djpark
Enlightened
Introducing Arc-AA MOD with 8mm Power LED from ISP
and PIC12F675 Micro Processor controlled Multi-Brightness Control
running on 14500 (AA sized) Li-Ion Rechargeable Battery!!
Goal of this MOD
<ul type="square"> [*]Minimum 5 hours bright current regulated light
[*]Minimum 24 hours of useful light (with dim mode)
[*]Multiple brightness control using PWM with micro processor
[*]Follow Arc-AA design philosophy as EDC
[*]Minimal physical modification to the original head with limited tools
[*]No fancy stuff due to limited space in the light head
[*]As simple operation as possible for practical use [/list]
Result
After having done quite a few mods to Arc-AA and Arc-AAA with various LEDs from ISP, the only possible way for me (as an amateur modder) to provide much brighter light from these lights is to provide higher battery voltage. So I decided to try on using Li-Ion rechargeable battery.
I went through many different design experiments and finally decided to stick to version 3.1 below. The final result is as follow.
<ul type="square"> [*]5.5 hours almost flat current regulated bright output (130mA) with high brightness mode
[*]Over 48 hours useful light (15mA) to half brightness with low brightness mode
[*]Simple high / low alternating brightness control (I did try 3 levels, but found that it is not very practical)
[/list]
R&D Versions and Runtime
Various stages of the R&D design and the run time charts are provided below. The version 4 and 5 are discarded and the current version is 3.1.
I put the battery retainer ring at the bottom of the body rather than stick to the head. This provided smoother turning of the head and also some pressure I needed to keep the led assembly firmly touching the head for better electrical contact.
Version 1 - Direct drive with series resistor
Just simple series register (3.3 ohm) provided about 5.5 hours of gently declining brightness with 14430 Li-Ion cell. This smooth curve is attributed to the Li-Ion cell characteristic.
Version 2 - Current regulation using FET + transistor
I used DIY circuit provided by "UK Owl" ( here ). Since I didn't have SMT 3.3 ohm resistor at that time, I used normal one as sense resistor (left in the picture below). The right hand side one is the direct drive version.
In fact, the run time was shorter than the direct drive using same 14430 cell (4.5 hours) and the regulation doesn't seem to be effective. I beleive that the poor regulation is caused by the combination of the components requiring bigger drop out. When the same unit was used with 4 pieces of NiMh batteries, the result was beatiful (11.5 hours very flat output), so I know the circuit actually works.
Version 3 - Version 2 + PIC dual brightness control
Based on the same current regulator circuit used in version 2 and added PIC12F675 micro-P to control the brightness with PWM. This time, I used all SMT components only in very tight arrangement. I also used an LED with quite low Vf (3.3V). this gave me 130mA current consumption during the regulation.
The result is much better. I got 2 hours flat regulation and another 2 hours gradual declining brightness to 80% followed by 1 hour of quick dimming into the moon mode using 14430 cell. Medium level (30%) duty cycle gave me over 14 hours of reasonably smooth output before falling to the moon mode. With low brightness mode, I stopped after 53 hours with half brightness of the original brightness (10% duty cycle).
After carrying it around a few days, I found that having 3 brightness levels is actually troublesome. So I changed to simple high and low only dual brightness as final version.
Towards the end of this version, I received 14500 Li-Ion rechargeable cells from a manufaturer and decided to try to compare the capacity with 14430 cell I used so far. As in the run time chart below, I got 3 hours flat current regulated output and another 3 hours drop to 80% followed by quick dimming into the moon mode. That's quite impressive.
Version 4 - PIC micro-P current regulation
PIC12F675 has an analog comparator which can output digital signal high or low depending on the voltage comparison and I thought it could be a good way to control the current regulation. Since the output is digital, it will actually cut off the power when the current is too high and immediately switch on again when the voltage drops. It worked quite well, it is a kind of PFM control. But when I combined it with PWM for brightness control, somehow I couldn't get it working correctly with some flickering, possible due to the lack of RC or LC to filter the output.
