I made a mistake, in my description of a ultra simple current limited LED driver.

Call the cyber police to take my PCs from me.

?

[Me making a mistake ought to make at least one person in this group smile. I live to please.]

?

I transposed the position of two components.

?

The capacitor goes between the AC Hot and full wave bridge, not the other way around.

?

The AC from the mains passes through the capacitor, the current is determined by the capacitor value (Xc) and mains frequency.

?

Higher capacitance equals higher current.

?

The now current limited AC passes through the full wave bridge producing 120Hz pulses of DC (100Hz pulses of DC in 50Hz nations]

?

This safely (if the capacitor value isn't too large) provides current to power/drive a LED or LEDs. My fake florescent LED replacement tubes have a lot of small SMD LEDs. I quit counting at 50ish. I'd guess there are about 200.

?

Note that is a guess and I won't pretend it is "mainly accurate," don't know, don't care.

?

I laid it all out on my work bench, out of the protective plastic tube, and used a capacitor substitution box to pick the 'right' value capacitor.

?

I kept feeling the LEDs and when they were just perceptible warm I backed down one step.

?

Is the brightness optimal? Snort hardly.

?

Is it efficient? Give me a break, nope, hardly.

?

But 4 of them provide about 10% more light as 2 real florescent lamp tubes.

?

Oh that 10% figure comes from a 50 or 60 year old professional light meter that, in 2008, agreed with a new, NIST traceable calibration, light meter, so I suspect it hasn't degraded in the 16 years....might well be wrong ...sort of "I hope it's mainly accurate."

?

Best I could do on a moment's notice. I need to take it in to the lab for calibration....but then what does the odd spectra do to calibration....

?

And yea I had to rewire the fixture. Oh how sad. I dispensed with sockets because it was cheaper and easier, I'd have had to go to a real electrical supply house for sockets and they might not sell to a civilian they don't know, my EE diploma would get laughs, and EE can't change a light switch for hire in Kentucky. So I hardwired everything and used home made Lexan spacers. Looks more then OK, actually looks store bought if you don't look too close....but who cranes their neck to inspect a florescent light fixture?

?

The LEDs face down so the reflector is sort of useless.

?

So yea WN4ISX made a mistake.

?

Yawn.

?

?

I'll tell on my self.... during a graduation when we used rented professional 2M business band radios and I ID with my ham call instead of my name. That was a really big no no, but I was used to using my ham call on an HT and habbit won over the law. And there were negative consequences.

?

The company monitored all comss with their radios and reported me to the FCC. The company owner was beyond upset. I have no idea why, it wasn't like anyone was going to waste time listening to 3 TV crews rush from staging area to staging areas and trying to set up with zero time.

I was part of it and it was BORING!

?

?But the Freaky Candy Company has zero sense of humor.

?

I still have the pink ticket framed on my wall in a place of pride, Paraphrased is says "Thou shalt not make this mistake again upon risk of forfiture of license."

They get real twitchy over IDing the wrong license in the wrong radio service.

?

[Again, damn sure wish I could spell.]

?

?

Sorry I'm far from perfect but mainly accurate. At least I got the link to the graphic of the diagram right...

How fast an oscilloscope must I have?

?

I used a DSO rated at 200MHz.

?

I used ~120V [118V RMS], peak voltage 168ish, a 0.5uf [0.503uF] poly capacitor, a 1 ohm resistor to give 1V per amp, the CT-1 current probe, the scope in one shot storage mode and a push button NO switch.

?

I've pushed the switch about 50 times and I never get more then ~0.04V across the resistor. That equates to ~0.4A, steady state was ~0.03V/A.

?

Is my scope too slow to catch the transient?

?

Perhaps I can't press at peak voltage... but I'd expect that at least once in 50 would result in a peak voltage. I might could rig a NE555 to start on zero crossing and turn on a SCR at peak, but that's more stuff on the 120V AC Mains then I like.

?

I'll think this over and try again tomorrow or Thursday, we have water exercise and I suspect the PT gal is going to add some new fun filled exercise and I might not be up to doing anything serious.

?

I've learned to not mess with 120V when you are tired. Too easy to make mistakes.