Acriche series Where to Buy? Some Pics added

[VidPro]

Acriche series Pics , Emitter pics that show its secrets

50 useless press releaces later, a 20% increase in output, and a reduction in costs, since 2004 the Seul AC driven led contraption has been out. whats it take to convert a press releace to a actual purchace :thinking:

anybody know a source for a few of these things?
i wanted to attempt an assembly of crees, and some sort of "ac" driver , then i saw stuff like this http://www.zywyn.com/files/zd831+zd832_flyer.pdf (only a measly 30ma)
where they convert without fat transformers, and high lossy stuff, which led me to thinking this Achiles heel oo: product thing might be easy to implement.

i want to recycle dead CFL frosted vanity bulbs into led bulbs , i will prefer to use many high powered led items, run at low power.
like say 4-8 crees run around 160ma or so.

i donno, but i am willing to try something to "mod my way into" SOME sort of Frosted decorator dome bulb, that uses cree or seul leds, doesnt cost an arm and a leg, runs forever, has a descent curcuit and all. everything out there now, would fry your retnas with little light spots in a bathroom vanity situation. Plus they drive Less LEDS at fuller currents, if i am going to plunk down $30 for area led item, i want Many many high powered leds, run at low efficent long lasting powers. not hard driven 5mms or less pieces.

If the Acriche thing is nothing more than a cheezey way to shove AC. PWM or Dirty DC into an emitter, and say it works, then i am not interested.
but if they have actually cleaned up AC completly, with some magic low loss stuff, better than a switching PS does, then mabey its ok.

if they do stuff similar to this http://www.ledsmagazine.com/features/3/5/2
in any way shape or form, i am totally uninterested, as i have tested different methodoligies of this method, and its like tossing pearls (leds) to swine, i am sure that methods similar to that will be used in a lot of the "AC" lighting cheap junk, but i already been there done that, and its worse than anything.

when are these people going to realise, they can pump out a bunch of spam press releaces till they are blue in the face, but until people can BUY them, were just standing in another Costco line. (costco line meaning, if your going to sell people stuff, the least you could do is let them PAY and LEAVE with it)

CPF search of the few people who have even seen or used the things so far, and so long ago, that the output and costs are now where the product might be usable.
http://www.candlepowerforums.com/vb/showthread.php?t=142399&highlight=Acriche
LED uses 40% less electricity than Fluorescents

http://www.candlepowerforums.com/vb/showthread.php?t=171466&highlight=Acriche
120VAC LEDs

http://www.candlepowerforums.com/vb/showthread.php?t=171728&highlight=Acriche
Durability of 110V LED bulbs?

Seoul Wizz bang product hype
http://www.seoulsemicon.com/_homepage/home_eng/asp/news_content.asp?news_idx=29

 

[James Jackson]

Re: Acriche series Where to Buy?

Well... I'm not sure if this is what you're looking for or not...

http://www.mouser.com/search/ProductDetail.aspx?R=AW2211virtualkey61510000virtualkey889-AW2211

Mouser has 'em (currently 17 in stock). I know 'cuz I bought 2 of the dual LEDs a few months ago, and have one mounted on a 'donor' CFL lightbulb socket. I am currently using it in my bathroom to light my mornings.

They are really neat. They work. They are real.

Regards,

James Jackson
Oztronics

 

[DaEmpty]

Re: Acriche series Where to Buy?

anybody know a source for a few of these things?

http://www.leds.de/index.php?cat=c2465&cPath=21_2465&language=en

 

[SteveDavis]

Re: Acriche series Where to Buy?

VidPro

If you do decide for whatever reason that the Seoul part is not for you, I have several driver circuits I can recommend that would not need a transformer. Power factor correction, and soon, standard incandescent dimming will be available. These are qualities you will not get out of an LED designed to run off of straight AC, and can make a big difference in terms of efficiency and functionality.

 

[VidPro]

Re: Acriche series Where to Buy?

THANKS.

there are things i want to "know" first. By buying a few, which i am hoping to find out.
i talked with the USA rep for Seoul and he seemed to indicate that it IS a multidie chip with series and reverse series (or something) , that "Runs off of AC" , not that its a embedded curcuit that does any conversion. :-(

so indeed i probably would need help "fixing input" if the method is not "perfect" for the led die, to clean it up a bit.

he is going to have some EE dude call me back, but the language barrier is very damaging to communication, they use different tech words, and i am not even a tech. and i think they might send me samples if i provide test data.

if at least it is a "high voltage" led array, then there is one less bunch of losses in conversions there, and less parts and pieces.

 

[lotsaluck]

Re: Acriche series Where to Buy?

