What Resistor Do I Need for These LEDs to Work?

120v ÷ 3.1v ≈ 39. How about two strings of 39 LEDs, with each string wired in series, and then the two strings in parallel? (If it is 240v, one series string of 78.)

Just to clarify -- LEDs want DC voltage; since the LEDs are diodes and will accept current in only one direction wouldn't you need to consider the net RMS voltage the diodes will see? I'm thinking 120 VAC * 0.707 ≈ 85 VDC RMS; then 85/3.1 = 27 LEDs.

(I've used LEDs with DC voltage, not AC; so I'm not sure, just asking.)

Good catch. My main thought was to avoid extraneous resistors by matching the collective resistance of the LEDs to the task.

?Huh? ... in replying to post2's:

"wouldn't you need to consider the net RMS voltage the diodes will see? I'm thinking 120 VAC * 0.707 ≈ 85 VDC RMS..."

I thought that meters (i.e., all multimeter/voltmeters) READ RMS VOLTAGES, 'period'.

As such, the typical 120v outlet IS delivering 120v (RMS).

Am I all wet on this?

I thought that meters (i.e., all multimeter/voltmeters) READ RMS VOLTAGES, 'period'

Unless your multimeter specifically states that it reads AC volts in RMS, then no it doesn't. It will show you an approxiamate voltage. Only certain Fluke models read RMS, there are other brands but only the higher end spec meters read RMS. If you paid $19.99 for your meter then it's unlikely that it will read true RMS.

Not necessarily all read RMS, but it's easy enough to check, by measuring a power point and seeing if it says 'other than' 240 ± 3 to 4V (or whatever it's nominal V is)

This just prompted me to check my "new" $19? Digitech Qm-1500 'bash about' DMM. It said 243 Vac = RMS

In the known to be not RMS were the Micronata (sp?) analog meters, but that's going back a ways. They would typically read ~ 300 Vac

While 'studying' how to correct for meter-errors in the measurement of cathodic potentials, I came across numerous references that clarified this misunderstanding that I had as well...

See:
http://enginova.com/true_rms_volts.htm

and

http://cp.literature.agilent.com/litweb/pdf/5988-5513EN.pdf

Yep - point taken. And some LED configurations could chop it to a square form.

I bow and grovel for your mercy.

(p.s. the second link was a blank)

Can you trust that the reverse voltage will be balanced and therefore within the specified reverse voltage range of the LED?

Oh don't you people "GET IT".....

I calculated all of this out but that was about 6 weeks back so I will give "what I recall" numbers, and not what I measured / calculated - it's just easier right now, because I did factor in and out safety levels of about 10 - 20%....

IN Aus we have 240V AC. Run that into a bridge diode for basic rectification (AC to DC) and from my own experience multiplying that by 1.4 gives a peak voltage I recall 380V DC... but that is the peak of the voltage - not the average voltage.

So I calculated out that with GREEN LED's running at 3.5V DC Max - I could UNDER DRIVE THEM - or give them insufficient voltage to achieve maximum brightness.

I say this from experience that a LED, with a maximum voltage of say 3V - will be almost as bright - sort of imperceptibly so at about 2.6V to 2.7V...

But I found that by assembling a string of LEDS running at about 3V, instead of 3.5V, sufficient to match this equation of 380V / 3V = 120. But that string was actually too dim, and I found that the multimeter read the DC as about 308V... (RMS?)

So I removed bunches of LED's at about 10 at a time, to watch the brightness go up, until they were about the same as a single LED on 3V DC - from 2 AA batteries.

Pretty simple.

Just under drive them by sufficient to cope with most transients and perhaps poke a fat 400V capacitor in the system to add some soak and some leveling of the DC...

Without going totally nutso about the exact frequency (505nm I think) I now have a HEAPS bright light, and being green (not the best like turquoise) that gives me about 2 times as much night visibility as white LEDS, and it uses SFA power in the first place.

Well for a start - that would be 320V smoothed adequately - so 91 @ 3.5V

And because it's mains supply and stable, no resistor is theoretically required.

