Mountain Bike Reviews Forum banner

1 - 18 of 18 Posts

·
Registered
Joined
·
324 Posts
Discussion Starter · #1 ·
It is autumn in the southern hemisphere, days are getting shorter & daylight savings ends for me next week, so it is time to bring out the bike lights. While I have dynamo lights on my commuter I run battery lights on other bikes, either an old magicshine or a lux RC XP-G triple, both of which use 8.4 V li-ion packs.

I have wanted to build a red rear light to go with it for a while now, but have struggled with how best to do it. An appropriate taskled driver won't fit in the easy2led housing. Suddenly yesterday I was looking at a triple red xp-e when it dawned on me that a poorman's circuit would work - a red xp-e is a low voltage led in the range of 2V - 2.4V, so 2 would drop the voltage by 4V - 4.8V. A poorman's circuit would give a voltage range between 4.8V - 2.6V, so a simple linear regulator at 700 mA should work just fine.

I got up early this morning & soldered it up as per this circuit & it appears to work perfectly
Red poormans circuit.png

416 theoretical lumens so whichever way you look at it, stupidly bright. The diffuser will help, but for now I will drop it to 350 mA. It will be a good candidate for a Dr Jones mobydrv.

Just waiting on some cable glands to arrive & I will post some pics.
 

·
Registered
Joined
·
324 Posts
Discussion Starter · #3 ·
I agree with you about flashing, but at the moment - no. It is the main reason for going to a mobydrv - I strongly dislike the rapid strobe on the cheap drivers & I like the mobydrv's flashing mode
Oscillating "on-and-flashing" strobes - 2 oscillating strobe modes (regularOoOoOoO and double-pulseooOoOoooooOoOooo) with 2 different power levels each in a 1 to 4 ratio (flash is 4x the drive current of steady state).
 

·
Registered
Joined
·
324 Posts
Discussion Starter · #6 ·
Ofroad'bent I'm a big fan of your little light. My copy replaces the aluminium reflector with a deodorant ball, self fusing silicone tape, but same same. If I could have figured out how to run it at 7.4V I would have done so. The minimalist housing is also the inspiration for my next dynamo light.

Do you know of a brass pill that would take a 20 mm led ?

Magnum, you are right about the UI of the lfex & other taskled drivers. The only downside to George's drivers is the price
 

·
Registered
Joined
·
1,497 Posts
FB, I never found a pill to take a 20mm star. Anybody want a red LED on a 20mm star??
One thing the aluminum reflector does is sink a bit of heat.
 

·
Registered
Joined
·
324 Posts
Discussion Starter · #8 ·
Mmmmm - a minor hiccup - the moby drive is not programming as it should - I suspect that the two extra leds are affecting it in some way. Will have to see if bypassing those 2 leds and using a single li-ion enables me to programme it correctly.

Oh & I might have found a housing to suit a 20mm star, but I will post that over on the crazy-bright little rear vent-light thread
 

·
Registered
Joined
·
9 Posts
There is a problem with the way you have the three leds wired. LED1 and LED2 are in series with the input to the driver. The current through 1 and 2 is a function of the current through LED3. If you were to remove LED3 from the circuit, you would notice a substantial dimming of #1 and #2; perhaps they will go completely dark. As the output current of the driver changes, the input current to the driver through #1 and #2 will change. It is also possible that #1 and #2 can limit the input current to a point where #3 will not receive the current you expect it to.
 

·
Registered
Joined
·
3,883 Posts
There is a problem with the way you have the three leds wired. LED1 and LED2 are in series with the input to the driver. The current through 1 and 2 is a function of the current through LED3. If you were to remove LED3 from the circuit, you would notice a substantial dimming of #1 and #2; perhaps they will go completely dark. As the output current of the driver changes, the input current to the driver through #1 and #2 will change. It is also possible that #1 and #2 can limit the input current to a point where #3 will not receive the current you expect it to.
Having the LED1 and LED2 in series on the input side of the driver is done to drop the 8.4V of the battery into the acceptable input range (2.7V-6V) of the driver selected by the OP.
 