We are currently testing several dozen of the 2 and 4 led versions for various applications. You must order the boards in the voltage you require. The boards are assembled with dropping resistors for the various ac primary lines in use. The regulation occurs inside the led. Our results thus far are very promising but cooling them is somewhat problematic. Many sources available.

 

[VidPro]

Re: Acriche series Where to Buy?

We are currently testing several dozen of the 2 and 4 led versions for various applications. You must order the boards in the voltage you require. The boards are assembled with dropping resistors for the various ac primary lines in use. The regulation occurs inside the led. Our results thus far are very promising but cooling them is somewhat problematic. Many sources available.

well just keep pushing out data , i am listening
pick your brain, i want to know anything you think about it, observed or measured, or can see in the led, or have seen in dissasebly.

so the resister was added to the 220 boards? or both of them? are you in 220 land?
did anyone measure the voltage across the emitter itself ?
any real specs on the actual current consumption?
any ideas from observing the led Die thing under the dome what is being attempted?
any strobing, of any section of the led?
effects of driving via cheap ac dimmers?

i know Nothing (as usual) any real user data that you have would be helpfull. Need Input.

 

[SemiMan]

Re: Acriche series Where to Buy?

Steve, I don't think you mean "standard incandescent dimming" but TRIAC dimming and yes that will work just fine for the Seoul product. Power factor is certainly not perfect for it, but as the regulations do not hit power factor at that low a power it is not an issue yet.

Semiman

VidPro

If you do decide for whatever reason that the Seoul part is not for you, I have several driver circuits I can recommend that would not need a transformer. Power factor correction, and soon, standard incandescent dimming will be available. These are qualities you will not get out of an LED designed to run off of straight AC, and can make a big difference in terms of efficiency and functionality.

 

[jtr1962]

Re: Acriche series Where to Buy?

We are currently testing several dozen of the 2 and 4 led versions for various applications. You must order the boards in the voltage you require. The boards are assembled with dropping resistors for the various ac primary lines in use. The regulation occurs inside the led. Our results thus far are very promising but cooling them is somewhat problematic. Many sources available.

Do they have any visible flicker? That's really my biggest concern with these. For years until electronic ballasts came people complained about the flicker from flourescent lamps. Now if this approach produces the same flicker with LEDs, then the LED manufacturers are just shooting themselves in the foot.

On another note, what's so bad about just using regular drivers? A 120VAC LED driver can be upwards of 95% efficient. Just because most currently aren't due to poor design doesn't mean they can't be. Frankly, I want to puke when I see some of the 120 VAC LED ballasts. They're as big and hot running as the fluorescent ballasts. Why is so much power wasted in them when off-line driver ICs exist such as the HV9910 which can be >90% efficient overall? And why are all these LED ballasts so big when all they need is an IC and a few small passive components? :shakehead It seems that the ballast designers just want to be able to use the same case sizes as their fluorescent ballasts, and couldn't care less about efficiency.

 

[SteveDavis]

Re: Acriche series Where to Buy?

SemiMan-

When I say standard incandescent, I mean either leading or trailing edge dimming, which is either triac or ELV. I have no doubt that these LEDs would turn on under a triac signal, but I do doubt that they would dim well.

My understanding of the internal structure is that there is a series string of tiny chips in either direction, and there is no circuit to regulate current, other than maybe a resistor. That means that for a good portion of the time in the AC wave, the LED is off. When you try to pass a phase dimmed AC signal into that, at the very low and very high end of the phase, you're going to be cutting the wave where the LED isn't on to begin with. The fewer dice in series, the less that's a problem, but also the fewer dice in series, the larger your resistor must be. So either you've got a really inefficient source, or a source that doesn't dim very well. Even if they use a linear regulator to regulate the current, they have potentially a very large voltage across the regulator, driving down efficiency, or more die, decreasing on time.

With a smart switching regulator, you can both power the LEDs and sense phase angle and dim accordingly, and not lose any angles to the LEDs being off as is. Plus, it's not hard to build in PFC, and you can actually make a fixture that is high power enough to light a room and still meet energy star.

 

[SemiMan]

Re: Acriche series Where to Buy?

Do they have any visible flicker? That's really my biggest concern with these. For years until electronic ballasts came people complained about the flicker from flourescent lamps. Now if this approach produces the same flicker with LEDs, then the LED manufacturers are just shooting themselves in the foot.

On another note, what's so bad about just using regular drivers? A 120VAC LED driver can be upwards of 95% efficient. Just because most currently aren't due to poor design doesn't mean they can't be. Frankly, I want to puke when I see some of the 120 VAC LED ballasts. They're as big and hot running as the fluorescent ballasts. Why is so much power wasted in them when off-line driver ICs exist such as the HV9910 which can be >90% efficient overall? And why are all these LED ballasts so big when all they need is an IC and a few small passive components? :shakehead It seems that the ballast designers just want to be able to use the same case sizes as their fluorescent ballasts, and couldn't care less about efficiency.

