

While I might occasioanally mod a P60 drop-in to a FET-based driver and add a mechanical switch based Ramping-UI, the modularity means I can just change out (and back) to a PflexPro, Malkoff, KDlitker,… or even a Surefire incan… P60 drop-in without concern for thread size or fit (for the most part). Much more complexity than that might loose even many flashlight enthusiasts. But if what you’re referring to is next gen replaceable light engines generically, I agree there’s space and it would be great!

For something like that to happen, it would probably require either - (a) a manufacturer with enough presence to somewhat dominate the market to impose a standard (the way Surefire did early on), or (b) the various global flashlight industry players to come together to agree on some standards, or © the emergence of some undeniable next best thing that everyone would want to migrate to, or (d) a drastic slow-down of the changes in the emitter/battery/driver technologies, which for the timing being are continuing.

Currently it seems like each time something resembling a light format emerges (as far as enthusiats are concerned, such as tubular S2-based single 18650 lights or multi-cell C8-based lights, …) they only last until the next best thing comes along. In the past several years battery sizes keep changing, emitter sizes/tint/shape/optics/voltage, driver electronics, UI and ID designs,… keep evolving and each new design tries to take advantage of what the newest developments offer.

Yes there is space for next gen replaceable light engines but for now we are in a (wonderful) period of disuption and development. The best we might be able to hope for at this time would be some forms of limited standarization to improve interchangeability.

5

Is unscrewing a MCPCB and putting a new one in too difficult?

6

Less? Definitely. Far less? Not really….
And quality hosts tend to be expensive. Fancy hosts - even more so. Some users would prefer not to buy them over and over.
Fancy light engines are expensive as well. It may be worthwhile to have just few top-notch ones and adjust host to the current needs. Want small and light? 18350 host. Want runtime? 21700 could be the right choice. Want something fancy? The fancyhost comes in.
If optics were a part of the host rather than the engine, one could share the best thrower engine between a small reflector and a big zoomie host.

Actually I viewed BLF as a possible driver for the next standard. We have good contacts with several manufacturers. If there was enough interest in our community we could design the format and convince the makers to implement it.

Some of the things that you mention caused me worries. Some didn’t come to my mind.
Yes, a platform like that absolutely needs stable interfaces. It would be able to incorporate any improvements that preserve them but nothing else.

So…looking at past improvements as a guide to what could be incorporated w/out breaking a standard, what would fit?

  • LED upgrades
  • driver upgrades, as long as the driver is not too large
  • UI software upgrades
  • most host format changes, though possibly suboptimally
  • aux LEDs
  • built-in charger (but it would have to be separate from the light engine)

What would not?

  • UI hardware upgrades (more switches? Twisty ring? LED indicator not placed on the engine?)
  • most optical upgrades

Any others?

Really it depends…it is possible to design a screw-in e-switch engine with surfaces that your fingers can grip to, as simple and nearly as fast to switch as P60. But it limits some design choices, both of the engine and the host. It may be worthwhile to sacrifice some usability. Or it may not.

7

Do you mean that such light engine could actually be an easy to assemble LED board?
Or that LED and driver could be placed on a single MCPCB and treated as a unit?

8

For some of us, yes. I have neuropathy in my fingers, and soldering is really tricky for me now. Enough that I’ve given up modding.

Also, I would guess the vast majority of people don’t own a soldering iron or know how to use one.

Maybe you’re referring to some (potentially future) MCPCB that doesn’t require soldering, I dunno.

9

P60 is a dead end from my view

in my opinion a lights construction that does not use the normal soldering of MCPCB connections could be the next gen solution,
with also stiff connectors on driver, so you can easily swap them out
add some pins into the shelf that fit connectos on the MCPCB, then simply push the MCPCB on them and the LED swap is done

10

Add the ability to do the same for the driver, and I’m all in! Lego drivers and MCPCBS, would be awesome. E-switches and chargers complicate things, but I’m sure they’re all solvable issues by someone smarter than me.

I’m still waiting for universal(-ish) firmware flashing connectors on drivers, though, so I’m not holding my breath.

11

many hobbies have this problem: the enthusiasts age out, and not enough youngsters get involved to keep up.

12

pc_light

If surefire was the company they claim to be, this shouldn’t be that hard of a thing to accomplish. With what they charge for their products they should be able to walk and chew gum.

