Two stage Firefly board information

[dat2zip]

Just FYI. Some question have come up regarding just what the heck this board is and how it works.

The board has a SMT resistor and a gold plated spring made from a connector contact pin. Just think of it as a spring.

At the bottom of the FF is a 2-56 screw. To install the two stage board requires you to remove the screw and insert the board and then put the screw through the hole in the board and screw it back into the back of the battery tube.

The spring on the board is higher than the screw head and thus the spring makes contact with the battery first as the battery tube is tightened down. The spring is wired in series with the SMT resistor which sets the LOW brightness mode. As you continue to screw the battery tube down compressing the spring the battery eventually makes direct contact with the screw head and is now in regular mode.

The board may need some trimming and flattening as it is only 0.007" thick FR4 board. The spring is set on the high side and it will probably engage several turns before full brightness is encountered. If this is too much travel for the low setting, the spring needs to be compressed or "tweeked" in the final flashlight to suit your needs.

You can use the spare screw or the screw from the flashlight to set the spring height. It should always be higher than the head. The head specifications is 0.042" in height and the spring height needs to be higher than that to ensure it engages before the head does. Squishing too much will warp or flatten the spring and it may need to be replaced if flattened too much.

In the image you can see a bare board, a fully assembly board and the contact pin used to make the spring portion of the assembly.

Wayne

 

[paulr]

Wayne, this thing works by limiting the battery current with a resistor. Electrically speaking is that a really satisfying design, or just a kludge? And will the SMT resistor get hot if you run the low mode for a long period?

Do you think it's feasible to someday make a sandwich-like board with a spring contact on the center connector, so it switches a sense resistor instead of a resistor directly in the battery path? Better still would be a PWM dimmer, of course.

Also, what does the 24R9 marking mean? (I.e. how is the code interpreted?) Thanks.

 

[dat2zip]

The resistor trick is one of the hot mods for many lights right now to get two levels of light out of the same flashlight.

The resistor drops the converter out of regulation and in this mode draws very little current. Somewhere in the 40-80mA range and the resistor value does not dissipate much heat since it is supply low power.

I haven't thought about making a sandwich two stage yet, so, something may yet crop up. You can get a Kroll two stage now from gadgetlover (sp?) if you need it now.

The 24R9 is 24.9 ohms 1% resistor. I believe that is Larrys original proto there that I sent back to him after I duplicated the board layout.

 

[cy]

Thanks Larry, Wayne,

I never realize the batt post was a removable screw. A very simple elegant design. Just bolt it in.

When will this board be available? Paypal ready.

The kroll tail cap two stage mod uses 8 ohm or 10 ohm resister, with most going with 10 ohm. How bright will the 24.5 ohm resister be with firefly?

Thanks,
CY

 

[dat2zip]

cy,

They are on the shoppe as we speak. The two models will be 16 ohm and 33 ohms.

This provides a softer low mode and longer run times.

Anyone handy with SMT resistors could change the resistor to 10 ohms or for that matter any other value if one of two provided doesn't suit your needs.

It's a shot in the dark what values to put on these and on the next batch I'll use any feedback I hear to modify the value I load.

 

[Hoghead]

Wayne,
Which of the two is brighter(on dim)?

 

[tvodrd]

I used a 25 Ohm, 1/4W for the proto- gives me a (whiter) Arc AAA level brightness on low beam. Credit for the concept really belongs to Gadget_Lover, and Doug S found the contact for use in another project. It just turned out to be a fairly easy implementation on my EDC CR2 light. As my orig 123 lights and DSpecks implementation of it (Firefly) use the offset #2-56 screw neg batt contact, I thought the "world" might just like it. (And who better than /ubbthreads/images/graemlins/bowdown.gif Wayne to make and provide it!)

Note that function is greatly improved with a foam rubber "washer" surrounding the assembly in the tail. This acts as a spring to keep the batt's + contact against the module and prevent battery rattle. Also, I "filled" the inside of the contact with silicone RTV to in theory provide additional resistance to side loading during the rotational transition between low and high beam.

Larry

 

[jtice]

Will the new FF 2s come with one of these already installed?

These are a great idea, thanks for your hard work Wayne, Larry, and Doug.

 

[cy]

Thanks wayne, I just ordered a 16 ohm kit. I'll post back how bright the 1st stage is.

Will you be offering the smt resisters?

 

[gadget_lover]

[ QUOTE ]
tvodrd said:
Credit for the concept really belongs to Gadget_Lover, and Doug S found the contact for use in another project.



[/ QUOTE ]

Thanks for the mention. /ubbthreads/images/graemlins/smile.gif It's always nice to see ideas that help others. I saw Wanye's implementation at the SF Bay area get-together last week. It is a VERY nicely done little add on. Well, as far as I could see it was. I did not bring my magnifiers and boy is that thing tiny.

Well Done Wayne! /ubbthreads/images/graemlins/thumbsup.gif


Daniel

 

[Billson]

I'm not familiar with electronics so forgive me if this sounds like a stupid question. What is the difference between the 16 and 33 ohm 2-level switches?

Is one more efficient than the other? Why is the 16 ohm brighter than the 33 ohm?

 

[springnr]

Resistors resist current flow so a higher resistance equals less current. Three volt 123 cell is used in the Firefly so

3volts/16ohms =.187 amps = brighter
3volts/33ohms =.090 amps = dimmer

Depends on sandwich used but around .650 amps is normal full brightness power.

