Mountain Bike Reviews Forum banner
1 - 9 of 9 Posts

·
fight the system
Joined
·
411 Posts
Discussion Starter · #1 ·
Greetings all.

I've been doing some power lurking the last few days on the forum and have some serious spreadsheet development trying to find the right balance of efficiency and simplicity for a DIY build(s) in the near future. On paper I think I'm starting to iron out a design but I think I might have a problem and am hoping some of you people can help me.

I want to build a 6x XP-G R5 wired 3 in series x2 parallel with a single 1000mA driver at 14.8V. Theoretically i'd end up with 500mA at each LED to yeild 1200L before losses.
I gain:
better efficiency with the Cree's
simplicity of one driver
grater range of settings between high and low if I use a UIB2Flex driver
lower voltage requirements

the potential problem I have been alerted to is one string of LED's will have a slightly different V drop than the other that will cause the current to become unbalanced and I'll end up with battery problems and more.

Can parallel light curcuits be pulled off? Is there something else i can add to the circuit to keep the current equal without using 2 drivers?
 

·
Registered
Joined
·
3,062 Posts
jakomonster said:
Greetings all.

I want to build a 6x XP-G R5 wired 3 in series x2 parallel with a single 1000mA driver at 14.8V. Theoretically i'd end up with 500mA at each LED to yeild 1200L before losses.
I gain:
better efficiency with the Cree's
simplicity of one driver
grater range of settings between high and low if I use a UIB2Flex driver
lower voltage requirements

the potential problem I have been alerted to is one string of LED's will have a slightly different V drop than the other that will cause the current to become unbalanced and I'll end up with battery problems and more.

Can parallel light curcuits be pulled off? Is there something else i can add to the circuit to keep the current equal without using 2 drivers?
Maxflex covers all your issues above .

Why limit your light to 1200 lumens when you have a potential 2100 available .
just use a Maxflex and your 14.8 volt pack you will have the same functionality as the bflex without the parallel issues .and be able to drive your XPGs to their full potential

though there have been no issues reported with parallel strings and the dreaded thermal runaway and quite a few build have used this method .
 

·
fight the system
Joined
·
411 Posts
Discussion Starter · #3 ·
troutie-mtb said:
Why limit your light to 1200 lumens when you have a potential 2100 available.
Efficiency. I hate big batteries so anything I can do to keep things lean is worth it. That also means limiting the number of lumens and running my LED's in a more efficient range. Wanting more lumens is a slippery slope that I know you are familliar with but I'm happy to just reach 1000. However after looking into the MaxFlex you suggested it gives me more options since I was limiting my design by not going beyond 3 LED's wired in series because of the voltage requirements.

Thanks for your help.
 

·
Registered
Joined
·
1,136 Posts
I think you will be fine splitting the strings jako. the thing to ensure is that the heatsinking for each string of lights is identical. otherwise, you'll end up with a thermal runaway issue.

I'm with you though, and I'm planning a light that's pretty much identical to yours. 3s2p XPG R5 on triple boards, run off a maxflex and a 2s lithium pack.

honestly, 1100-1200 lumens is a lot of light, and I can build a smaller, more efficient light.

mostly, though, the worry is heat for me. I could just as easily run a 6s LED setup on a maxflex, but I'd be losing efficiency in the driver (unless I bought 5s pack, which are bigger, more expensive, and heavier), and in a housing sized for 6 XPGs, there's no way you're going to have a safe amount of heatsinking capacity.

just looking at the measurements on the taskled site, I'm looking at around 10% higher losses in the driver alone, assuming the same battery pack (2s).

I'll just make more lights if I want more light. :thumbsup:
 

·
Registered
Joined
·
2,017 Posts
I wouldn't bother with 6. Three at 1A will get you 1050lm, vs 1200 for the six at 500mA.

Its been done before for a few reasons.
  • Maxflex put out 1200mA, MCE/XPE only rated for 700mA
  • XRE/XPE efficiency difference between 600mA/1.2A ~ 23%
 

·
Registered
Joined
·
190 Posts
I would go with troutie's advice. Wiring all 6 leds in series and running them at, say, 700 mA with maxFlex, will put out incredible amount of light, yet still highly efficient with around 120 lm/W. End you still have your doors wide open to either increase or decrease the current if heat allows/dictates it.
 

·
Registered
Joined
·
596 Posts
Hi Jaco, We want to run our leds in series if possible. This encourges the current through all of the leds to be the same. I would use the Maxflex and run all the leds in series. Letssee here, 6 leds in series is a Vf of 19.2V. Run at 500mA is 9.6W delivered to leds. Maxflex will need a watt or so for a total battery draw of 10.6W. A battery of two 18650 cells (2s) has about 20WH in it. You will get almost 2 hours run time. If you want to run the leds at 1A, then you'll get about 1 hour on the battery. easypeasy
 

·
fight the system
Joined
·
411 Posts
Discussion Starter · #8 ·
MaxFlex - ok, how's this thing work?

Thanks for the feedback everyone.
My revised plan is to avoid parallel circuts and run:

4 XP-G's in series
7.4V input
MaxFlex driver UIB2
750mA / 350mA
=1120L / 580L theoretical
=9.6W / 4.2W at LED's
with est 1W loss in driver i'll be looking at 1.7h / 3.4h burn time with 2Sx 18650's

But I still have some questions about the maxflex.
after looking on their site with there input current calculator i get:
4 LED's, .75A, 3.2Vf, .88eff, 5.5 lowest input V = 2.0A
4 LED's, .35A, 3.0Vf, .88eff, 5.5 lowest input V = .9A

on Batteryspace the 18650 packs discharge at 4.2mA.
MaxFlex input current is 3.0A max but from above I need 2.0A/.9A input.
-I'm assuming the maxflex can handle a current range?
-So can Batteryspace use a 2.0A polyswitch on the batteries?
-will the MaxFlex be in an efficient range at 2.0A input current?
 
1 - 9 of 9 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top