Questions on Direct Drive

If I have a Samsung 30Q and an unregulated DD driver, connected to 3x XP-L HI in parallel, will I burn the LEDs? How is the power distributed among the LEDs in series vs in parallel? I’m just really confused, but want to get a Convoy S2+ triple with a FET+7135 driver… :smiley: Which will be my first pocket rocket EDC.

Thanks in advance!

Not by current. Each of the LEDs will take one third of the delivered current.
But it will get very hot very fast at full power.

You should be okay if you have copper DTP boards and a good thermal path. The Samsung is rated for 15A continuous discharge, so it should get pretty hot, but with good heat path, it should be a nice lumen blaster! :cowboy_hat_face:

My light is a triple XP-L HI, Noctigon DTP PCB, FET DD driver, in a Convoy S6, and I have no problems.

LEDs in a series connection won’t work from a single cell without a boost driver.

In series, voltages add. For 3x series connected LEDs, you would need the Vf x 3 to make the LEDs function. In parallel, the voltage is the same across all three LEDs, and the current is evenly divided between the three.

So it would be around 5A for each LED, without counting in the resistance in the flashlight and connections?? I finally understand now the power distribution now. But what is Vf (I’m guessing forward voltage?) Thanks a bunch for clarifying this.

With a 30Q you should expect about 10 amp draw, depending on your wiring and multimeter setup. With my cheap multimeter I never got past 10.3A for a triple XP-L HI, but some people have reported up to 12A.

Yes, Vf is “forward voltage”. If you look at the voltage that will drop across the emitter (a diode, basically) as the current increases, the voltage increases, until it (the voltage) starts flattening out (or it could be the other way around, i.e., as you increase voltage, the amount of current increases until it levels out). I think they specify the “maximum” Vf, which is the amount of voltage that the mfr will “guarantee” the LED, i.e., if you keep increasing V past the Vf (max) and the LED dies, then it’s “on-you”.

Up to 14 Amps with good cell here, measured with clampmeter. But still ok for 3 LEDs.

Thanks for all the details! I think around 3.33A for each LED is good enough for me. Output would be around 3000 lumens, right? :+1: How do you measure current draw with a DMM (or what the correct terminology is)?

The current flowing through each emitter would be dependent/determined by the Vf of each particular LED. There could be small variations between the emitters, but probably nothing significant.

For instance, say the Vf’s are:

3.87
3.80
3.85

The emitter with the Vf of 3.80 is going to pass more current than the other 2. Depending on the variance of Vf, the difference could range from small/minuscule to large enough to cause a significant problem.

Running emitters in parallel was generally frowned upon not too many years ago because of the danger of thermal runaway causing catastrophic failure due to mismatched Vf. What happens is in a scenario like the one I posted above, is if there is a large variation between the Vf’s of said emitters, the one with the lowest Vf will by nature allow more current to flow across it, which in turn translates to higher heat. If the case is extreme the emitter will reach a stage of thermal runaway and burn itself out, the remaining emitters have to “absorb” the “excess” current, causing those emitters to then experience thermal runaway themselves.

I haven’t seen anyone caution against this in the flashlight world in forever, and even though tons of people have emitters ran in parallel, probably the majority of us here, I haven’t heard of any issues or problems. Perhaps LED manufacturers have gotten better with making diodes with less variation of Vf, or perhaps the emitters themselves have become so robust it’s less of an issue?

I know your other question has already been answered, but I was bored while waiting to pick up my kid from school so I thought I’d share lol, sorry.

I’d say search YouTube for “how to measure current with a multimeter”, there are some good vids there, and if you’re like me it’d be easier to understand watching someone while explaining it than understanding someone here trying and explain it over text IMO.

-Michael

I think this is definitely something to keep in mind. The danger depends on the LED type. For XPLs, people run single emitters direct drive and the resulting current does not burn the emitter. So there is really no danger with XPLs in parallel.

With something like the 219C there is the potential for more danger because of its much lower Vf.

Put your meter in current mode, then move the leads so one is plugged into the 10A socket and the other in the + then touch one lead to the battery “-” side and the other lead to the exposed part of the body. That will measure total current flow. If you have a triple, then divide that current by 3.

On that note, something that’s always bothered me is that strangely, Cree doesn’t offer LED’s that are binned by Vf, but Nichia and some other manufacturers do. You would think that it’d be easier or less risky to run those binned Vf LED’s in parallel, but that doesn’t seem to be the case. I’ve wondered many times why that is, does Cree do a better job of maintaining Vf across a product line? Or is there something else at play here? Also curious, is that some of those manufacturers have used Cree chips to manufacture their emitters, like the EZ1000.

IDK, maybe someone with more knowledge could shed some light on it, I don’t have an explanation for why that is, and no longer have any contacts or access to resources to possibly find out.

My partial answer or guess at an answer would be in the nonlinear way in which LED’s pass current. The danger increasing the lower the Vf due to increased “overhead” available from the battery, in a single cell linear or DD light.

-Michael

Thanks for your input! Every little bit of info is helpful to me. As for running LEDs in parallel, I guess it’s not too much of a problem. Haven’t heard any issues with it so far. Hopefully technology has improved by now so that the consistent manufacturing of LED’s are consistent! Ahh, and I totally forgot about Youtube. Thanks for reminding me, although ohaya’s already clarified that.

:+1:

I put 219C’s (single emitters) into a Convoy S2 and Convoy S2+ along with FET driver’s and haven’t had any issues so far. Both draw over 7A using a 30Q on my cheapie DMM with thick leads. I’ve run them until they’re too hot to hold but so far so good :slight_smile:

Good to know! Well, I guess I’ll just go for the “WOW” factor, instead of high CRI. Can’t wait to get my hands on one!

This article on measuring flashlight amps could be helpful: Measurement on Flashlight