 

[Billson]

Ok. Thanks for explaining.

 

[Doug S]

[ QUOTE ]
springnr said:
Resistors resist current flow so a higher resistance equals less current. Three volt 123 cell is used in the Firefly so

3volts/16ohms =.187 amps = brighter
3volts/33ohms =.090 amps = dimmer

Depends on sandwich used but around .650 amps is normal full brightness power.

[/ QUOTE ]

Well at least the arithmitic is correct /ubbthreads/images/graemlins/grin.gif
This isn't the problem to be solved though. From the Duracell 123 datasheet a CR123 cell puts out a fairly flat 2.8V at loads of around 10-100mA. When driven by a current limited source where the available current is a few 100 mA or less the LT1618 IC will pull whatever current is necessary through the resistor to pull the voltage at the IC input down to 1.5V. This leaves (2.8-1.5)=1.3V across the resistor. Current drawn from the battery is thus:

1.3V/16 Ohms = 81mA or
1.3V/33 Ohms = 39mA

So what does the LED get? My testing shows that under these operating conditions the BB board is about 70% (0.7) efficiency. The BB is receiving :

(1.3V)(.081A) = 0.106W or
(1.3V)(0.039A = 0.051W

At these low drive levels the LED Vf will be down around 3.0V so the currents seen by the LED are:

(0.106W/3.0V)(0.7) = 25 mA or
(0.051W/3.0V)(0.7) = 12 mA

FYI, the LT1618 datasheet shows a minimum voltage of 1.6V but under these condition my testing of a limited number of samples showed that they all pulled the IC input voltage down to 1.5V.

Yes E=IR is the Law but it must be applied judiciously /ubbthreads/images/graemlins/tongue.gif

 

[dat2zip]

Actually, the first equation is almost correct.

In the 16 ohm case it closer to:

3volts/16ohms =.187 amps - 0.020 = .167 amps.

the IC when brought below 1.6 goes into moon mode and is at this point is running but not doing much. So, the typical idle current is around 0.020.

The rest of the current is going to the LED.

Since there is a diode in series, the current is actually:

((3V - 0.4) / 16) -.020 = .145A (An approximation)

*Edited*
That's assuming the Vf of the LED is below Vin. In this case, it is not. Thus, the iC is in moon mode and how it runs and behaves below 1.6V is not a know formula. I can tell you that the 20mA is constant. Assume the rest is going to the LED. Someone would have to plot Vin vs Iout for the voltages below 1.6V. Best is to set this up with a 1/4W axial resistor and a battery. Since it is not drawing a lot of power, the test should be easy to setup and measure the voltage drop across the resistor to get a more accurate reading of the LED current.

Practically speaking tho. Underdriving a 1W or 3W down around the 100mA level is still very bright as seen by the NLS and other underdriven lights. The resistor selection of 16 and 33 may not be a linear 1/2 or 2X the other since the equation is not necessarily linear. It could be exponential for example in which case the 33 could be more than 2X dimmer.

The 33 ohm I'm usinig now is about the same brightness as one of the MM lites with a 1.5V single battery and a P ranked Luxeon.

The NLS in this case is brighter than the 33 ohm resistor if that helps any. I believe my NLS has a R2H in it.

 

[PeterB]

Would it be possible to use an inductor to bring the converter board into moon mode (without having the resistive loss)? Depends probably strongly on the input capacitor.

 

[dat2zip]

PeterB,

Yes, but since the converter is still running the efficiency equation won't change much. The converter in moon mode is very efficient if it didn't draw the 20mA idle current.

The two stage also works on the MM module as well. I would think that the MM would be more efficient in this case. The brightness levels I'm sure will be different between the BB and MM modules.

 

[springnr]

I'm not familiar with electronics.... to circuit analysis. /ubbthreads/images/graemlins/smile.gif gotta love CPF

 

[coyote]

i'm an electronic idiot so any help in answering this would be appriecated:

will these adapters work with both Madmax and Badboys?

what are the run-time multiplication factors for type of each adapter?
(ie: say the 16 ohm version increases runtime by approximately "???" times: if my MMWOLux3 normally ran for 1 hour, i could multiply that by the "???" factor and figure out the extended low-beam run-time)

thanx

 

[Doug S]

[ QUOTE ]
coyote said:
i'm an electronic idiot so any help in answering this would be appriecated:

will these adapters work with both Madmax and Badboys?

what are the run-time multiplication factors for type of each adapter?
(ie: say the 16 ohm version increases runtime by approximately "???" times: if my MMWOLux3 normally ran for 1 hour, i could multiply that by the "???" factor and figure out the extended low-beam run-time)

thanx

[/ QUOTE ]
For the BB you can calculate the *approximate* mA draw per my post above and divide this into the mA capacity of your battery. For example if draw is 81 mA with the 16 ohm resistor, a 1300 mAhr capacity 123 cell will give 1300/81 = 16 hours runtime.
The adapter will probably work with a Madmax but it will very *very* inefficient [<10%] since the Madmax terminal voltage will be pulled down to a very low value, perhaps as low a 0.2V.

EDIT: I have subsequently tested a MM using a 22.1 ohm resistor and an input voltage of 2.8V. The MM IC does not pull the voltage at its input terminal down nearly as far as I speculate above. With the 22.1 ohms and Vin of 2.8V the measured I/O efficiency was 30.3%.