Re: CMG Infinity Mod.....Lots of Pictures, updated
Hi guys,
I finally got my mod done last weekend successfully but not up to my expectation. This was due to I screwed up the stocked circuit board while I was doing the inductor experiment. But the Luxeon LXHL-BW01 BIN Code Q2K, which I bought from Lambda for $5.00 might be an emitter not meeting his requirement. That means I am having a second quality emitter in my mod.
Here look at some circuit board photos first.
CMG Infinity stock circuit board with LED
Circuit board turned on
Stock circuit board w/o LED
On my post at the same thread on 4/8/04, I guessed that the inductor on the stock circuit board might be too less of power (low saturation?). Even though I got the brightness reached to the level of Ill Pill, the power consumption was very high (1.2 to 1.4 A from 2AA) and the Zetex 310 IC chip was burning hot. So I did some test with different hand-made inductors.
[Some words about the Ill Pill that I assembled. I don't know whether it is the same as the others, mine is very low efficiency, drawing about 1.4A from 2 pretty fresh AA batteries. But Lambda is having a new Ill Pill available soon.]
Part of tested hand winded inductors.
I wanted to control my cost low so I only used the ferrite beads and magnet wires I had on hand to do this test. The picture above showing part of my tested inductors. I winded and re-winded those beads with #30 (red) and #26 (green) with different turns for testing. The best effect would be that one most left at the bottom row. It was the same (except I used #26 wire) as that one used in Douggg's MiniMag 2AA Conversion [
The actual core is a ferrite bead manufactured by JW Miller Magnetics, model number FB73-422. It's about 11mm in length and 5mm in outer diameter. We got ours from Digi-Key, part number M2308, $0.22…..Run the wire through the bead's center hole four times]. It was the brightest one because of the bead size volume and produced least heat that I almost couldn't feel at all. I measured the current draw form the battery was about 0.7A with 2 cm of sense wire. At the time of testing, I repeatedly soldered and de-soldered the hand made inductors; finally I messed up getting the SMD capacitor out from the board. It was so tiny and my soldering iron tip was so huge comparing with it, I couldn't put it back. Finally I decided to remove the diode also to make the circuit a "Maximum Battery Life Solution". At this moment I found that I could not fit this inductor with the board into the bezel. I felt embarrassed and tired, so I simply used one like used in the Ill Pill (Grey Toroid in picture, #30 wire, 40 turns). But the current so as the power both dropped. That was why I was not completely satisfied. The sense resistor was 1.8 cm including both end soldering tip. I tried a shorter one but the current either remained around 0.5A or suddenly jumped up to 1.4A and the IC was burning hot. I guessed the inductor was not good enough. I had no choice, the others were even worse.
I followed what the predecessors did. I composed the emitter and circuit board in a sandwich form. The emitter was soldered onto a copper plated circuit board, so the board functioned double as a heat sink and negative contact while the flash was in turn on status. I also sanded down the NX-05 with sandpaper and final finished with toothpaste. I didn't use NX-05 at the beginning instead I did like what darkzero did by sanding the bezel shining. But the effect of light beam was not good because of lacking a hot spot. So I followed koala's trace, putting the NX-05 onto the head.
Modded bezel, circuit board w/ emitter & sanded down NX-05
Front view of assembled head
Another view of the head
The circuit board came out a little bit from the bezel because the thickness of the inductor. But it wouldn't drop off the bezel because the epoxy made the sandwich a little bit tight to fit.
Assembled mod-Infinity
Mod-Infinity turned on
Below are some photos of beam shot, all taken at 1/60 sec F2.7 except the last one. I positioned two flashes about one foot apart and the camera in the middle of two flashes. In these pictures the bigger circle of light spots of both flashes are the light beams spotted on the wall and the smaller circle of light spots are the reflection of the flashes. The effect of the pictures is so bad but can provide some idea of the brightness through comparison. In order to have a picture closest to my human eye perception, after a few attempts, I get the last picture. (edited: It can only show 2 pics, the others have problem to post. Since the pics effect is bad though, so I only keep the best one.)