HV9910 "CAN" be greater than 90% efficient is the operating word. If you are only powering a few LEDs, say 4-5 or less, then getting 95% efficiency in a switching power supply is extremely difficult. Even 90% is considered quite good for lower voltages. If you have a higher voltage string, say 40-50V, then you can get the efficiencies up, however, there is the issue of cost, size, etc. That HV9910 for a low voltage string may only be 80% or less efficient.

Why are they so big? Well part of it is cost. It is cheaper to make a big one in some ways then a fast switching smaller one. However, a proper AC\DC ...isolated that is, has a lot of things in it. If it has power factor correction, even more. You will have chokes on the input to reduce radiated emissions (absolutely necessary), possibly some protection circuitry, switching circuitry, transformers, possibly some more chokes to keep the radiated emissions down, etc.

Semiman

 

[lotsaluck]

Re: Acriche series Where to Buy?

No visible flicker from them and the double are drawing 10ma. We have a problem with one of the four led boards now after it heats up in use one led starts to strobe at about 1 hz. All others are still burning bright after over 1k hours on them. I have put some in as under counter lights now with great results.

 

[VidPro]

The part is a AW2211 110v 2 emitter board

Note:the pics might not show up on some moments, due to the comcast server thing
they are there, and will show eventually , but a high load might stop them from showing :-(

Look Real close, and you can see the foil traces there, it might be a clue as to how the stuff is connected
see the slightly darker lines comming from the connection points.
the base is enamel gloss white not raw aluminum
Width 20mm
Length 40.1mm
thickness of board 1.9mm
total height with emitter dome 7.4mm
the dome is a squishy

Question, are those 47ohm or 470ohm resisters in there?

 

[VidPro]

a closeup of the emitter, squshed into a flatbed scanner, the dome gets crushed a bit, the 2 yellow emitter slat items seen are the SAME size , although the pic magnifies them different sizes.
everything else you see exists, 1 wire goes to an emitter thing, one goes to the next emitter thing, and one wire back out.

Powering the emitter item only with 1-30V DC does nothing, i will have to rig something to get more voltages to see what lights when dc is applied.

another picture here https://www.leds.de/popup_image.php?pID=476&imgID=0 of a 230V emitter only

the squishy dome is about 5mm in diameter, and 3.5mm in height from the black plastic and up (so you can compare it to other squishy domes)
the black thing is ~8mm in diameter (minus any wire stuff)

 

[LEDite]

VidPro;

There are TWO 470 ohm resistors in series with each LED !!

Not very efficient.

I would use a small switching supply. In my AC lights, I have found some that are 90% efficient:

9 Watt AC Light

Larry Cobb

 

[VidPro]

i like it larry.

so it IS 470ohm resister, hmmm.
unfortunatly most of the led Calculators dont go that high voltages , so i would have to do nasty math stuff to guess the voltage drops, and total losses, anybody assume what the approximate voltage is across the led item, so i can test doing DC?

here is a plan i have, as useless as it might be. if anyone already knows anything it would be helpfull.

1) find out the actual trigger voltage of the emitter , and drive voltage at or near spec

2) find out how it runs on DC not AC. as a high voltage emitter, so to use less pieces to get to high voltage without transformer or switching power supply.

3) make up a bridge curcuit with caps only to drive it, simple, cheap, small, and no frequencies other than the 60cycle AC.

4) use multiples of the emitter , to get to the ~100v or ~160V DC drive, without much stuff, or extensive electrical knowlege.

5) drive it at 1/2 spec current by voltage matching, plus capacitors for averaging or some minor resistance,

6) wedge 2 Sets (4 emitters) into "vanity" CFL base (frosted dome), so the aim is 240* approximatly in a V pattern. (i dont need it to hit the BACK of the fixture, and i dont need it to go up/down much, but i do need it to spread from the sides, not point straight into the person)

7) have the "driver" or conversion be less than 1 cubic inch total approx, so it fits where the CFL driver is, and doesnt take heat sinc space, and still wedges into the ~3.5" domed bulb =O .

8) have a "folded" heat sinc with chimney effects, then the outer caseing of the fully enclosed CFL bulb (frosted dome) will have holes drilled to let air in the bottom to travel out the top.

9) dream smaller next time, so its easier to do

 

[VidPro]

ok, some preliminary info, i am working on it still.

