Trying to quantify mods to a Q3

OK,

I'm just posting this because I've been a lurker for a couple years, and only recently a poster. I was reading a bit about the easily moddable Q3 and thought I'd try it out. Most of the easy mods consisted of changing the lens to a UCL and changing to an IMS reflector for more throw. Nobody seemed to quantify this (or I simply missed the threads). Curiosity went beyond my ability to just hope someone did... I went to eBay and bought an elcheapo CHICOM Lux Meter. I got this after I made the first mods, so there's no starting reading, so ASSUMING its roughly equivalent to the one that FlashlightReviews has, that means that a completely unmodded Q3 has about 600 lux at the center of the beam at 1m. That's the theoretical start.

I got two Q3s. I also got two IMS 20mm reflectors and two 22.6mm UCLs from Flashlight Lens. I dropped these into the Q3s first thing (easy). I also made the switch sit deeper with a #21 o-ring from Home depot (doesn't affect output so it will not be mentioned here in).

I then modded the circuit with an output cap and diode. I was unimpressed initially because, my naked eye can't tell the difference. I then went on to change only one of these (one had an SYOL the other had an SYOJ Lux III, I changed the SYOJ). The first thing I changed was the sense resistor. It had an R020 resistor, I put a .033 and .047 ohm resistor in parallel. Then I got a TV1J from the SandwichShoppe (TWOJs are on order from other threads), I replaced the SYOJ star with the TV1J.

This is when the lux meter arrived in the mail.

So, lux readings at the center of the beam are as follows for the following changes:

unmodded Q3 (FlashlightReviews): <font color="blue">600</font>lux/1m
ETA:unmodded Q3, SYOJ (HaveBlue later in thread): <font color="blue">680</font>lux/1m
ETA: Q3, factory circuit, with SYOL, IMS, 22.6mm UCL, R020 factory sense: <font color="blue">750</font>lux/1m
ETA: Q3, factory circuit, with SYOL, IMS, 22.6mm UCL, R020 in parallel with .033ohm: <font color="blue">800</font>lux/1m
Output diode/caps, SYOL, IMS, 22.6mmUCL, R020 factory sense resistor: <font color="blue">880</font>lux/1m on fresh battery, about <font color="blue">700</font> on a used one.
Output diode/caps, SYOL, IMS, 22.6mmUCL, R020 in parallel with a .033ohm resistor: <font color="blue">1050</font>lux/1m on a fresh battery.
Output diode/caps, TV1J, IMS, 22.6mmUCL, R020 in parallel with a .033 and .047ohm resistor: <font color="blue">1300</font>lux/1m on a fresh battery.
ETA: Output diode/caps, TWOJ, IMS, 22.6mmUCL, R020 in parallel with a .033 and .047ohm resistor: <font color="blue">1300-1350</font>lux/1m on a fresh battery.

ETA: NexGen 2x.12ohm board, TWOJ, IMS, 22.6mmUCL: <font color="blue">about 1700</font>lux/1m on a fresh battery.

ETA: With the Zetex ZHSC310 based Q3:
factory (Z310 circuit), SX1K Lux3, factory reflector/lens: <font color="blue">370-400</font> lux/1m
Z310 Q3, removed capacitors, TV1J Lux3, factory reflector/lens, factory R020 sense resistor: <font color="blue">430</font> lux/1m
Z310 Q3, removed capacitors, TV1J Lux3, factory reflector/lens, sense resistor R020 with .033ohm in parallel: <font color="blue">530</font> lux/1m
Z310 Q3, removed capacitors, TV1J Lux3, IMS reflector, 22.6mm UCL, sense resistor R020 with .033ohm in parallel: <font color="blue">900</font> lux/1m

The TW0J/NexGen modded TV1J modded one is the one I carry around right now as EDC (its my first EDC, I usually made due with my ASP navigator and a SF in my bag). The TV1J one will not run on some batteries that the SYOL one will. Those batteries are mostly run down, they work in my SF KL3 body, but they _BLINK_ in the modded Q3 with TV1J. They will run in the SYOL one.

Also, this thing produces a lot of heat. Not BBQ-your-hand-heat, but the unmodded Q3 never really got warm for me. The TV1J one I've only ever felt like running upto 25 min, so I have no true runtime data on it. I know that the original Q3 ran for quite a while before I changed it. Changing from the original circuit to the output caps/diode did change the heat output of the light. Changing the sense resistor merely increased the heat. I do not have an unmodded Q3 to try changing the sense resistor only. Someone else can try that an post some measured results.

With the "blinking" of the TV1J (it blinked when it was only a SYOJ with modded sense resistor), I don't think I can play with using this to run down the batteries I pull from my SF Z3/P91 HOLA light.

I just wish I could find a (cheap) drop in momentary on switch now...

How about it's mod runtime. I see the output twice from factory output.