13

With maybe some adjustability to help raise/lower for optimal focus on the emitter side. AND perhaps a lateral circular clamp on the driver side to accomodate various driver diameters (even though the P60 standard of 17mm works pretty good.)

Nice idea Lexel.

14

While Surefire was definitely a leader at one point by being the first to adapt the cr123a format cell to make small powerful lights with great lumens to weight ratio. As well as the P60 lamp module that would fit between the various Surefire models, and eventually other brands. I doubt Surefire has any interest in establishing standards which would frankly make it easier for other manufactures/brands to clone/copy/compete. In doing so, they’d lose some competitive advantage.

I think Surefire is also more interested in reliability and fitness for robust field use over being on the cutting edge. Just look how long it took for them to adopt the 18650 battery or LED alternatives other than single (cool) white emitters, etc.

I think flashlight forums such as CPF, BLF, TLF, and Chinese sites serve almost like skunkworks for the industry, equally able to drive the needed innovation/standardization.

Now if we could only agree on the best tint or throw versus flood :wink:

15

I guess we agree, they aren’t the company they once were. I would disagree about loosing any competitive advantage. Without government buying their products they would be out of business. Most people can’t afford their products. I think more people would spend money on something they could upgrade, as technology advances. Why couldn’t they do both?

16

No. They should be designed to have LES at particular height. Emitter is lower? Make MCPCB thicker to compensate, so bottom-LES distance is constant. This will enable perfect focus in lens-based lights. Reflector lights need plastic spacers so…additionaly make the distance from LES to the surface that centering ring rests on constant. This normally varies with LED height so there should be additional shelf soldered around the LED. Centering ring of well-defined height sits on the shelf and perfectly focuses the reflector.

17

How does MCPCB diameter affect thermal performance?
I ask because now makers need to have different MCPCB diameters to clear different reflectors.

If it had no real impact on performance, with Lexels idea they could resort to one-size-fits all. Or at least one-size-fits-all-larger-than-X. Which would cut costs and also make it more feasible to order batches of PCBs for less popular LED footprints.
It could also save some weight.

Then - my idea increases costs. You really need a different MCPCB for each LED model. It also reduces manufacturer flexibility - when a new LED comes out they can’t just re-use their old MCPCBs. Note that reducing the number of diameters somewhat counters it.
Is it worth it? I’m not sure. I really like precise centering by design. Seems to go as far as possible towards user-friendly upgradability with 0 performance loss. But cost is higher.

18

P60 is a dead end from my view

in my opinion a lights construction that does not use the normal soldering of MCPCB connections could be the next gen solution,
with also stiff connectors on driver, so you can easily swap them out
add some pins into the shelf that fit connectos on the MCPCB, then simply push the MCPCB on them and the LED swap is done


Surefire and ArmyTech already does this….plug and play, drivers and mcpcb’s. Also their reflectors make contact with the DTP contacts on the mcpcb which are oversize and bare, not covered with masking, to help in thermal transfer. And they don’t use a centering ring, it helps with shock proofing the LED, which could sheer the LED off the board, during heavy recoil/shock. The reflectors are centered by the body/head of the light and the machining in the mcpcb shelf (Surefire) or Pill (ArmyTech)

19

These manufacturers can’t even get the threading to be consistent between versions and sometimes just between different colors of a model so do not expect them to ever be able and or willing to agree on a standard.
Not gonna happen.

P60 was nice in that there was no threading to muddy up the process but that also meant there was no heat transfer.

Personally I would like to see just heads sold with different combos from the same maker rather than have to buy another entire light, but that seems to much of a reach as well.

So it goes.
At least most of these lights are relatively cheap.

Later,
Keith

20

Any pics? I tried to look up Armytek but found only a few pics where leads are soldered. But the come through rivets - maybe that’s what you mean? Also - yeah, there’s no centering ring.

Actually now that I think about it - centering ring is not needed. Correct profiling of the reflector base is enough.
And this opens up a good way to do good thermal transfer to the reflector. Which may enable improving thermal transfer to fins in the front of the head. And also add some thermal mass, alu stores heat well. But at the same time - heating up reflector base will make it expand. It will be hotter than the front so the expansion will be uneven which will affect reflector shape. Is that meaningful?

I would not be concerned about mod-unfriendly makers. Yes, they are unlikely to adopt any kind of standard. I don’t think their participation is necessary.