[<<][electronics][>>][..]
Wed Apr 19 22:55:01 EDT 2017

Pulsed LED power supply

What if it were simpler to create a pulsed supply and use it to create
brightness profile, than to gate a constant supply with a mosfet.

Googling, some more about current vs. brightness curves - not what I wanted.
https://electronics.stackexchange.com/questions/17528/does-pulsing-an-led-at-higher-current-yield-greater-apparent-brightness

It seems better to go for switched current sources like:
http://www.linear.com/product/LT3518

The 15uH power inductor seems expensive comparted to rest of circuit.
This is likely why using a series resistor is still a common thing.


The other LED light I have uses R-L-L-L series, equally distributed.

Doing the same for a 10W LED (1A, 9V) with 3V resistor drop would
require a 3Ohm, 3Watt resistor.  That is a lot.



Ajna light material is pretty woo-woo.  Looking for some clearer material, found this:

Dr.med. Dipl.-Psych. Dirk Proeckl

https://www.vice.com/en_uk/article/futur-humans-lucia-LED-psychedelic
http://old.ilacolor.org/node/523
http://www.lucialightexperience.com/development.html

Hmm not less wonky.

EDIT:
current mode converters
http://www.ti.com/lit/an/snva555/snva555.pdf

Q:
- in a current-mode buck converter, why is there a capacitor?

it filters the ripple current.  for LED driver, ripple might not be
important?

I found at least one circuit that doesn't have the cap, and uses a
switch to ground to charge the inductor.

Circuit could be calibrated to run in feedforward PWM only.

https://www.digikey.com/en/articles/techzone/2011/jun/buck-regulators-make-driving-high-brightness-leds-easy


- how to pick an inductor?

http://www.ti.com/lit/an/snva038b/snva038b.pdf

higher frequencies allow for smaller inductors, but create more
switching losses.  where is the optimum?




[Reply][About]
[<<][electronics][>>][..]