Atomic_Chicken
Enlightened
Greetings!
Earlier this week, after returning to work from almost a week of illness, I was excited to get back into the swing of things. In between catching up on Aviatrix mods and Glowfob machining, I decided to design and build a precision light meter and data-logging rig to gather data on the Surefire A2 Aviator runtimes with various batteries. After about 3 hours of design, soldering, machining, fitting, and testing, this was the result:
Presenting... The squawk-o-matic Light Measurement system!
These are thumbnails - Click for full-sized images
The squawk-o-matic is designed around a precision light sensor I.C., the Burr Brown (now a division of Texas Instruments) OPT-101. The hand-soldered sensor circuitry sits inside of the rear part of the tube (the end with the cable coming out to the connector) held in place by a piece of hand-carved polyethylene foam, keeping it rigid and immobile. The other end of the tube (where the flashlight fits) is precision machined from solid aluminum, and is designed to fit the bezel of the Surefire A2 Aviator perfectly (0.002" slip-fit). There is a channel in front of the flashlight bezel where light filters can be placed, in order to cut down the amount of light reaching the sensor (making it less sensitive for incandescent output measurement) or more sensitive (absence of filter, allowing full LED light to reach the sensor for LED runtime tests). This is a specialized instrument, designed SPECIFICALLY for testing the light-output vs. time of the Surefire A2 Aviator, in all of it's various light output modes. Think of it as a "2-range" light meter, one range being adapted for the high output incandescent mode of the A2 (with light-output filter) and the other "range" being adapted for low-level LED output measurement (without the filter).
The output of the sensor goes through a cable as a voltage between 0 and 5V, and is fed into a Mastech 8226T DMM with optically-isolated RS-232 computer interface. The computer logs the data, which I then feed into a spreadsheet to convert into comma-separated text data... which is then fed into a piece of freeware charting software (RMChart - which I HIGHLY recommend...) in order to produce the final chart graphics.
The first test-data gathered by the squawk-o-matic was a series of 4 incandescent runtime tests using different brand batteries (rechargeable and primary). Each test was done with fresh-out-of-the-package (and recent manufacture date) batteries. The rechargeable Li-ion test was done using batteries fresh from the charger, with less than 1 minute delay between removing it from the charger and starting the test. All tests were done using one of my EDC Aviators - modified with an Aviatrix board containing 1 Deep-Red (660nm) and 2 White THC3 LEDs. Since the Aviatrix was in "all LEDs on" mode during the tests, similar results should be obtained using a stock Surefire A2 Aviator with the factory LED ring. NOTE THAT THE TIME AXIS IS DIFFERENT for each of these tests, to really compare battery life and incandescent output duration between these tests, READ THE TIME (X) SCALE MINUTES and don't just look at the graph image when comparing! Each of these tests will now be described:
Test 001 - AW Li-ion (750mah) protected RCR123
This is a thumbnail, click for full-size graph
Battery pair starting voltage = 8.35V
No light-output filter used
This test was done with no filter in place, allowing the full light output to reach the sensor. As a result, the incandescent beam "swamps out" the sensor, clipping the high output and not allowing the full detail of the incandescent brightness changes to be visible. You can see the protection kicking in at the end of the incandescent cycle, and after I reset the batteries (let sit for about a minute, then turned the light back on) you can see the LEDs-only light output followed by the battery protection circuit "kicking in" again at about 43.5 minutes. All-in-all, fairly poor performance when compared to the following Lithium primary (non-rechargeable) CR123 battery results! Basically, when using these cells you get less than 1/2 the incandescent output time, followed by a few dismal minutes of LED output before cell protection kicks in.
Test 002 - Panasonic Brand Primary Lithium CR123
This is a thumbnail, click for full-size graph
Battery pair starting voltage = 6.52V
No light-output filter used
This test was also done with no filter in place, allowing the full light output to reach the sensor. As a result, the incandescent beam "swamps out" the sensor, clipping the high output and not allowing the full detail of the incandescent brightness changes to be visible. These batteries perform surprisingly well... and are readily available from Digi-Key Corporation (Part# P151-ND) through online purchase.
Test 003 - Surefire Brand Primary Lithium CR123
This is a thumbnail, click for full-size graph
Battery pair starting voltage = 6.48V
Test done with light-output filter in place
This test was done with a filter in place, preventing full light output from reach the sensor. As a result, the incandescent beam brightness details are readily visible, and quite interesting! You can see the self-correction of the PWM modulation (not the ACTUAL modulation, as that's MUCH faster) as it slowly rises and falls in a sawtooth-like pattern. An interesting test - notice the end of the incandescent regulation, as the light output swings wildly (between full-output and dim-output incandescent modes) for several minutes before the incandescent light output finally shuts down completely.
