SST-50 DD - does this make sense?

[JohnF]

I guess I have a hole in my knowledge (many, really) figuring out current limiting resistors. I've got an SST-50 mounted on a heat sink, and tested current draw from a 26650 with a resting voltage of 3.9 v. Current draw with emitter straight to the battery was ~ 4a. With a .47ohm 5watt resistor in series with the emitter, total current draw was only ~1.4a.

Did I mis-measure, or if not, can you explain why this is so?

My plan is to use this SST-50 in a Mag3D direct driven from alkalines, and am trying to determine if I really have to use a series resistor, and if so what value would work best. I'm using the 26650 just because it was much easier... Might there be enough voltage sag, even with fresh alkalines, so that I'd be safe without a series resistor?

If it went from ~4a. to ~1.4a., with some loss (how much)? to the resistor, I'm assuming I'm losing a lot of output.

Here is another question that will show my basic lack of LED operation knowledge - if I stay below the 5a. that the SST-50 is rated for, even though it might be that from 3 fresh D alkalines at 4.5v resting, am I safe?

I will be using a dFlex, so I could use a slightly lower PWM level than max until the battery voltage has dropped a little to make it safe on max, right?

Thanks for any help,

John F

 

[Techjunkie]

Personally, I don't think it makes sense to drive an SST-50 from Alkalines, at least not at high drive currents. They sag like crazy with steadily decreasing Vout throughout the discharge. You could always add one or two more of the 0.47 ohm 5W resistors in parallel with the first one to reduce resistance until you find a drive current you like.

 

[JohnF]

how about using a HipCC at 2.4a regulated in conjuction with the dFlef with alkalines? Will that be more satisfactory? Or should I just plan on some other battery configuration other than alkalines? Thanks.

 

[ti-force]

Just like previously stated, alkalines just aren't a good choice for what you want to do. Use your 26650 with a couple of spacers or one custom made spacer to take up the space.

 

[JohnF]

Just like previously stated, alkalines just aren't a good choice for what you want to do. Use your 26650 with a couple of spacers or one custom made spacer to take up the space.

I just did testing with 3 D alkalines, and I see what you mean, totally inadequate for the SST-50. 1.2a is the most they would deliver.

I've going to make a getto 2x 26650 parallel adapter to fit in the 3D mag. This, I think, will work pretty well direct drive with an appropriate resistor if I decide not to use the HipCC (just sitting around anyway) which will limit current to 2.4a.

After playing on the bench tonight with a reflector taped to the britelumens heatsink and aiming it down the street, it seems the difference to my eyes between the SST-50 driven at 4a. and driven at 2.4a is pretty minimal. Much less heat is produced, so I think 2.4a might be the way to go. Even at 2.4a, that chunk of aluminum gets hot pretty quickly. The maglite body will absorb much more when the heatsink is installed, but still...

John F

 

[Aircraft800]

how about using a HipCC at 2.4a regulated in conjuction with the dFlef with alkalines? Will that be more satisfactory? Or should I just plan on some other battery configuration other than alkalines? Thanks.

It won't work, it would instantly drop out of regulation, the hipCC needs more headroom than that to regulate, it's best used with 2 series IMR26650 cells (not parallel) in a sleve in your mag. I don't know what a dFlef is, but on 3 alkaline D you could use a d2Flex, but you'd underdrive the LED alot. Alkalines just won't discharge fast enough.

Either way, 2.8A of a hipCC just doesn't do that SST-50 justice, get more like 4~5A to it DD or with a resistor.

 

[JohnF]

It won't work, it would instantly drop out of regulation, the hipCC needs more headroom than that to regulate, it's best used with 2 series IMR26650 cells (not parallel) in a sleve in your mag. I don't know what a dFlef is, but on 3 alkaline D you could use a d2Flex, but you'd underdrive the LED alot. Alkalines just won't discharge fast enough.

Either way, 2.8A of a hipCC just doesn't do that SST-50 justice, get more like 4~5A to it DD or with a resistor.

Sorry left out the '2' in d2Flex. Your way sounds easier and better, so be it. Any idea of a good resistor value starting point for 2x26650's in series? Thanks for your help.

