LED Driver

Dodging the efficiency question for the moment, be very careful with this method of driving them - remember that if one LED goes open circuit, there with be full mains voltage between all the wirring attached to its cathode and that attached to its anode.

"energy "wasted" in a driver circuit"

Energy wasted = 'Volts drop across driver circuit' times 'Amps flowing in it' = Watts wasted.

"if this would be a good "driver" design"

No. it's not. A constant current driver is a good driver for Led lighting system.

Please keep in mind, the previous poster's comments, if you are bent on doing it your way.

Unfortunately there are so many ways that one can change the 240VAC to the approximately 2+ VDC across each LED. Because of this there will be many different efficiencies and reliability expectations that will change with each circuit topology.

I just noticed that D1 needs to be reversed to make a full bridge.

consider it an easter egg for those who are knowledgeable enough to find it.

Actually I have used the same basic circuit with out R1. Even more effecient

Please check energetic forum Joule Ringer Thread for LED Drivers and Joule Thief for Wireless LED's

for quick and dirty the cap in series with a FWB and a cap on the output is great. You might also try IRF.com although others have chips as well. I just happen to use a lot of IRF chips in my work.

https://ec.irf.com/v6/en/US/adirect/ir?cmd=eneNavigation&N=0+4294837988+4294953424

'Quick and Dirty' should also include a rectifier diode in series with the string. Why?

As a rule, LEDs have a very low Peak Reverse Voltage. The first silicon-carbide blue LEDs had a PRV of less than the forward voltage. Something like 1 volt.

You want a rectifier in series with the string to protect it from catastrophic failure should a transient exceed the net PRV of the string. If it does, there's a chance it'll hose the entire string. If there are a few weak units present, they'll blow first - as shorts, usually. This, in turn stresses the remaining units and some of them will blow next, then some more, then more, like dominoes. By the time the transient has gone, so has your string.

When working with diodes, especially specialty diodes like Esaki/tunnel diodes and such, you cannot assume they can withstand any reverse bias. Reverse biasing some types of diodes can instantly destroy them unless the current is externally limited in some way.

As a general rule, 'diode' does not necessarily mean 'rectifier'. Some diodes don't rectify at all. For these devices 'diode' refers to the number of terminals on the device and that is all.

Thanks everybody for the inputs.

I want to know the system efficiency of the LED lighting system. What is power consumption of a good reputed constant current driver available commercially Vs power consumed by the LEDs.

Does this compare well with conventional fluoroscent lamps with electronic ballast? Good fluoroscent lamps are said to waste about only one watt in ballast while rest of the power(about 28 to 30 watt) is actually consumed in the lamp.

There are some good Led chain drivers out there just check some specifications to find out the one that fits your needs

Start with tcmtech's (corected) ckt then disconnect c2. Light output and efficiency will be higher. Check flickering. Then adjust the current with smaller values for c1 till R1 can be reduced, maybe to zero. But you must add a varistor in parallel to the LED string, this is needed for voltage limiting during switch-on.

brgds

Snel

Sizing C1 to match the current desired from the mains frequency and eliminating R1 will improve efficiency more than anything. DO NOT get rid of C2 and don't bother with adding a varistor. A varistor is a dissipation element and C2 will absorb the spikes from turning on at a peak of the sine wave or other transients on the line and then delver the power the chain. It is easy to find a suitable value for C1 as there is a substantial range of current that you can pass through the LED's.

BTW

Cap = Desired current / [2*Pi*frequencyHz*(mains voltage- string Voltage)]

Dont't forget luminous efficiency is higher at smaller duty cycles. Getting rid of C2 will give you smaller conducting angles w/o a pulse generator. Since rsalaskar wants to have 80+ LEDs in series this will not be real current source operation, light output will be sensitive to mains voltage (as it is with incandescent lamps), he must decide how much is tolerable. The same for "bilinking" at 100 Hz, a little worse than at 120 Hz. A properly dimensioned varistor will be cool at steady state, just clip off spikes as you mentioned (especially at turning on).

brgds

Snel

I had not noticed the 80+ led's in the string. I think he would be best off using an IC driver. If he is just looking for lots of light he would be better off splitting them into parallel strings to get a better duty cycle out of the cap FWB circuit / less flicker.

tcmtech is technically correct but that is one cheep cap if the ESR is high enough to cause problems and it is almost a case of why bother with a cap like that. Actually you would be better off with a small inductor in that case.

Since there seems to be some debate on the circuit I posted here is the logic behind the components. C2 is there as a ripple filter and spike absorber and R1 is mostly for ESR bias.

R1 is to increase the series resistance of the LED string to a value greater than the ESR of C2 so that when there is a spike or a short pulse of HF noise the path of least resistance is still into the C2 rather than creating a high current spike into the LED string.

Depending on the size and ESR ratings of C2 that's sort of how you size up R1. The more ESR in C2 the more resistance in R1.

As far as running without C2 and R1 the issue I have had with that is as the number of LED's in the string increases the forward voltage drop increases creating a shorter and shorter on time during each half of the cycle which I for one can sense as a 120 Hz flicker out of the corner of my eye and it gets annoying to me after while.

FWIW I built one of these circuits over 10 years ago from scavenged CFL ballast parts and have had it in continuous operation with a string of 40 green LED's ever since.

Sure there are simpler and equally more complicated LED string driver circuits but for the most part this one gives solid reliable relatively constant current with high efficiency to any common LED set with minimal parts and no visually perceivable flicker.