AC -----
at a VF on the led emitter thing (minus the resisters)
90V AC (rms) i get 50ma of ac and 10ma of dc showing when in Ma setting
on the 10amp shunt i get .02 amps. my ac is around 120V (not 110)

DC --------
putting 60 volts DC into the emitter, and i still get nothing.

finnaly i bundle together enough power supplies and get light with DC

Now i am going through the resisters (dont forget that)
at 85volts DC on the emitter the gates Just start triggering, just like .02ma

95volts DC 100ma its glowing pretty good but its still Cool in temperature?
err isnt that some 9 watts for the 2 items?
i am confused, but its not neer its possibilities, so wish me luck

i am sooo confused
at 103V DC its at about 200ma, and barely getting warm
hmmm i know reverse it, what the heck, i got nothing to lose, i will reverse the polarity

Resisters --------
voltage drop across onboard resistance (above readings were VF of emitter) at ~111V input there is ~103V at the emitter, or ~8V drop on the resister


ok, i think i have it, i saw it with a 30X magnifyer, and i think i got a picture of it, i think it might just answer the elusive question of "how does it do that" Pictures in next post, gimme a while

 

[VidPro]

Re: Acriche series Emitter Closeups

Ok here are the macro type pics
drive current is very low so i can see the seperated items.

ALL of these pictures are done when driven with DC, so dont be fooled by the pic, there is 2X as many gates, and the whole emitters base is covered in them , without the space seen between them.

this is a VERy precision set of series items to have this kind of balanced output at the low current point, other series items would have much more differences on the array, this certannly isnt Sloppy like other series items can be.
on the other hand it looks like where the wires plops down, effects the first gate the wire is at, slightly.


This is what it looks like when its light up, just so you know

these last 2 pics below are Very special.
the first one is with one polarity (+ -)
and the second one is with reverse polarity (- +)

Ok now reverse the polarity , and wala, the other set lights up

as you can see in the last 2 pics, is that one array is lit on one pic, and the other array is lit up in the other pic. 2 interleaved multi arrays , working sorta like cheap AC LED xmas light strings.
when run with AC each array group lights 1/2 the time.

 

[VidPro]

so how many gates do you count , per array?

ok soooo, if i am going to drive this with DC, to keep Spikes off the gates from the sine wave stuff, i would be stuck lighting only 1/2 of it.
unless
after i convert to dc , trim down the current with something like say a resister, and some PWM curcuit reverses the dc.

IMO this item would run "best" on some sort of Pure Square Wave AC , with a high frequency like say over 300, and trimmed to ~110v max (not rms).

so anybody have any ideas on running this with dc?
run 1/2 of it, and include a switch to trade arrays when its aged.
try and make some kinda square wave driver
give it up, waste of time, run it off the ac, and just have it socketed for replacment.

remember i want to run it with high voltages, and minimal conversion and transduction (aka stupidly simple) but at the same time i dont think that an AC sine wave is a good way to constant current to the gates.


another problem becomes that i cant use my cute trick method of simple rectification (no transformation) of wall socket and (almost) direct drive, because there is no matching combination to get to a drive voltage.
160-170V DC would not light up 2 of the items. :sigh: resistering even with say a capacitor, would waste 40% of the power. :sick2: so its not the high voltage array i was seeking. mixing it up with other led items, in series, as "resisters with light" would be difficult too.

 

[VidPro]

i just connected up a very loose (guess and hope) rectification for it, that works ok.
i use a 1.3K 1/2W resister on the ac side
a slow blow fuse is on the ac side too, in case something goes wrong
that goes into a very small bridge rectifyer 300V 3A
on the dc side , it goes into a 100uf 350V capacitor
then that side is resistered down with another 1.5K 1/2W resister
then i plugged that into the AC wall socket (of course)
i get about 103V DC to the led, and it looks ok.
there was about 17V or so dropped at my added resister, heat on the resister was probably just right for the 1/2w, for long term use.
i figure i want about 20-30ma current on the array total?

then i will add in a tiny fast blow glass fuse of say .1amp to keep led failure from occuring during a wild surge

so from my LOOSE attempt, i figure that mabey 2 x 1K resisters should about do it, by putting one of them before the charging capacitor, and one after, the capacitor charging doesnt get slammed so bad on turn-on.

its quite a large array area, so i figure i can probably push it to 3W+ per item without the ac spikes, and running 1/2 the arrays cleanly (DC) should be ok.

of course i am guessing, and need major help with the Math stuff , but heck its working, and its not getting to hot.

next thing was to hack up one of the dead Frosted vanity CFLs, and then more of my plans went to heck errr the cfl bulbs all have GLASS domes on them, i thought they were plastic (with glass FL tubes of course). sooo now i gotta find frosted plastic globes and sockets too.