L. Mod-Infinity, half CR-V3 3.00V
R. Ultra G, 1 AA 1.60V
6ft from wall, 1/4 sec. F2.8
This picture shows the difference of two light is very close to my perception.
Here are some measuring of the current drawing from the batteries (All voltage are measured with the batteries unloaded):
Flash__________Battery_______Voltage___Current
Mod Infinity___1 AA Alkaline____1.595V___0.49A
_____________2 AA Alkaline____3.20V____0.51A
_____________2 AA Alkaline____2.84V____0.53A__(another sets of battery)
_____________L91 AA Lithium__1.641V___0.50A
_____________1/2 CR-V3 Li.____2.95V____0.52A
Ultra G_______1 AA Alkaline____1.595V___0.24A
_____________L91 AA Lithium__1.657V___0.25A
Ill Pill_________2 AA Alkaline____3.20V____1.65A
_____________2 AA Alkaline____2.84V____1.40A__(another sets of battery)
/ubbthreads/images/graemlins/confused.gif<font color="orange">How can I align those letters without typing lines (I want spaces)</font>
Mod Infinity run-time test, the Ultra G hosted with a fresh AA and only turn on 2 to 3 seconds 4 to 5 times an hour for comparison.
Battery__________Run-time(hr)___Human eye perception
1/2 CR-V3 Li._____0 to 1_________No brightness drop after first few minutes
________________1 to 2 1/2______Started to drop, much brighter than Ultra G
________________2 1/2 to 4 1/2___Un-attended
________________4 1/2 to 5 1/2___Dimmer than Ultra G, still bright
________________5 1/2 to 8 1/2___Dimmer, brighter than Arc AAA "moon-mode"
________________Test stopped
1 AA alkaline_____0 to 1 1/2_______No brightness drop after the first few minutes,
___________________________________brighter than Ultra G
________________1 1/2 to 2 1/2____Started brightness drop, hot spot much brighter
_______________________________than Ultra G
________________2 1/2 to 3 1/2____Overall dimmer but hot spot brighter than Ultra G
________________4______________Hot spot same as Ultra G
________________Stopped test
Run-time test with 2 AA 1.25 V NiCa rechargeable battery (5 year old, never used)
Sandwich______Run-time(min)___Human eye perception
Ill Pill__________0 to 36_______Brightness dropped from the first min all the way down
___________________________to the level of Ultra G
______________36 to 40______Brightness dropped rapidly
______________Stopped test
Mod Infinity____0 to 60_______Comparatively more stable, not watching close.
______________80___________About the level of Ill Pill at 40 min.
Run-time test with 2 AA Alkaline (fresh)
Sandwich______Run-time(min)___Human eye perception
Ill Pill__________0 to 3__________Dropped from the brightest level
______________4 to 60_________Brightness dropped every min, but very bright
______________61 to 110_______Overall and hot spot brighter than Ultra G
______________110 to 120______Same as Ultra G
______________121 to 140______Dimmer than Ultra G
______________Stopped test
Mod Infinity____Not test yet, expect to be brighter and double the time of Ill Pill.
Conclusion:
This mod proved the correctness of my first thought. That is the mod flash would be very flexible in power source. Though we sacrifice some burnt time, we still get 4+ hours of brightness which is above and up to the level of a fresh battery in an Ultra G. When you can get a CR-V3 or 3.6V AA easy and cheap, than use it to get full brightness and longer battery in-shelf life. Otherwise using an ordinary AA, with 4+ hours run time and more brightness, it still can be a light of everyday use.
Things to do:
Mod the Ultra G by replacing the inductor and sense resistor with commercial products to get higher efficiency and higher brightness. Take better photos of beam shots. Try to take pictures in every 15 min while doing run-time test.