Test 004 - Titanium Brand Primary Lithium CR123
This is a thumbnail, click for full-size graph
Battery pair starting voltage = 6.59V
Test done with light-output filter in place
This test was also done with a filter in place, preventing full light output from reach the sensor. As a result, the incandescent beam brightness details are readily visible, and quite interesting! You can see the self-correction of the PWM modulation (not the ACTUAL modulation, as that's MUCH faster) as it slowly rises and falls in a sawtooth-like pattern. In this test, you can see that the Titanium brand CR123 cells do not wildly oscillate as much at the end of the incandescent runtime... they kind of "give up the ghost" faster than the Surefire brand CR123s, dropping rapidly to LED-only output. You can also see that they are not quite as bright as the Surefire's batteries during the life of the incandescent output, HOWEVER... they do last longer in full-output (non-fluctuating) incandescent mode than the Surefire cells.
Test 005 - Panasonic Brand Primary Lithium CR123 (With light output filter)
This is a thumbnail, click for full-size graph
Battery pair starting voltage = 6.51V
Test done with light-output filter in place
This is a re-test of the Panasonic brand CR123's from Test 002 - using batteries purchased from the same batch. The difference is that this test was also done with a light-output filter in place (just like the Surefire and Titanium cell tests 003-004), preventing full light output from reach the sensor. As a result, the incandescent beam brightness details are readily visible, and quite interesting! You can see the self-correction of the PWM modulation (not the ACTUAL modulation, as that's MUCH faster) as it slowly rises and falls in a sawtooth-like pattern... as well as the swinging oscillation between the full-incandescent and dim-incandescent output as the battery life reaches its end. Comparing the results of this test with the results from Test 002, it becomes apparent that the "clipping point" of the sensor when run without a filter is actually at about the 5% of "full output" point... meaning that when there is no light-output filter, 5% light output looks like 100% output to the sensor due to the "swamping out" or "overdriving" of the sensor input because of it's extreme sensitivity. Basically, the light-output filter I'm using to capture the full incandescent mode details is cutting out approx. 95% of the emitted light!
Even with these "more modest" results (actually the SAME results with different vertical "magnification") obtained by re-scaling the vertical-axis light output results through the use of a filter, the Panasonic cells still look like the best of the 3 when it comes to long-duration incandescent output. However, overall incandescent light output is the lowest of the 3 brands of primary cells tested so far... making the Panasonic cells the least desireable choice if brightness is your primary concern.
I hope you've enjoyed these test results, more battery types and output modes (3 LED without incandescent, Aviatrix Single LED mode, Etc.) to follow in the future!
Best wishes,
Bawko
Earlier this week, after returning to work from almost a week of illness, I was excited to get back into the swing of things. In between catching up on Aviatrix mods and Glowfob machining, I decided to design and build a precision light meter and data-logging rig to gather data on the Surefire A2 Aviator runtimes with various batteries. After about 3 hours of design, soldering, machining, fitting, and testing, this was the result:
Presenting... The squawk-o-matic Light Measurement system!
These are thumbnails - Click for full-sized images
The squawk-o-matic is designed around a precision light sensor I.C., the Burr Brown (now a division of Texas Instruments) OPT-101. The hand-soldered sensor circuitry sits inside of the rear part of the tube (the end with the cable coming out to the connector) held in place by a piece of hand-carved polyethylene foam, keeping it rigid and immobile. The other end of the tube (where the flashlight fits) is precision machined from solid aluminum, and is designed to fit the bezel of the Surefire A2 Aviator perfectly (0.002" slip-fit). There is a channel in front of the flashlight bezel where light filters can be placed, in order to cut down the amount of light reaching the sensor (making it less sensitive for incandescent output measurement) or more sensitive (absence of filter, allowing full LED light to reach the sensor for LED runtime tests). This is a specialized instrument, designed SPECIFICALLY for testing the light-output vs. time of the Surefire A2 Aviator, in all of it's various light output modes. Think of it as a "2-range" light meter, one range being adapted for the high output incandescent mode of the A2 (with light-output filter) and the other "range" being adapted for low-level LED output measurement (without the filter).
The output of the sensor goes through a cable as a voltage between 0 and 5V, and is fed into a Mastech 8226T DMM with optically-isolated RS-232 computer interface. The computer logs the data, which I then feed into a spreadsheet to convert into comma-separated text data... which is then fed into a piece of freeware charting software (RMChart - which I HIGHLY recommend...) in order to produce the final chart graphics.