 

[Benson]

Sorry left out the '2' in d2Flex. Your way sounds easier and better, so be it. Any idea of a good resistor value starting point for 2x26650's in series? Thanks for your help.

There isn't one.

See, the resistor-drive method has the resistors in series with the LED, so they both share the same current -- about 5A is your target. Since the LED has a Vf of around 3.7V at 5A, and the 2s26650 battery delivers about 8V when fully charged, you've got more than 4V drop across the resistor. (Now you've got the info calculate the resistor value using Ohm's law: R = 4V/5A=0.8ohm, but hang on a minute.) So the resistor will dissipate 4V * 5A = 20W of power. This will definitely be a heat problem.

Now delivering current to a low-voltage LED (or series string of LEDs) from a significantly higher voltage source is done sometimes, especially with 5mm LEDs; ir's about 50% efficient, as seen here, but sometimes the difference between 30 hours and 15 hours runtime just isn't worth the cost of a buck circuit. But that's dealing with 10s and 100s of milliwatts in the resistor -- even if you can stand the inefficiency, at 5A, the way that electrical inefficiency is manifested as needless heat in a flashlight already at risk of overheating makes it completely unfeasible.

If you can't get your voltage source below about 5V, series resistors or linear regulators are generally completely impractical for handheld flashlights -- you need a buck converter.

 

[ahx66]

You might want to check this out:

10 A Buck Switching Regulator

I think it's the best thing at the moment for these high current leds.

If you really want to use the 2 26650's in series with a resistor dropping out 4 V you can buy large wattage resistors with a built in heat sink/cooling fins. But I think you would be much happier with the temperature and battery run time with current regulation.

 

[JohnF]

Well then, I'm seeing now why folks are using multiple bucks in parallel to get more current handling, but from what I've read this might be more than I want to tackle at this time. So, I'm back to a couple of options:

1) Use the d2Flex, using 1 26650, or maybe 2 in parallel, direct drive, with or without resistor, depending on if I see angry blue. Since the supply voltage will be much closer to the emitter vf there will be much less loss to the resistor (and less heat) if one is used.

2) Use the d2Flex with HipCC and 2 26650's in series. The HipCC chart shows it should operate at around 84% efficiency in this case, so heat should be manageable. The downside is only 2.8a for the SST-50.

Am I getting closer? I really do appreciate your help in this learning experience.

John F

-------edit--------
just tried a new IMR 26650 fresh off the charger, 4.19v resting. With no resistor, I'm measuring 4.78a and the emitter did not turn blue after a couple minutes of running, so it looks like that might be safe. It looks very bright, BTW.

 

[Techjunkie]

Two cells in series with buck drivers will give you regulated (constant) output level. A single 4.2v cell, or even two in parallel will give you diminishing output as the cell voltage drops. If you're up to soldering a d2flex and a hipCC then you can just as easily solder two buck drivers in parallel. Myself and ma_sha1 have done it with cheap DX drivers with very satisfying results. Check out my build thread here.

 

[Aircraft800]

Well then, I'm seeing now why folks are using multiple bucks in parallel to get more current handling, but from what I've read this might be more than I want to tackle at this time. So, I'm back to a couple of options:

1) Use the d2Flex, using 1 26650, or maybe 2 in parallel, direct drive, with or without resistor, depending on if I see angry blue. Since the supply voltage will be much closer to the emitter vf there will be much less loss to the resistor (and less heat) if one is used.

2) Use the d2Flex with HipCC and 2 26650's in series. The HipCC chart shows it should operate at around 84% efficiency in this case, so heat should be manageable. The downside is only 2.8a for the SST-50.

Am I getting closer? I really do appreciate your help in this learning experience.

John F

-------edit--------
just tried a new IMR 26650 fresh off the charger, 4.19v resting. With no resistor, I'm measuring 4.78a and the emitter did not turn blue after a couple minutes of running, so it looks like that might be safe. It looks very bright, BTW.

All of your thinking here sounds right. I have one SST-50 light regulated with two parallel buckSharks, and I have one light DD with a d2Flex and 26650 IMR cell. The DD setup was much cheaper to build and that cell does a great job of not sagging, so I don't even notice the drop in output over time, but I don't use this light everyday either, and I check the voltage often.

