Big Project LED's

LScamper wrote:
Gdetrailer wrote:

"Half Wave rectification without filter caps will result in HALF the voltage at 89V in theory but in practice generally you can count on about half of the AC RMS or about 60V-70V.
Assuming those LEDS are designed from 32V to 36 V or 96V-108V for three in series which should be well high enough."

Not correct but not going into it.

The LEDs will see the peak voltage of about 180V. At that time the current will be at maximum, well over what it is at 127V. The LEDs are sensitive to peak current and will smoke.

Also, dissipating 400W will need a huge heat sink with fans on it. Or maybe water cooling.

Half wave DIODE will only LET ONE HALF of the CYCLE through, PERIOD.

If you don't believe me then perhaps you should break out the O scope...

Putting a STANDARD DIODE in series with the LED string will block the remaining LEDS in series with the blocking diode from ever seeing the PIV. The PIV of the ALL diodes in COMBINATION (because series in this case will ADD to the PIV of the ENTIRE STRING).

The LEDS do not see the PIV at all since the blocking diode will prevent any current flow through the diodes and the PIV of the blocking diode is well above the PIV of the AC.

Now placing the LEDS straight across the mains would be rather spectacular to say the least... Which you could do if you had enough in diode junctions in series but not really a good thing to do..

Where you can get into trouble is if you place a CAPACITOR across the rectifying diode. The filter capacitor not only filters the 60 cycle half wave DC but it will increase the average voltage of the DC..

Now if you used a full wave rectifier then you WILL get BOTH "halves" of the AC giving you a higher voltage unfiltered (120 HZ)DC..

There is a VOLTAGE difference between half wave and full wave rectification.

I have been doing the electronic thing since I was a kid, went to school for it then repaired too many electronic things over the years.

I would be more inclined to just get a fixture that works.


http://www.superbrightleds.com/moreinfo/high-powered-fixtures/high-power-50w-led-flood-light-fixture/1548/

http://www.superbrightleds.com/moreinfo/off-road-lights/9-quad-row-heavy-duty-off-road-led-light-with-comboselectable-beam--108w-/1586/

OK, FIVE 100 watt plates? Hoping the watts area under curve would be small enough for the devices to take it. Dang, four power supplies are almost 200 dollars plus shipping. Way more costly than the plates.

The warehouse has 3 85 watt 4,500K CFL's and still looks like an undertaker's parlor.

Here's the heatsink I was thinking of...

http://www.allelectronics.com/make-a-store/item/HS-176/17.25-ALUMINUM-HEATSINK/1.html

Gdetrailer wrote:

"Half Wave rectification without filter caps will result in HALF the voltage at 89V in theory but in practice generally you can count on about half of the AC RMS or about 60V-70V.
Assuming those LEDS are designed from 32V to 36 V or 96V-108V for three in series which should be well high enough."

Not correct but not going into it.

The LEDs will see the peak voltage of about 180V. At that time the current will be at maximum, well over what it is at 127V. The LEDs are sensitive to peak current and will smoke.

Also, dissipating 400W will need a huge heat sink with fans on it. Or maybe water cooling.

Although (thinking outside the box..)

For what Mex is trying to do, perhaps instead of going the high wattage LED route and having to blaze a DIY path into uncharted territories..

Take a long look at high wattage CFLs...

For instance you can get a 65W CFL HERE for about $11. This gives off about 4500 lumens.

Then use a reflector to concentrate and direct the light a bit..

Like THIS for $7

Keeping in mind that the large wattage CFLs are about 12" in length and often will not fit into regular lamps but as long as you get the medium Edison base bulb it will fit any standard bulb socket. You do need to watch with these as they do come in larger base style and or may not have a built in ballast..

I have a 85W CFL in my garage which puts out more light than a four foot long twin tube shop light..

Need even more light, they make CFLs in even higher wattage levels like 105W and if I remember correctly 200W also.

Where I work at, they replaced ALL the high bay HID lights a few years back with six tube four foot fluorecsent fixtures.

We gained more light, less glare, better color rendition with a vast savings in power consumption and maintenance (paid a third party to rent a man lift and relamp the burned out HID bulbs every 3 months).

A year and a half ago, I changed out a 175W mercury vapor outdoor light at my parents house to a 42W CFL outdoor light. The 42W CFL is considerably brighter than the old 175 mercury vapor light ever was...

LScamper wrote:
MEX wrote:

"(4) 33 volt 100 watt LED's wired in series, with a 25 amp 800 PIV stud mounted rectifier for 127 vac operation. Dang I wish this would work, it would save me a lot of money. Or then it might make the prettiest smoke signal"

Don't forget that the peak voltage is 1.4 times the RMS voltage. 127 x 1.4 = 178V! It will make a pretty smoke signal.

Half Wave rectification without filter caps will result in HALF the voltage at 89V in theory but in practice generally you can count on about half of the AC RMS or about 60V-70V.
Assuming those LEDS are designed from 32V to 36 V or 96V-108V for three in series which should be well high enough.

However with the conditions of the electric service there I would be prudent to use a current limiting resistor (but that would require a very large one). I would recommend using a switching power supply (designed for such use) for each one which would protect the LEDs and provide a constant current source... Not a totally cheap solution but would be better in the long run...

MEX wrote:

"(4) 33 volt 100 watt LED's wired in series, with a 25 amp 800 PIV stud mounted rectifier for 127 vac operation. Dang I wish this would work, it would save me a lot of money. Or then it might make the prettiest smoke signal"

Don't forget that the peak voltage is 1.4 times the RMS voltage. 127 x 1.4 = 178V! It will make a pretty smoke signal.

Heatsinks tend to be rated using temp per watt for dissipation.

I found a calculator that might be helpful but it doesn't give the square inches required per watt. They do list some info for dissipation of copper on a circuit board which could be a bit helpful..

HEAT SINK CALCULATOR

Most likely if you do a search using "heat sink calculator" you most likely will trip across one that gives square inches of material required.

On edit..

Found a real nice calculator which looks like it should give you the info you need..

ANOTHER HEATSINK CALC

More info on heatsink design..

HEATSINK DESIGN

75 cents kWh, wow! If they allow net metering then a grid tie system would be a slam dunk with maybe three year payback.