Host electrical resistance

Hello.
I’m digging resistance of hosts to do some ballpark output calculations. I am able to find the commonly modded components (springs, switches, drivers, wires), but adding them up skips some others:

  • battery contact points
  • threads
  • host body
  • LED PCB
  • lead wire solder joints
  • driver-body joint
  • other?
    Anyway, for what I know these elements may add up to something meaningless or to something substantial. I don’t have any data. OK I can do ballpark calculation of host body (I did and the result was very low), but with others I’m at loss. Did anyone measure or calculate resistances of entire hosts? Can you share the results?

Don’t know most of those off-hand individually, but total circuit resistance including a FET driver, but not including cell internal resistance might be around 10-20 mOhms, pretty low. That is with bypassed springs, a low resistance FET, and not-too-long 22 or 20AWG wire leads.

I’ve done some modeling of direct drive LED circuits; see here and here for some discussion.

Great, thanks

The things like tube and connection of tube to heads are so big, current can flow there really easy, I don’t think that will add significant resistance in a normal light

I tested the battery to spring resistance here and its a lot lower than the AWG22 bypass

Why not just use a constant current driver?
Direct drive is bad and inconsistent, definitely not what a good flashlight should have.

For turbo mode FET is great, otherwise very bad.
I haven’t decided on whether I want linear + FET with parallel LEDs or boost with LEDs in series, but I’m collecting the data (brightness vs. runtime curves) to make an educated decision.
And host resistance matters with all kind of drivers. It affects boost driver efficiency (and buck too). In linear, lower resistance allows better work very near the end of battery capacity.

But with a direct drive FET driver the lower resistance may fry the LED…

I believe it’s better to add resistance in form of battery IR than in host - because this way you can gain some capacity. At least sometimes. And if you do it wrong, you may fry the battery.
So yes, a FET setup must be reasonably designed. That’s all.

Regulation is always the proper way. Direct drive… LoL! Go dee dee your emitters with your car's battery, always fun to do!

In fact, I was recently suggesting some folks the convenience of boost-buck designs: drivers consisting of boost stage with adequate cell undervoltage protection logic plus fully regulated multiple modes buck engine. A li-ion cell neither is 3.7/4.2V, after effective loaded output from the cell and path losses a good driver should be designed to work right from as low as 2.8V (in my opinion).

What I love about FET is efficiency. It will have the highest current from any type of drivers.
In a sensible design, this current won’t overload the LED (s), they will still run fairly efficiently and the output will be better than from any other type of driver.
Normally, this highest output will be thermally unsustainable, thus turbo mode only.

Some drivers PWM the FET to gain some intermediate levels between the maximum they can regulate and non-PWM FET. I don’t find it good design; IMHO Turbo is to get the most out of the light. If it’s too overpowered to be useful, it’s better to add resistance than PWM. If the highest regulated level is too low, make it higher. Most lights can have a regulated mode that will exceed their thermal capabilities anyway.
BTW there’s drive to use the lowest-resistance FETs available and then regulate current with wires.
It’s the easiest way, but suboptimal. It’s better to use a smaller FET instead because smaller wires hurt your regulated mode.

As to buck-boost, I used to like them too. But now I think they are good only for 1 cell 1 led.
If you have more of either, you can configure the light to use either pure buck or pure boost. You’ll save money and board space. In most cases (or maybe all? IDK) you’ll improve efficiency too. Though it doesn’t discount them. Personally I’m not much interested in 1-led lights, but I see there are many who think otherwise.

You want a current limiter i.e. a constant current driver.

Sometimes yes, but for some current / driver size / cost combinations it’s either suboptimal or entirely infeasible. Here is an example. 15A that this light drew was more than the builder wanted. But limiting it to, say, 10A, wouldn’t be possible with a regulated driver small enough to fit.

How i see it:

Resistance in the battery will make the battery go hot, bad idea. (depends on the ambient temperature of course, you’ll get away with it in winter).
We all want the best electrical path for when the battery runs lower and we need every 100 mV to have the LED run 100%.
We don’t want the lousy efficiency of a heavily overclocked LED.

The exception would be XP-G2 dedomed throwers, because on 5 Amperes or more the Vf is higher than 4.2 Volts.
But in fact it would be better to run an XP-G2 on 2x 18650 with a buck driver like the FX-30 (Convoy L6).

I think the LD3 driver from led4power would be nice for regulated high current.
They’re 17mm and can do up to 12 Amperes regulated. (iirc).

I stand corrected. If LD3 supports 6V, it would indeed work better there.

In my opinion people take the best fet and then thin the wires on some builds. Is for the next upgrade or build they can just change the wires and emitters and not need different drivers or smd reflow. If your planing a one time build that’s one thing. A lot of people here change out LEDs at will when something new comes out. It may not be the most efficient but it works and still get great run time on low modes with fast pwm. Hours and hours on lower modes

Well, it’s just my opinion. :slight_smile:

I think it does actually.
But ideally you use a buck driver, because it doesn’t ‘burn off’ the excess Voltage.
But you always need some overhead Voltage for a buck driver.
But, the newer LEDs have increasingly lower Vf, so buck drivers will be the more interesting option in the future.
Iirc there are new Nichia LEDs with a Vf of only 2.4 Volts now.

And I agree with it. For this light LD-3 would be a superb option. Small enough, powerful enough, about the best thermal control on the market.
Though I guess a boost driver with VTC6 would be a better option than a buck with a pair of Aspires. GXB17 won’t handle this power though.