I'll bet if you do build a parallel 26650 pack that you would easily draw more current than the 4.78A you saw. A very small voltage increase makes a big difference in Amps. When mine comes hot off the charger, 4.21V, I can see as many as 5.31A in a very low resistance 1D Mag setup and a d2Flex, so I be careful not to run it that way. Now I set my charger to it's 3.6V setting top-off at 4.1v.

I really like the ideas that Techjunkie has, I've been following his threads ^^^^^^^^^

 

[JohnF]

I'll follow Techjunkie's lead in the future, but for now I'll be quite happy with 2 x 26650's in parallel. I just tested with freshly charged, new-ish, IMR's. With no resistor, it's close to 5a but drops to 3.8a after 5 minutes. It stays there for quite a while, I didn't clock it. So, runtime and brightness will be pretty good with only the d2Flex in the mix. Also, I'll be running in a mode lower than max much of the time, so I'll not worry about the health of the SST-50. My testing is with a lot of extra wire that won't be present in the finished built (including the DMM) so I expect it'll be slightly better than what I'm seeing before it is put together. I was a little concerned with thermal runaway, but no sign of that happening.

I'm just finishing up the mag momentary switch mod, that was challenging enough for a first-timer, but it works beautifully so far.

My 2x26650 parallel pack will be just wired and taped together with a suitable spacer to make up the difference in length between 3xD cells and 2x26650's. I'm thinking about drilling and wiring in a mini-plug for charging without removing the taped-together pack, but we'll see.

With the aluminum reflector just taped to the heat sink, it is a throwing out a lot more light than my new XPG drop in from Malkoff (which I really like, BTW.)

John F

 

[Techjunkie]

If you go the two parallel direct drive route, you'll definitely want the d2flex in there to add some small amount of resistance. I had to stop direct driving my SST-90 from two parallel 26650 cells because after the emitter was broken-in and it's Vf lowered, it would draw way over the specified max current on a full charge. I assume the SST-50 will probably also burn-in over time and it's Vf will also drop.

 

[JohnF]

If you go the two parallel direct drive route, you'll definitely want the d2flex in there to add some small amount of resistance. I had to stop direct driving my SST-90 from two parallel 26650 cells because after the emitter was broken-in and it's Vf lowered, it would draw way over the specified max current on a full charge. I assume the SST-50 will probably also burn-in over time and it's Vf will also drop.

Finished the momentary mag switch mod, seated the britelumens heat sink with SST-50 mounted, wired the d2Flex, loaded with 3 alkalines, and somewhat to my amazement everything works! Don't want to rob my FM hosts for the 2 26650's, but batteryspace shows them in stock, so I ordered a few. With alkalines I'm only drawing 1.26a at the tail, which jives with what it drew on the bench, so I've got no serious losses. When I get the 26650's I'll do the getto parallel mod with a spacer, and see what happens...

d2Flex Question: does it need any heat sinking (like mounting it on the bottom of the britelumes?) I don't want to do a permanent mount as I might be adding a high current driver some day - for now the d2Flex is just sitting loose in the space between the switch body and the heat sink. I don't see anything about heat sinking the d2Flex in the manual like I do for the hipCC (which I'll be using for another project.)

Thanks,
John F

 

[JohnF]

I used my 2 available 26650's to see how this was going to turn out. Wired them parallel, and they fit fine in a Mag2D with a slight tail spring trimming.

The d2Flex is really quite a nice driver, so many options! I haven't played too much with it yet, tonight for sure.

I've tried a cut stock smooth mag reflector, and the britelumens aluminum OP reflector - more throw with the smooth of course, but I'll be keeping the OP in full time for its wider coverage. The beam is beautiful.

The SST-50 from photonfanatic is a great tint, ever so slightly on the warm side, still very 'white'.

I'm measuring 4.7a at the tail with the 26650's not quite fully charged, direct drive with no resistor. That is right in the ballpark of what I wanted, so I'd have to call this a success.

Thanks for all of your help.

John F