Besides, needed a zener diode to provide the absolute voltage reference and this took space in the head. <font color="red"> So this version is dropped.</font>
Version 5 - LDO regulator + PIC 3 stage brightness
I wanted to try using LDO regulator to provide 3.3V output from the battery voltage 4-3.5V and using an LED with Vf 3.3V or below. On the bread board it worked quite acceptable. With reduced components count, I could also put a schottkey diode and 47uF capacitor to hold some charges during momentary power off to select the brightness level. Everything seemed fine till I found that the current output from the regulator I used is not big enough to drive the LED over the specification, I got about 80mA only. <font color="red"> So this design is also dropped.</font>
<font color="blue"> Version 3.1 - Version 3 + 8mm Power LED </font>
Around this time, I received some newer Power LEDs from ISP, Korea. These 8mm LEDs are using a single die and rated at 150mA current. The output is very white and has quite low Vf (3.2-3.4V). So I took one piece with Vf about 3.2V and put to the same head I made for version 3.
The current consumption at high mode is still 130mA, this means that the current regulation works fine. If I change the sense resistor to 2.2ohm instead of 3.3ohm, I would get about 200mA.
With the lower LED Vf, I could get much longer run time under regulation -- full 5.5 hours flat output under regulation till the battery voltage dropped to 3.55V. Also the XY plot confirms that the current regulation is done till the battery voltage 3.55V.
The small increase of brightness as time progress is caused by the ambient light, I started measuring at 2AM and finished at 10AM.
Battery
I am using these batteries. (700mAH)
Charger
And I use this home made battery charger.
Beam shots
Finally some beam shots. The left one is a standard (unmodded) Arc-AA with a new battery and the right is version 3.1 with fully charged cell on high brightness mode.
Wall at 50cm distance (1 sec F2.8)
Wall at 1m distance (1 sec F2.8)
Wall at 2m distance (1 sec F2.8)
Bed at 2m distance (4sec F2.8)
Aircon at 3m distance (4sec F2.8)
Thank you for viewing.
-- dj
and PIC12F675 Micro Processor controlled Multi-Brightness Control
running on 14500 (AA sized) Li-Ion Rechargeable Battery!!
Goal of this MOD
<ul type="square"> [*]Minimum 5 hours bright current regulated light
[*]Minimum 24 hours of useful light (with dim mode)
[*]Multiple brightness control using PWM with micro processor
[*]Follow Arc-AA design philosophy as EDC
[*]Minimal physical modification to the original head with limited tools
[*]No fancy stuff due to limited space in the light head
[*]As simple operation as possible for practical use [/list]
Result
After having done quite a few mods to Arc-AA and Arc-AAA with various LEDs from ISP, the only possible way for me (as an amateur modder) to provide much brighter light from these lights is to provide higher battery voltage. So I decided to try on using Li-Ion rechargeable battery.
I went through many different design experiments and finally decided to stick to version 3.1 below. The final result is as follow.
<ul type="square"> [*]5.5 hours almost flat current regulated bright output (130mA) with high brightness mode
[*]Over 48 hours useful light (15mA) to half brightness with low brightness mode
[*]Simple high / low alternating brightness control (I did try 3 levels, but found that it is not very practical)
[/list]
R&D Versions and Runtime
Various stages of the R&D design and the run time charts are provided below. The version 4 and 5 are discarded and the current version is 3.1.
I put the battery retainer ring at the bottom of the body rather than stick to the head. This provided smoother turning of the head and also some pressure I needed to keep the led assembly firmly touching the head for better electrical contact.
Version 1 - Direct drive with series resistor
Just simple series register (3.3 ohm) provided about 5.5 hours of gently declining brightness with 14430 Li-Ion cell. This smooth curve is attributed to the Li-Ion cell characteristic.
Version 2 - Current regulation using FET + transistor
I used DIY circuit provided by "UK Owl" ( here ). Since I didn't have SMT 3.3 ohm resistor at that time, I used normal one as sense resistor (left in the picture below). The right hand side one is the direct drive version.