I don't know. I'm guessing that it has a bit over an hour of runtime. I've only run it for 25min (and it got quite warm, the TV1J one) straight. I'm _guessing_ that it has about 940mA. If the circuit is only 70% efficient, that means that its drawing, what, 1350mA from the batt. That's about what 55min on a 1250mA battery? (ETA: That also assumes the battery can be 100% drained... that seems like a lot of assumptions. /ubbthreads/images/graemlins/frown.gif )

I really don't know how long it works, but that 25min was on a battery that measured 2.85V (it was used from my SF) before it started blinking. It (the battery) still runs in the modded SYOL Q3 and in the unmodded SF KL3/Z3 and now measures 2.63V.

If I'm wrong about the current draw and the effiency... I don't know about full time. I'm more concerned about heat over the long term of the run if I run it more than using a battery. Maybe I can put it in the freezer and run it.

If you mod to give much current to LuxIII, possible to make 2 satages switch for give it to normal brightness as 1 step for cool temp operated and long runtime. One step for high as spacial use as very bright need works.

I probably will do this, eventually. I like the idea of a second mode for an EDC, but I don't like the solution in the form of a switch that has two modes. If I have to make sudden use of this modded Q3 as a shooting light outside (say at the in-laws farm if something attacks the chickens), I don't want a low mode in the cycle.

I have a Kroll, but it doesn't fit (diameter or length), but it has the momentary on I'd like. But the Kroll can be modded to a two stage switch (rotation), which is also good. maybe some sort of sleeve and I can look for some way to reduce the Kroll's length... I don't know... Nothing is ever as easy as it appears.

Maybe if the original switch could be changed somehow (if you open it up, its like a little pen cap system. A three pointed platform connects the ground to a pin, but it only touches when the switch is unpressed (it touches from the bottom). That's why there's a momentary off when on with the switch. Maybe if the contact could be changed to a spring ala the Kroll, it could then just press down on it for "on." Then one could follow the other instructions for changing the cap to a two stage mod.)

WOW! 1300lux with no L-ion rechargable? That's nice. I'm getting a little over 1300 using the Q3 with SYOL, 20mm reflector and rechargable R123 battery. I get about 45min of runtime on high and about 2 hours on low (if I remember correctly) Check out the "Q3 Rechargable" thread for all the details

Are you sure that you got 1300 lux on the 1X123? Did you take a current reading to the luxeon? I'm only asking because a stock Q3 runs at about 500mA. The most that we can get out of a single 123, with some pretty efficient converters, is in the neighborhood for about 700mA. Your lux reading would indicate that you are driving the luxeon at a current level in excess of 900mA. Either there is a mistake some place, or you've stumbled upon a way to get more than 700mA or so out of a single 123. Did you take a current draw reading?

I'm very interested.

No, I don't have a current draw reading, just the one from when I shined the light at the lux meter sensor from 1m away.

To get the current draw what do I need? I've only got a small radio shack DMM, a soldering pencil, and some spare surface mount resistors. I think one might be .1ohm. Is this where I solder one onto a pad of the Lux and solder the lead to that and take a voltage reading while on across the resistor? How do I keep enough pressure on the assembly to get a reading (The spring pressure from below the battery seems to press the pill out so it doesn't turn on unless you press back down).

I can try to take a reading, if it's explained in simple terms...

Maybe I missed it, but did you measure the Lux on your lightmeter before modding? Might just be your particular lightmeter is slightly optimistic.

[ QUOTE ]
cratz2 said:
Maybe I missed it, but did you measure the Lux on your lightmeter before modding? Might just be your particular lightmeter is slightly optimistic.

[/ QUOTE ]

No. I got the lux meter a couple weeks after I'd already altered the circuit with the output caps and diode and put in the UCL and IMS reflector. I did that to both of the two Q3s. I altered the sense resistor value on one (left the other alone) and the lux meter arrived. Thus, I only have FlashlightReview's lux readings of an unmodded Q3, my slightly altered Q3, the readings while it had its sense resistor changed, and the one that I had altered with new resistor and new T-bin Lux III (the one causing the questions).

It is possible that the lux meter is off, but to be off by what, 300 lux? I'd really hate to think it's about 20% off. That would also suggest that the modified for throw and output caps/diode reading of 880ish is off and it really should have been nearer 600. Fine, but that means that the IMS and UCL had no change in the output of the Q3, also that the Output caps/diode did nothing either. All of which goes against what people think (otherwise the circuit design of "max brightness" for the Zetex, and the replacing of reflectors and lenses is for naught)...

I'm more inclined to think that I've got an abberant circuit that is massively inefficient and is unhealthy to the Lux, but I'll try to find a way to get a current output if someone can make sure I've got the idea right and the meager equipment to do it.

To get a draw reading, just remove the tailcap and complete the connection with the leads of the DMM. If we can get a draw reading, we can get an I idea of what the emitter is getting.

I tests using the Mc20, a lux reading of about 1100 lux is typical when running the luxeon at about 500mA. At 1A, you'll see about a 1300 lux reading, IIRC.