·
Registered
Joined
·
9 Posts
Having the LED1 and LED2 in series on the input side of the driver is done to drop the 8.4V of the battery into the acceptable input range (2.7V-6V) of the driver selected by the OP.
I realize why it was done this way but it is a poor way to accomplish it. It makes the current through #1 and #2 dependent on #3. Should you decide to have #3 flash, #1 and #2 will also flash. Also, depending on the driver, if it is a constant current type, as the battery voltage drops with usage, #1 and #2 will get brighter and #3 will stay the same.
 

·
Registered
Joined
·
3,883 Posts
I realize why it was done this way but it is a poor way to accomplish it.
Maybe one of the reasons this circuit is commonly referred to in the LED light world as "Poormans".

Just a method to use a really cheap constant current linear driver with a low input voltage threshold to drive more than one LED. While not electrically elegant or optimum it gets the job done. Not as widely used today as say 4-5 years ago, there have been lots and lots of DIY lights built using this circuit.
 

·
Registered
Joined
·
324 Posts
Discussion Starter · #14 ·
It makes the current through #1 and #2 dependent on #3. Should you decide to have #3 flash, #1 and #2 will also flash.
Both of these statements are correct & exactly what I wanted to achieve
Also, depending on the driver, if it is a constant current type, as the battery voltage drops with usage, #1 and #2 will get brighter and #3 will stay the same.
I don't think this is right, at least not with the linear driver I am using. As far as I can tell current remains constant as battery voltage drops.

As Vancbiker said, it used to be used a lot more widely - it relies upon particular quirk of the combination of voltages which worked for 2 or 3 P7 leds, but not as well for XM-L leds. It happens that the lower Vf of red XP-E leds meant I could use it in this way.
 

·
Registered
Joined
·
3,883 Posts
The current through #3 remains constant, NOT #1 and #2.
I don't agree with this either. This driver a linear regulator. Current draw by the driver from the battery is constant as long as the battery voltage, less the voltage drop of LED1 and LED2, is higher than the Vf of LED3. The excess energy is dissipated as heat by the 7135 chips on the driver.

If this driver were a switching regulator then LED1 and LED2 would see an increase in current as Battery voltage drops.
 

·
Registered
Joined
·
9 Posts
That's the way linear regulators work. Only the output is constant current, not the input. You can't have both input and output be constant current with a battery that changes voltage over time. If you have a driver that is constant current output; 500ma, 750ma, 1000ma, whatever, then the input current has to change as the input voltage drops.
 

·
Registered
Joined
·
324 Posts
Discussion Starter · #18 ·
Sorry mate I think you have a linear driver confused with something else. What you describe is correct for a buck driver.

The way a linear driver works is that it maintains a constant current, both input and output and burns the excess voltage as heat - this is the reason for recommendation that when using a linear driver the battery voltage should be as close as possible to the Vf. Using my example above, @ 0.700 A, the Vf of a red XP-E is 2.33V, hence the 2 on the input side drop the voltage by 4.66V. With a freshly charged 2s2p li-ion pack @ 8.4V, the driver will see (8.4-4.66) = 4.66V, so the driver will have to burn off 1.41V @ 0.7A = 0.987W.

When the pack voltage drops to 7.5V, the regulator will see (7.5-4.66) = 2.84V so will only have to burn off 0.51V @ 0.7A = 0.357W.

At 7V there is no longer sufficient voltage to drive the leds at 0.7A and all three will dim as the voltage continues to drop to the limit of the pack.

Interesting in typing this I have realised that I am wrong about this being a good candidate for a moby drive. One of the groovy features of the moby drive is that it progressively drops the current once the battery drops to 3V. In my case the driver will see that voltage when the pack drops to 7.5V & it is going to progressively limit the current so that by 7V the leds will only be getting 0.2A.

Not to worry, the driver I have will do fine for a solid light & if I want a flash sequence I a taskled lflex will do the job, and can be wired in a conventional manner all 3 leds in series.
 
1 - 18 of 18 Posts
Top