Calling out to Weezard for LED advice

[Weezard]

Looks good on paper and I see where you are going with this.

Got some bad news for ya.

Just scaling up the parts falls apart fast.
A one amp emitter, or string of same uses 100 times the current of those tiny leds.
Just scaling the capacitor still up leaves huge inrush current, Vmax., and voltage spike problems.

A capacitors impedance is frequency dependant.
Higher frequency - lower impedance.
Voltage spikes sail right through them!

A tenth of a volt change is a big deal to a non-linear resistance like a led juction, and most line voltage is very far from stable or "clean".

Coupled with the peak voltage being 1.414 times the RMS voltage it becomes apparent that un-regulated AC is not a viable source for high power LEDs.

No shoot da messenger.

Aloha,
Weezard (electronics dick)

[mx4intx]

Ya I didnt say it was going to be easy but it makes me go hmmm.

[Weezard]

Ya I didnt say it was going to be easy but it makes me go hmmm.

If you just must try it, here are some tips:

You need to hang a MOV, (Metal oxine varistor, and a small capacitor across the input, use a diode bridge with 4 more caps for rectification, a large electrolytic cap for DC filtering along with a tantalum cap for HF noise on the output. And a series inductor across the electrolytic to smooth the "sawtooth" ripple.

The inductor adds yet more problems at high current, so now you need "steering diodes to sink inductive spikes.
All that helps, but is still a risky supply for non-linear loads.

That can be addressed with an AC line regulator,
They are not inexpensive.

And without pre-regulation and transient suppression,
the complexity increases as does the parts count.

I found it less expensive and less worrisome to just do it right the first time.
(well, in my case, the second time)

A DC current limit supply is the best way to drive leds.
YMCV*, but I doubt it. (*Your Mileage Could vary. You might be a very lucky person.)

Do it any way you wish, of course, but keep in mind, once that "magic smoke" gets out da leds are toast.

Aloha,
Weezard

[khyberkitsune]

I dont think you understand the circuit.

heres another one

Circuit - AC Powered White LED Strings

I understand the circuit just fine, sir. I design LED and Induction lighting for a living.

You're not understanding something.

You can't simply scale things up and expect it to work, electrical engineering simply doesn't work like that. You've barely even figured out your conductor resistances, how are you going to match the circuitry to that to avoid creating inductive loads?

Sure you MIGHT get it to work with what you're thinking but you're about guaranteed to destroy the diode in the process.

[Weezard]

For me at least.

Cheap is mandatory.
Lack of Safety :blueknife: is a deal breaker.:toilet_claw:
And, Simple, unto elegance is my ultimate goal.
Easy, would be very nice as well.:rastasmoke:

This combo will usually require compromise.

Just think about it fo' a while.

Weeze

[mx4intx]

My conductor resistances in case of inductive loads?

An emp?

Dude, you dont design anything I think anyone who has read your posts will agree.

Weez, safe and simple is what I am looking for. I am just looking towards the future and a much larger array.

[mx4intx]

Maybe I will just see if Seoul gives away the secret in an Acriche pdf of how to run high power off AC.

[khyberkitsune]

My conductor resistances in case of inductive loads?

An emp?

Dude, you dont design anything I think anyone who has read your posts will agree.

Weez, safe and simple is what I am looking for. I am just looking towards the future and a much larger array.

Video - TinyPic - Free Image Hosting, Photo Sharing & Video Hosting

Welcome to my research facility. You do not know anything about me and I wish you would quit assuming such.

We've run the AC and DC tests, multiple times. We even made an LED T8 retrofit that runs directly off of a fluorescent ballast OR can run directly from mains voltage. BTW we were using Ariche diodes for those bars.

Guess what? 70% failure rate. We dropped Seoul Semiconductor like a rock and I am now working with Nichia to create the first true multi-band horticultural diode.

I don't design anything, such a laugh!

[mx4intx]

Video - TinyPic - Free Image Hosting, Photo Sharing & Video Hosting

Welcome to my research facility. You do not know anything about me and I wish you would quit assuming such.

We've run the AC and DC tests, multiple times. We even made an LED T8 retrofit that runs directly off of a fluorescent ballast OR can run directly from mains voltage. BTW we were using Ariche diodes for those bars.

Guess what? 70% failure rate. We dropped Seoul Semiconductor like a rock and I am now working with Nichia to create the first true multi-band horticultural diode.

I don't design anything, such a laugh!

I have seen your video before. Like I said I read all your threads. Funny you didnt have anything to say about my conductor resistances any more?

Also funny that it's "my facility" and the rest is "we have" done this and that. I doubt you are anything more than a helper. Reading your threads saying you do this and that but any time a technical question is asked you refer them to Weez or Knna or that one guy Weez helped before.

Me? I did 10 months electronic training to fix radios in the USMC. A hardcore 10 months, equal to a 2 year AA.

Calling up some Chinese manufacturer doesnt denote design and certainly hasnt helped you know what you are talking about.

[mx4intx]

This thread has so much awesome information. Thanks Weez and DH. I am reading it again...sad I know. But I read it twice and I'm still seeing new things and things I came in and asked.

Sorry if I am gunking it up last few posts.

Still waiting on the tape. Bah humbug.