Facebook
Google
GitHub
Linkedin
The lovely wife Morticia tasked me with disposing (responsibly) of dust buster that the bit the dust. Of course it had a Lithium Ion battery pack. A 5S 18650 pack with BMS and charging. This was obviously too good to throw away. The battery power supply had an on button that went off - half - full. Putting it on the scope it was simple PWM at 3.9KHz 55% for the 1/2 power, and full on for the full power.
I also had an old LED fluorescent tube that I was wanting to repurpose. The two together ought to make a nice emergency light.
The fluorescent replacement tube (circular) consisted of 6 sections, each of which has a series/parallel LED board running at about 10V 300mA. This sounded like a good fit, adding current controlled buck converter.
I came up with the following circuit (some components not was was actually used, but close):
This is a hysteretic (bang bang) converter that is as simple as can be, designed to current control at about 300mA. It simulated well, so I built it on a breadboard.
When the power was turned on full, it performed as expected, giving me what looks like full brightness out of one of the LED boards.
But, surprisingly, it worked beautifully on the 1/2 power setting as well! Scoping it showed the same nice triangle wave running at a lower the current than the full power mode. I was very puzzled as to why the PWM power input did not not seem to disturb it and, indeed, worked perfectly as I would want it. The same triangle wave at less current.
So I went back to the simulation and made the power source the same PWM as the battery pack was producing. Here is the result:
A few minutes, and I understood what was going on. My entire control circuit was out of the picture, because it simply kept the MOSFET on all the time, since the current never rose to the threshold of the comparator. So the pwm input was basically directly running the inductor/Shottky output stage, producing a lower average current.
Sometimes you get lucky.
I also had an old LED fluorescent tube that I was wanting to repurpose. The two together ought to make a nice emergency light.
The fluorescent replacement tube (circular) consisted of 6 sections, each of which has a series/parallel LED board running at about 10V 300mA. This sounded like a good fit, adding current controlled buck converter.
I came up with the following circuit (some components not was was actually used, but close):
This is a hysteretic (bang bang) converter that is as simple as can be, designed to current control at about 300mA. It simulated well, so I built it on a breadboard.
When the power was turned on full, it performed as expected, giving me what looks like full brightness out of one of the LED boards.
But, surprisingly, it worked beautifully on the 1/2 power setting as well! Scoping it showed the same nice triangle wave running at a lower the current than the full power mode. I was very puzzled as to why the PWM power input did not not seem to disturb it and, indeed, worked perfectly as I would want it. The same triangle wave at less current.
So I went back to the simulation and made the power source the same PWM as the battery pack was producing. Here is the result:
A few minutes, and I understood what was going on. My entire control circuit was out of the picture, because it simply kept the MOSFET on all the time, since the current never rose to the threshold of the comparator. So the pwm input was basically directly running the inductor/Shottky output stage, producing a lower average current.
Sometimes you get lucky.
