Why aren't P60 dropin drivers more efficient?

Is it just my perception or do most P60 drop in drivers not even come close to matching the efficiency of drivers in retail lights known for their superior FLAT regulation and long runtimes, e.g. Zebralight?

Is the space constraint of the pill a reason? I have trouble imagining it to be while looking at my Zebralight.
I know it can't be a technical hurdle. Right?
I don't think anyone would turn down better FLAT regulation and longer runtimes, so it can't be lack of interest or demand right?

So what gives?

scs said:

Is it just my perception or do most P60 drop in drivers not even come close to matching the efficiency of drivers in retail lights known for their superior FLAT regulation and long runtimes,

Click to expand...

Are you comparing D26 (1" diameter) with 1.5", 2" or larger head size???

fivemega said:

Are you comparing D26 (1" diameter) with 1.5", 2" or larger head size???

Click to expand...

fivemega, D26, but please do elaborate on the other form factors as well. Thanks.

Constraint of the pill, but mostly because of isolated thermal relief. Have to keep in mind drivers produce heat just like LEDs. There's barely enough heatsink for a high powered LED, and you're piggybacking a heating element on it. This is also why companies like Eagletac, Malkoff, Oveready are making "oversized" P60 modules, ones that is way bigger than the lamps with the original "P60" designation. With a drop-in, you also take into account of additional series resistance, namely resistance across touching contacts.

That brings two issues
The efficiency of the LEDs in the face of accumulated heat
The efficiency of the driver feeding power from the battery to the LED

Many lights are made with proprietary drivers
Many dropins are made with generic/off the shelf drivers, like the 7135
High power dropins may also stack these same chips, sapping additional energy

Thanks for the replies, guys. Aside from the heat mentioned, which I'm guessing might be more than adequately managed with copper tape wrap and custom bezels like the Cryos, if the output is kept to no more than say, oh, 500 OTF lumens, it sounds like the underlying reason that cripples dropin drivers is their design, which as Dan mentioned, doesn't match up to that of proprietary drivers. That goes back to one of my original questions regarding technological hurdles.

I don't know anything about circuit design, but how difficult is it really for electrical engineers to design an efficient driver for a dropin?

I can definite understand the existence of proprietary technologies developed through intense R&D when it comes to complex machines and electronic devices, but for flashlight drivers? Aren't they simple enough (relative to the more complicated and sophisticated electronic devices) that everyone in the business and in the trade knows the "secrets."

How much exclusive knowledge about flashlight driver circuit design and manufacturing can there be and is held by retail light makers but unknown to those who design drivers for dropins? Do the dropin driver designers simply not know how to design better drivers, or there's no financial incentive for them to do so?

scs said:

I don't know anything about circuit design, but how difficult is it really for electrical engineers to design an efficient driver for a dropin?

Click to expand...

What looks like a technical challenge is actually an economic one. Think of producers in three groups:

1. To small to make their own drivers (most custom dudes)
2. Small enough to make drop ins + large enough to make their own drivers
3. To large to make drop ins (most flashlight companies)

The flashlight shops you're looking at (3) with good drivers make thousands of a model at a time. Some go to CPF but many go to stores around the world. This volume means an investment of $10k+ for a design can be amortized over 10k+ customers (less than $1 per).

But you can't sell drop ins at Home Depot or even REI. So a given model is unlikely to sell more than 100, never mind 1000, never mind 10k. Amortization of $100 per just for the drivers simply won't work, so you (1) buy whatever drivers can be had that the other drops are also made from, for the same reason.

That leaves #2. Which on paper looks like the sweet spot. Make 500 copies and that same design investment is only $20 amortized per copy (plus actual parts plus actual construction). But then work and risks of all sorts go up. And in the end, you've gotta charge at least twice what the other drop in shops are charging. And then you've gotta explain the benefits of efficiency so people are willing to pay extra for it*. Because one of those risks is that you won't sell enough to even break even.


*shop's lumen numbers are usually based on total power drawn from the battery, as if drivers were 100% efficient and all that power was hitting the LED. So for that extra investment to even show up on a cost/benefit for customers, you've gotta explain all that first.

That was an excellent exposition electronguru. Thank you for that!

Yes, great explanation, thanks.

Another factor in the generic driver efficiency wars is voltage range. If you are designing a driver for a specific light, you know what the possible input voltages will be, which will usually be a narrow range. Thus you can tune the driver for best performance in that range. That's why the user manual for my TM16, for example, says that I can use 18650 cells, but not CR123 or RCR123. I have a couple of D26 drop ins that will work over a range of 3 volts to 13 volts, so obviously the driver hasn't been optimized for a specific voltage range, thus less efficiency. I have a couple of other drop ins that very specifically state that they can be used only with a single Li-Ion cell, restricting what I can use as cells, and somewhat the market they can sell to. It all adds up. Personally, I have those two drop ins for reasons other than efficiency, and am perfectly willing to accept that trade-off to get the benefits I want from the lamps. YMMV

Sure thing gents.

BTW, I'm working on #2 as we speak. I don't expect even an efficient 3500 lumen triple to sell 500 units so I'm getting around the amortization problem by using the same driver in multiple products. Should be appropriately epic.

Thanks, Dan, for the explanation. It sounds like the bottom line is a more efficient driver not only requires a better design, but a more expensive design, and there just isn't enough demand for them to recoup the added material/fabrication costs.

I was under the impression that given the relative simplicity of flashlight circuitry design, it shouldn't cost substantially more to make something better.

scs said:

Thanks, Dan, for the explanation. It sounds like the bottom line is a more efficient driver not only requires a better design, but a more expensive design, and there just isn't enough demand for them to recoup the added material/fabrication costs.

Click to expand...

Yes and no. If everyone made their own drivers already, it would just be a question of making better drivers. Then the only extra cost would be [what is required for more efficiency]. But most P60 shops don't already make their own drivers. Most buy a box of 7135's and call it a day. So in practice its:
One option is maybe $100 with minimal development, the other is 10k (100x as much) to make a circuit design from scratch and then test it and then produce it. Thats a lot of additional work for a single feature that by itself is hard to explain, near impossible to actually see, won't produce 100x the sales, and is difficult to charge extra for.

I don't know the history of the 7135 but if it was made for flashlights, the producer may have 100k units on the shelf and see little reason to make them obsolete by making an improved version. If they were made for something else and adapted to flashlights, the producer may not need efficiency for the larger application.

scs said:

I was under the impression that given the relative simplicity of flashlight circuitry design, it shouldn't cost substantially more to make something better.

Click to expand...

There are different ways to make flashlight boards. A 7135 is a near complete part. On the plus side, this means cheap and simple. But on the neg side, this locks shops into what it can do and how. My v5 board is a set of different chips that work together. On the plus side, there is huge flexibility. But on the neg side, there are also huge logistical considerations putting it all together (plus all the costs and risks).

The central point here is that in most cases the people deciding to make what you're buying are not the same people who are deciding what the driver is capable of doing.