Are there capacitors that would fit in a pill and could usefully filter out pulse width modulation? I imagine that would only work for PWM that is too fast to be directly bothersome, but it would improve the efficiency, because the light output per current curve has smaller slope at higher current, and because the driver would see less voltage when putting out its current pulse. The led would be more efficient, boost and buck drivers would be more efficient, and current regulating drivers would be less sensitive to cell voltage. I am inspired in this by the single old fashioned wire lead capacitor that stuck out of the surface mount VDSL transceiver I worked on. I think the gain would not be as much as one would get from the data sheets, because their data is for constant current, so heating effects are included. But led efficiencies do drop off, even at constant junction temperature, at least because of detrapping.
It would decrease radio frequency emissions.
One can also think of adding capacitors to the input to make it easier on the battery.
I thought about trying the same thing, using a small Electrolyte 10v 250mf cap across the LED wires to “smooth” out the PWM. it would likely make the lower modes brighter & less spaced, as its the “on-off” characteristics of Pulse-width-modulation is what helps lower the brightness of the emitter in the lower modes.
Nope…
That sounds like an excellent idea, try it and see!
solarbotics has a TINY supercap (1F) that might do it if you wired two in series.
this has been mentioned before, and someone came up with some number (2300?) microfarad minimum before you start to see a difference in the waveform.
It would take 2 wired in series to handle the voltage, 3.3vdc is nominal output, you would need the combined at 5vdc to be “safe”
Interesting idea though for sure
That seems like it would do it, when there is that much left over space. I wonder how sensitive it is to heat.
All those tiny supercaps are meant to supply microamps of power for memory backup. They have a HUGE ESR value. Utterly useless…
In the example of an 8*7135 driver it would be much better to disable some 7135 chips to get a lower current. No PWM and the current would be perfectly constant… but you can’t go below 350mA.
This solution would take only a little bit more space and require 3 a µcontroler with enough outputs to control each 7135. but it would take less space and be cheaper than adding a cap.
That does seem to do it fir 7135 circuits, though it is a more complicated modification. What about boost and buck drivers, though?
What is an ESR value?
Equivalent Series Resistance… it makes a perfect cap look like a cap in series with a resistor.
PilotPTK and texaspyro contributed a lot about a thread regarding caps to smoothen PWM. I’ll link to it if I get ton a computer.
Edit: Here it is: Can a capacitor be used to smooth PWM?
Unfortunately Pilot’s photos are no longer available.
I honestly don’t know anything about ESR values, but it was designed for “hybrid battery applications”, with an ESR value of .4 ohms
For comparison HKJ reports the internal resistance of an Eneloop XXX AA to be 0.05 Ω.
hope he’s alright. He disappeared off the face of the earth.
true, but what is the resistance of an xm-l?
Since an XM-L is a semiconductor, its voltage versus current relationship is not related by a constant such as the resistance for a length of wire.