But let's leave at that, until you acquire an attitude adjustment - then maybe I might tell you how it actually works, and possibly how to actuality do whatever it is you are actually seeking to achieve.

Well, 1.41 x 240 = 338.4V and not 380V

380V / 3 = 126 LEDS while 338.4V / 3 =~113 LEDs

This will explain (maybe ) the need to remove some of the LEDs to get some brightness. Also the fact that the peak is 338.4V and the rms is 240V will dictate a reduction again but must be carefull because the peak will still be there!

Therefore, if the peak for the LED is 3.5V and the minimum for brightness is 2.6V, then 339/3.5 = ~97, and 97x2.6 = 252 then I would say anything between 96 and 98 LEDS will probably work fine (? experiment?).

100 LEDs can be a starter.

I think my point was missed here - I factored in for SURGES up to about 270VAC, and then converted across to DC peak voltages.... and then I found that my multimeter only read RMS in full wave rectification DC...

Since if I only did linear conversions with no safety factor and to the stated limits of every component by calculating EXACTLY 230V or 240V into DC and then running the string so that each LED got exactly 3.5V DC (full wave not ripple reduced), then rather than risk damage by surges and near enough lightning strikes, to blow the string or damage parts of it.

I opted to add in too many LED's into the initial string (the excess safety factor) and then wound back on that amount, until I felt I had a reasonably good trade off between sufficient illumination AND longevity of the components with some excess for unforeseen circumstances like the silent thunderstorm that floats over ever so slowly on an almost windless night, and then dumps a huge one within a few K of my home...

Like it will not survive EVERYTHING that comes down the line, but I feel that it will do better than average, considering the limitations of LED's to excess voltage and current.

I also expect this ONE light to last me the next 75 years of my life.

Hallo everybody 1) Sorry but what type meter do you use. To my knowledge RMS is the average of the P-P value of a sine wave or alternating-current (A/C). Measuring direct current (D/C) is measuring a voltage, either positive or negative with reference to a 0 volt or other fixed reference point. this voltage is fixed at a certain level so there is no hi-lo change. this means no average or RMS to measure. 2) LED,s are current dependent devices as they are basically diodes.when operated from a constant voltage source (D/C) there should be protection provided by means of a current limiting resistor. The value of this resistor can be calculated as: Rl=Vcc-Vf/If Vf and If you will find in the LED data sheet. Keep in mind the power rating for this resistor. Operating from A/C it must be remembered that LED's must only be operated in the forward direction and therefor the LED circuit must provide reverse voltage protection if the maximum Vr of the LED will be exceeded. This is achieved by connecting a normal silicone diode back to back with the LED. Do not forget the current limiting resistor must also be used for A/C operation. Hope this will help.

I fear for this (the OP) man's life!

He doesn't know his supply voltage, and apparently plans on hooking up LED's to the AC mains!

Oh dear! C,mon you electronics gurus, tell him where he's going wrong.

Sounds like low power LED's too, are these Christmas lights?

I responded to this thread because I could see that the OP was being somewhat "reckless" due to his lack of knowledge.

I'll respond to your "rebuke" with what my reasoning and knowledge led me to provide the responses I did. I don't take any offense to your comments, just like you to know where I'm at.

1/ I agree that merely putting a fuse incircuit may not save the OP from direct harm, but it may save further misadventure should there be an issue with his project. Yes ideally a ELCB protected Epod would be ideal in this situation. Having no idea how old his home is or if there is a (retrofitted)RCD protecting the power point having a fuse on the project is better than relying on the bent nail in the fuse box..

2/ This piece of knowledge came from my Radio Mechanic Apprenticeship days at TAFE some 30 odd years ago. Yes, things, people, attitudes and regulations change, however Electronics Technicians are still not allowed to do electrical work despite being for the most part better qualified.. That said, it was hammered into me that you weren't allowed to have rectification directly on the mains, part of the explanation was the imposition of DC onto the mains supply. There was an issue in the early 70's with an imported brand of TV that did just that, not to mention a live chassis... I had been told it was a regulatory requirement that the rectification circuit had to be isolated from the mains by a transformer.

