next topic »

[Reply #15 on:]

About cheap Ebay leds. They don't stop working,

The ones I got did. Kept below 20mA as well. These were RGB and it was always the red that was dim anyway except two which were very bright. As I said rejects.
The dimming of the LEDs is the normal ware out phenomena, this is accelerated by low doping levels.

[Reply #16 on:]

How can you have a ground on the scope AND measure the difference between points A & B?

Sorry, I was not precise. Ground probe was connected to point B all the time. The second probe was connected to point A. With this setup I measured  12 V DC (pwm value = 255) and that sinusoid voltage (pwm value = 0).

Btw, if place a rectifier diode between +12 V and point A on my picture to cut the negative halfwave, would that be any help for LEDs (they would not be exposed to reverse voltage)? Price of a voltage drop due to diode forward voltage is acceptable. Do I miss some other drawbacks?

[Reply #17 on:]

Btw, if place a rectifier diode between +12 V and point A on my picture to cut the negative halfwave, would that be any help for LEDs (they would not be exposed to reverse voltage)? Price of a voltage drop due to diode forward voltage is acceptable. Do I miss some other drawbacks?

Other than the increased part count and voltage drop, both of which seem accecptable to you, there really aren't drawbacks to adding a rectifing diode for this type of application.  It would be nice to find-out exactly what's causing the ~5 V_AC sinusoid, but that might not be practical.

[Reply #18 on:]

Having a scope ground at point B probbly resulted in shorting out the FET through the scope unless your scope and arduino supply were floating. Scopes are normally connected to ground and so is the computer powering the arduino. I doubt if the voltages you were seeing we're real. One way to find out is to put a 10K resistor across the scope to lower the load impedance.