The first test-data gathered by the squawk-o-matic was a series of 4 incandescent runtime tests using different brand batteries (rechargeable and primary). Each test was done with fresh-out-of-the-package (and recent manufacture date) batteries. The rechargeable Li-ion test was done using batteries fresh from the charger, with less than 1 minute delay between removing it from the charger and starting the test. All tests were done using one of my EDC Aviators - modified with an Aviatrix board containing 1 Deep-Red (660nm) and 2 White THC3 LEDs. Since the Aviatrix was in "all LEDs on" mode during the tests, similar results should be obtained using a stock Surefire A2 Aviator with the factory LED ring. NOTE THAT THE TIME AXIS IS DIFFERENT for each of these tests, to really compare battery life and incandescent output duration between these tests, READ THE TIME (X) SCALE MINUTES and don't just look at the graph image when comparing! Each of these tests will now be described:
Test 001 - AW Li-ion (750mah) protected RCR123
This is a thumbnail, click for full-size graph
Battery pair starting voltage = 8.35V
No light-output filter used
This test was done with no filter in place, allowing the full light output to reach the sensor. As a result, the incandescent beam "swamps out" the sensor, clipping the high output and not allowing the full detail of the incandescent brightness changes to be visible. You can see the protection kicking in at the end of the incandescent cycle, and after I reset the batteries (let sit for about a minute, then turned the light back on) you can see the LEDs-only light output followed by the battery protection circuit "kicking in" again at about 43.5 minutes. All-in-all, fairly poor performance when compared to the following Lithium primary (non-rechargeable) CR123 battery results! Basically, when using these cells you get less than 1/2 the incandescent output time, followed by a few dismal minutes of LED output before cell protection kicks in.
Test 002 - Panasonic Brand Primary Lithium CR123
This is a thumbnail, click for full-size graph
Battery pair starting voltage = 6.52V
No light-output filter used
This test was also done with no filter in place, allowing the full light output to reach the sensor. As a result, the incandescent beam "swamps out" the sensor, clipping the high output and not allowing the full detail of the incandescent brightness changes to be visible. These batteries perform surprisingly well... and are readily available from Digi-Key Corporation (Part# P151-ND) through online purchase.
Test 003 - Surefire Brand Primary Lithium CR123
This is a thumbnail, click for full-size graph
Battery pair starting voltage = 6.48V
Test done with light-output filter in place
This test was done with a filter in place, preventing full light output from reach the sensor. As a result, the incandescent beam brightness details are readily visible, and quite interesting! You can see the self-correction of the PWM modulation (not the ACTUAL modulation, as that's MUCH faster) as it slowly rises and falls in a sawtooth-like pattern. An interesting test - notice the end of the incandescent regulation, as the light output swings wildly (between full-output and dim-output incandescent modes) for several minutes before the incandescent light output finally shuts down completely.
Test 004 - Titanium Brand Primary Lithium CR123
This is a thumbnail, click for full-size graph
Battery pair starting voltage = 6.59V
Test done with light-output filter in place
This test was also done with a filter in place, preventing full light output from reach the sensor. As a result, the incandescent beam brightness details are readily visible, and quite interesting! You can see the self-correction of the PWM modulation (not the ACTUAL modulation, as that's MUCH faster) as it slowly rises and falls in a sawtooth-like pattern. In this test, you can see that the Titanium brand CR123 cells do not wildly oscillate as much at the end of the incandescent runtime... they kind of "give up the ghost" faster than the Surefire brand CR123s, dropping rapidly to LED-only output. You can also see that they are not quite as bright as the Surefire's batteries during the life of the incandescent output, HOWEVER... they do last longer in full-output (non-fluctuating) incandescent mode than the Surefire cells.
Test 005 - Panasonic Brand Primary Lithium CR123 (With light output filter)
This is a thumbnail, click for full-size graph
Battery pair starting voltage = 6.51V
Test done with light-output filter in place
This is a re-test of the Panasonic brand CR123's from Test 002 - using batteries purchased from the same batch. The difference is that this test was also done with a light-output filter in place (just like the Surefire and Titanium cell tests 003-004), preventing full light output from reach the sensor. As a result, the incandescent beam brightness details are readily visible, and quite interesting! You can see the self-correction of the PWM modulation (not the ACTUAL modulation, as that's MUCH faster) as it slowly rises and falls in a sawtooth-like pattern... as well as the swinging oscillation between the full-incandescent and dim-incandescent output as the battery life reaches its end. Comparing the results of this test with the results from Test 002, it becomes apparent that the "clipping point" of the sensor when run without a filter is actually at about the 5% of "full output" point... meaning that when there is no light-output filter, 5% light output looks like 100% output to the sensor due to the "swamping out" or "overdriving" of the sensor input because of it's extreme sensitivity. Basically, the light-output filter I'm using to capture the full incandescent mode details is cutting out approx. 95% of the emitted light!
Even with these "more modest" results (actually the SAME results with different vertical "magnification") obtained by re-scaling the vertical-axis light output results through the use of a filter, the Panasonic cells still look like the best of the 3 when it comes to long-duration incandescent output. However, overall incandescent light output is the lowest of the 3 brands of primary cells tested so far... making the Panasonic cells the least desireable choice if brightness is your primary concern.
I hope you've enjoyed these test results, more battery types and output modes (3 LED without incandescent, Aviatrix Single LED mode, Etc.) to follow in the future!
Best wishes,
Bawko
Last edited:
