# Brinkmann LED Flashlight Explained, + Circuit diagram

#### MrAl

##### Flashlight Enthusiast
Hello all,
I got interested in the LED flashlight too
and decided to do a little investigation.
I took it apart as per instructions i found
on this site (thanks to all who posted)
although i had to keep pressure on the
barrel in order to keep the plastic
tabs from popping back in while i tweeked
the other 3 tabs.
Anyway, i did a fairly
complete analysis and posted the circuit
diagram complete with parts list and waveforms
Here is the explanation site: http://forum.howstuffworks.com/showthreaded.php?Cat=&Board=elec&Number=46128
reach from that url too) http://members.aol.com/Xaxo/
with the full schematic/parts list/waveforms.

Enjoy, and feed back, results etc would
be appreciated.

--Al

#### KenBar

##### Enlightened
What a great drawing!
I see why my attempt did not work.
Good job!

#### Steelwolf

##### Flashlight Enthusiast
Wow!! I'm still trying to draw out the diagram, but you have not only drawn it out, but also a schematic and a full explaination.

Thanks!

I'm still trying to digest it though

#### PeLu

##### Flashlight Enthusiast
<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by MrAl:
Anyway, i did a fairly
complete analysis and posted the circuit
diagram complete with parts list and waveform
<HR></BLOCKQUOTE>

Very well done Al!
I have a few remarks:
<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>
While L dumps its current through the LED, the voltage across the LED falls only slightly because of the nonlinearity of the LED
<HR></BLOCKQUOTE>

If you look only at the interesting part of the LED's U/I diagram (lets say in between +20mA and +80mA) it is pretty linear. It is just no resistive load. You could look at it like a 3V voltage 'sink' with a 25 Ohm resistor in parallel (values estimated). Linearity does mean that (in the part we look at) the U/I diagram makes an even line, but not necessarly that this line hits the 0/0 point. If you learnt another definition of 'linearity' at school (in it's electrotechnical meaning) please let me know.

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>
In other words, if you were to supply about 40ma continuous uninterrupted
current flow through the LED you would get the same level of light output from the LED
as you do when pulsing it at a higher current which averages 40ma
<HR></BLOCKQUOTE>

First you do not state which kind of average you are referring to )or have I overlooked it?). Is it arithmetic average or RMS or whatever?

And then, the LED's light output over current is not linear. At higher currents the efficiency is lower. So pulsing a white LED with a perfect square wave and a 50% duty cycle will NOT give the same light as half the current. (besides, as you did not state that: Also the LEDs power consumption is higher in pulsed mode)

Don't get me wrong: I'm very happy that you did this more than exemplary work. I would not have made it almost as nice and will probably have many other errors in it.

#### MrAl

##### Flashlight Enthusiast
I got interested in the flashlight and then i found all the info on
this site about different flashlights and tests and all, so i thought i
would contribute to the information pool also by determining how the
flashlight circuit was made and how it worked so that others could
build their own home made flashlights and power them up with circuits
similar to this one and install them in whatever flashlight they wanted
to. Having posted the schematic and brief explanation, i feel i have
acheived that goal.

Some of the points you made were interesting, and i would like to
explain a few things pertaining to those remarks.

First, the definition of 'linear' is:
the multiplication of the voltage by a constant k results
in the multiplication of the current by the same constant k.
The LED is certainly not linear, and doing a nonlinear analysis
anyway, i saw no need to use a piecewise linear approximation.
If you still are not comfortable with calling the LED v/i curve
nonlinear, then simply substitute the words
"v/i curve of the LED"
for the words
"nonlinearity of the LED"

Secondly,
when i say "average" i mean "average". Whenever someone says average, they
mean the arithmetic average and not the geometric average, harmonic average,
square root of the mean of the square, ninth root of the velocity of
the speed of light in a vacuum times the current cubed, or anything else

Third,
when i said the pulsed current produces the same light as about
the same average current through the LED would produce, i meant just that too.
The reason for this is quite simple: when you half the time of the light output
you half the total amount of light being output, and when you double the current
to the LED you approximately double the light output from the LED. When you do
both, you get the same output as if you supplied about the same continuous current
to the LED as the average of the pulsed current. Yes the led is more inefficient
at higher currents, but it will be hard to notice a difference. Also, 'about'
means just that too.
For most practical purposes, when you supply 'about' the same constant current
to the LED as an averaged pulsed current, you get 'about' the same light output.

Lastly, i will post updates when i do more with this circuit.
I did find already that the circuit can operate on a single
AA battery with slight mod's. I built up a similar circuit
using the parts list shown (with the schematic) and wound
results as the original circuit. The thing i really like
much cheaper then buying evaluation boards for parts that will
cost \$4.00 just for the ic. Im thinking of using a MOSFET
output transistor to increase efficiency. Only problem is,
that will raise the price tag a little. With the transistors
in the parts list, this circuit can be built for under
75 cents.

As i do more tests ill post the results here so everyone can
build the circuit and modify it to fit any number of LEDs
and battery cells they wish to use with their flashlights.
I would hope that anyone else working with this circuit would
post their results as well.

Good luck with it,
--Al

PS. Special thanks to Stingmon for the use of the WCT graphic.

#### KenBar

##### Enlightened
<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by MrAl:
when i say "average" i mean "average". Whenever someone says average, they mean the arithmetic average and not the geometric average, harmonic average, square root of the mean of the square, ninth root of the velocity of the speed of light in a vacuum times the current cubed, or anything else <HR></BLOCKQUOTE>

LOL!! This is my kind of humor..

D

#### **DONOTDELETE**

##### Guest
Great work MrAl!

The Brinkman circuit is nearly identical to the "solar powered LED walkway light" mentioned in this thread:
Want a LED flashlight with only 1 AA?

The circuit (minus a few parts which are the photocell switch) is shown here.

The parts have slightly different values and the solar light circuit has one more resistor and capacitor from the battery + side to the base of the PNP.

Thanks for the explanation of how the circuit works. It sounds as if you ran it through a PSpice circuit simulator.

One thing about all these types of circuits is the large peak current through the LED. This peak is usually much higher than the LED manufacturer's "absolute maximum rating" for the LED current. I've exceeded these maximum ratings before with bad results, i.e. destroyed parts, so I have great respect for them.

One modification to the circuit which would smooth out the current pulses in the LED and lower the peak LED current would be to add a Schottky diode between the Q2 colector and the LED anode. Then add a 10uF (2.2uF might work just as well), low ESR, tantalum capacitor from the junction of the cathode of the new Schottky/LED anode, and ground (emmiter of Q2). This would stabilize the current pulses in the LED.

Unfortunately it would also decrease the overall effiency of the circuit because of the drop across the diode, leading to slightly shorter battery life.

You might be able to get this effeciency back by using a Zetex ZXT14N20DX transistor in place of Q2. This part has a saturation voltage of ony 30mV at a colector current of 1A! Wow! A bit costly though. Also using an inductor with a very low DC resistance would also help.

#### jeff1500

##### Enlightened
Thanks. I really appreciate everyone's willingness to share.

I'm most interested in a light circuit with three or four leds and maybe two batteries. I've looked at the Brinkman light in the store a couple of times, but one led doesn't seem like enough for a package of that size and weight.

What would it take to make it work for multiple leds?

#### MrAl

##### Flashlight Enthusiast
Yes thanks for all the inputs in the various threads; it is very interesting to hear about other's experiences with these things.

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by jeff1500:

What would it take to make it work for multiple leds?
<HR></BLOCKQUOTE>

I'm going to try that soon too
Right now i think you can just connect two
in series. Just how many you can connect
in series is probably dependent on the design
of the coil (Brinkmann). The coil will have
to sustain a higher voltage while pumping the
LED's. If its made well enough it could handle several in series.
My guess is that a better coil with lower inductance will handle more LED's in series,
and lower series R will handle a few in parallel too.
I guess it will have to be tried to find out
As usual, ill post any results i get here in one of the threads.

Good luck with your LED circuits,
--Al