In any case, even if your meter is a bit overly optismistic, it would have been helpful to have had a before reading. At least that way we could get an idea of the relative improvement.

I've clowned around with the Q3 circuit a bit and found that adding or replacing resistors doesn't get you much added current to the emitter. The driver appears to be a very simple voltage booster and the more you put in, the more you get out. That's why guys have been getting the 1300 lux readings using a li-ion cell.

If you can take a current reading, let us know.

My QIII is drawing 1.27A from a R123. What is the emitter getting? Don't have a lux meter, but this thing is very bright!

OK, I took some quick readings. Of course your actual readings may/will differ from mine. I am using a JSB R123. When running my Q3 converter through a Lux III with a K Vf luxeon, I'm getting around a 900mA current draw and just over 800mA to the emitter. When I wired the converter up to the bench power supply and set it to 3.7v, I was getting a 900mA draw and 900mA to the emitter.

It appears that at voltages of 3.5v or so and higher, the converter goes into direct drive. The reason I didn't get direct drive with the battery is probably due to voltage drop of the battery under load.

AW,

You probably have a J Vf luxeon to be getting a current reading like that. The light is probably running direct drive, so your current draw reading should be close to what the emitter is getting. If your reading was correct, I think it would be safe to assume that you are getting in excess of 1A to the emitter on the li-ion cell.

Mine is modded with a TWOK /ubbthreads/images/graemlins/thinking.gif /ubbthreads/images/graemlins/thinking.gif

It is also dependent upon your battery's ability to supply current. The JSB123 is kind of on the low side as far as capacity goes and a battery's capacity has a direct relationship to its ability to supply current.

Plus, it's always possible that your luxeon falls in the lower range of the K Vf, while mine might be at the higher portion of the range.

Understood. Thanks for the input!

OK, GN/BN?

A friend has a Fluke 73-III with ammeter (my RadioShack one doesn't). I did the ammeter thing in place of the cap. It drew 1.2A. It did this four or five times. Then I put the cap back and thought, maybe I can get an output amps reading by putting this ammeter into the circuit (recall, I don't have a bench current/source, or clips, etc. I can only take readings if I press stuff together within the flaslight body).

I desolder one lead to the Lux and start pressing down on the lux to press the board into contact with the battery. I touch the loose lead... nothing. I try again. Nothing. I pull the circuit and solder the lead back... nothing.

I drop the board/LED from the other one in. It works. Somehow, taking the Amp reading, I managed to burn out either the Zetex ZXSC300 or its 617 transistor. I pull both sense resistors to put it back to the factory R020. It brightly comes on like a flash, then dims to a very light blue (you can see the LED die glow and die).

Not to be out done, I did the same with the less modded Q3 (only one change to the sense resistor). It read 1.01A on the Fluke. I put in the TV1J from the dead Q3. It reads 1250lux at 1m (actually, it fluctuated upto 1300, but back down again to 1280, I'll deduct a couple lux and call it 1250 for now).

Good news? /ubbthreads/images/graemlins/smile.gif

I couldn't let the thing lie so I wanted to check out what would happen to the dead Q3 if I removed the output diode/caps rendering the Q3 back to an unmodded circuit (still with the IMS and UCL). Hoping, but not expecting it to turn on... It turned on. Hooray, the ZXSC300 and FMTT617 are fine, one of the caps or diode got fried (I have a couple replacements buried on my workbench at home).

Anyway, so people have a reference (I'll add this back to the first post), a Q3 with UCL and IMS only seems to give 750 lux/1m on my lux meter with the SYOL LuxIII with a current draw of .65A (Fluke 73-III DMM replacing the tailcap switch).

So, how do I properly take an output reading of amps?

Man, I was reading down the thread thinking, man, this guy just killed his QIII and I'm partially responsible (though the real blame would go on Chop /ubbthreads/images/graemlins/wink.gif ).

And I was definately not suggesting your meter was 20% off... QuickBeams could be off n one direction, yours could be off in the other direction. He could have had a less choice QIII, you could have the best one ever. There are so many variables. That's why I was asking if you were able to take readings of the stock QIII.

Still, just judging by the heat, it sounds like it's very bright for a single cell light!

Oh look, if you need a converter for the Q3, let me know. Mine's running a NexGen667 now. I'm just using the converter for testing.

Anyway, I was just tinkering with the converter some more, and I'm getting around a 700mA draw from the stock converter with 500mA going to the luxeon. If you are getting a 1.27A draw, you've got to be getting something around 800mA to an amp to the luxeon. If you are, this is unheard of. An amp to the luxeon on a 123 cell?

I would like more details on exactly what it is that you did. If you wind up not needing my converter, I want to do some testing myself to see if I can replicate your results. Even the NexGen or badboy can do over 750mA or so to the luxeon on a 123.

This is sounding too good to be true, but I'm hoping that it is true.

Can any of you math wizzes out there crunch these numbers and tell us something? I'm an electronics idiot.