In fact, the run time was shorter than the direct drive using same 14430 cell (4.5 hours) and the regulation doesn't seem to be effective. I beleive that the poor regulation is caused by the combination of the components requiring bigger drop out. When the same unit was used with 4 pieces of NiMh batteries, the result was beatiful (11.5 hours very flat output), so I know the circuit actually works.
Version 3 - Version 2 + PIC dual brightness control
Based on the same current regulator circuit used in version 2 and added PIC12F675 micro-P to control the brightness with PWM. This time, I used all SMT components only in very tight arrangement. I also used an LED with quite low Vf (3.3V). this gave me 130mA current consumption during the regulation.
The result is much better. I got 2 hours flat regulation and another 2 hours gradual declining brightness to 80% followed by 1 hour of quick dimming into the moon mode using 14430 cell. Medium level (30%) duty cycle gave me over 14 hours of reasonably smooth output before falling to the moon mode. With low brightness mode, I stopped after 53 hours with half brightness of the original brightness (10% duty cycle).
After carrying it around a few days, I found that having 3 brightness levels is actually troublesome. So I changed to simple high and low only dual brightness as final version.
Towards the end of this version, I received 14500 Li-Ion rechargeable cells from a manufaturer and decided to try to compare the capacity with 14430 cell I used so far. As in the run time chart below, I got 3 hours flat current regulated output and another 3 hours drop to 80% followed by quick dimming into the moon mode. That's quite impressive.
Version 4 - PIC micro-P current regulation
PIC12F675 has an analog comparator which can output digital signal high or low depending on the voltage comparison and I thought it could be a good way to control the current regulation. Since the output is digital, it will actually cut off the power when the current is too high and immediately switch on again when the voltage drops. It worked quite well, it is a kind of PFM control. But when I combined it with PWM for brightness control, somehow I couldn't get it working correctly with some flickering, possible due to the lack of RC or LC to filter the output.
Besides, needed a zener diode to provide the absolute voltage reference and this took space in the head. <font color="red"> So this version is dropped.</font>
Version 5 - LDO regulator + PIC 3 stage brightness
I wanted to try using LDO regulator to provide 3.3V output from the battery voltage 4-3.5V and using an LED with Vf 3.3V or below. On the bread board it worked quite acceptable. With reduced components count, I could also put a schottkey diode and 47uF capacitor to hold some charges during momentary power off to select the brightness level. Everything seemed fine till I found that the current output from the regulator I used is not big enough to drive the LED over the specification, I got about 80mA only. <font color="red"> So this design is also dropped.</font>
<font color="blue"> Version 3.1 - Version 3 + 8mm Power LED </font>
Around this time, I received some newer Power LEDs from ISP, Korea. These 8mm LEDs are using a single die and rated at 150mA current. The output is very white and has quite low Vf (3.2-3.4V). So I took one piece with Vf about 3.2V and put to the same head I made for version 3.
The current consumption at high mode is still 130mA, this means that the current regulation works fine. If I change the sense resistor to 2.2ohm instead of 3.3ohm, I would get about 200mA.
With the lower LED Vf, I could get much longer run time under regulation -- full 5.5 hours flat output under regulation till the battery voltage dropped to 3.55V. Also the XY plot confirms that the current regulation is done till the battery voltage 3.55V.
The small increase of brightness as time progress is caused by the ambient light, I started measuring at 2AM and finished at 10AM.
Battery
I am using these batteries. (700mAH)
Charger
And I use this home made battery charger.
Beam shots
Finally some beam shots. The left one is a standard (unmodded) Arc-AA with a new battery and the right is version 3.1 with fully charged cell on high brightness mode.
Wall at 50cm distance (1 sec F2.8)
Wall at 1m distance (1 sec F2.8)
Wall at 2m distance (1 sec F2.8)
Bed at 2m distance (4sec F2.8)
Aircon at 3m distance (4sec F2.8)
Thank you for viewing.
-- dj