3/ You are quiet right, "Luminaires" are treated a little bit differently to "Appliances" under the regulations, I have been there and done that with regards to getting devices "approved" for commercial/consumer use. My intention was to alert the OP to the fact that there was an off the shelf opportunity that would perhaps be cheaper(maybe a safer) alternative. Which is why I bought my lights rather than conjure them up myself, out of my parts bins... Ok so i plan on hooking up 70+ leds to house wall socket. This to me infers a decoration rather than interior lighting system. although his 2nd post, Without going totally nutso about the exact frequency (505nm I think) I now have a HEAPS bright light, and being green (not the best like turquoise) that gives me about 2 times as much night visibility as white LEDS, and it uses SFA power in the first place. Makes me wonder what he hopes to achieve.

4/ Yes you're right, however I was hoping to steer the OP into a safer experimenting environment. I have been "plated" several times in my career occasionally with mains, sometimes with HT, the odd LT and RF. It's the hazard of working with energetic electrons, they're always looking for an easier career path.. As a boy apprentice I worked with Thermonic Valves, Often transmitter "tubes" operated with "Plate" (Anode) voltages in the KV range, not a lot of current but geez you knew about it when it bit you. You learned real quick to respect it... These days with OH&S requirements I'm no longer allowed to work on my own if I'm dealing with Mains voltages, at least here at my current job. I suspect the OP is working on his own and may not have any supervision. So I'm trying to to get the OP to consider working at lower voltages where the opportunity for misadventure is less.

So, essentially the OP may be just reinventing the wheel, but should have correct technical advice, or none at all.

I'm all for someone learning and experimenting, however we can no longer be bucket alchemists when we do so. These days safe work practices are just as important as anything else. Unfortunately the interweb has a lot of knowledge on offer but not much wisdom. I have tried in my responses to provide the same guiding technical advice that I would give to any of my Apprentices.

We are basically the same page

Should we send the OP a 'respect' kit?

In series: E/I; In parallel: (70+)(E)/I.

Correction: In series - E/I; In parallel - E/(70+)(I). For the configuration you're describing, you would use the current through the diodes. Residential wall socket voltage in the USA is 120V, unless the electrician installed 240V for certain appliances that need it.

The OP is in Australia.

Domestic mains supply is Single phase MEN 240Vac 50Hz. Although you can guarentee it will vary between 236 to 248 Vac unless your in South Australia where it is 250Vac and can be as high as 260Vac.

That is measured between The Active and Neutral with the Neutral tied to Earth at the switch board.

Three phase is considered to be 415Vac as measured across the phases, whilst it still is 240Vac between the phase and neutral. Again the Earth is tied to Neutral at the switch board.

While it is now mandated in the regulations for houses to be protected by earth leakage breakers this only applies to new houses and any new wiring in existing buildings. There is a push to mandate the retro fitting of ELCB's into all domestic housing. It is mandatory for Industrial/Commercial premises to be so equipped.

I submit that it's a bad idea to stick LEDs into mains power. AC current voltage ratings are average - in the US we have stuff called "120V," but peak voltage is around 170V, so you'd best build for that. It will flicker at 60 Hz, or 120 Hz if you don't rectify.

'Christmas light' quality LEDs degrade quickly, often losing 70% of their brightness over a few hundred hours or worse. This degradation is more rapid over the most-commonly rated 20 mA drive current. Without a datasheet you're guessing and hoping. Buy decent LEDs. You are probably better off investing in a constant-current driver and better LEDs than making a DIY project.

If you insist on DIY, put a proper fuse in it. Here is how to calculate forward voltage:

Vin = Vres + Vled

Vled will be the sum of the forward voltages of the LEDs; around 3.5v each for white/blue/green, 3v for yellow, and 2.2v for red. Note that the forward voltage changes with current, nonlinearly. I strongly suggest the use of LEDs ordered from a reputable manufacturer with a datasheet for all DIY projects.

Vres will be V=IR as normal.

Vin will be the voltage supplied. For longevity I suggest you set this as the maximum voltage input - 340V in the AUS.

Please consult with an electrician before you plug this in though.