Adding LEDs to projects - calculating series resistors.

[Yorkie]

It’s more than a year ago now since I posted my homebrew ESR meter on another forum, (based on a design by a Canadian radio amateur, VE7IT), since when scores have been built. Over a period of many months, based on feedback, queries and suggestions that I received from constructors who built one, I modified the PCB in various ways to make it easier to construct and use, until it reach the final version as posted on this forum.

One of the queries was from those who wished to add an LED as a warning that the meter was switched on. LEDs are a simple and useful addition to any project but they require a series resistor, and some constructors were unsure how to calculate that. I did modify the PCB layout to add a series resistor on the PCB to make construction easier, but it occurs to me that home-brewers might occasionally like to add an LED to other projects, so some explanation of how to calculate the series resistor might be of interest.

The formula for calculating the series resistor for an LED is: Supply Volts minus LED volts, divided by LED current in mA = series resistor value in Ohms. So for example, say an LED which drops 3.3V, draws a current of 20mA, and is to be run from a 9 Volt supply. 9V – 3.3V = 5.7V to be dropped. Divide this by 20 mA (.02A) and we get 285. Nearest higher preferred value to that would be 330 Ohms, so that’ what we would choose.

As to the Wattage of the resistor, given that it drops 5.7 Volts, and the current is .02A, Watts = Volts x Amps, so 5.7 x .02 = 0.114. Hence, the nearest 330 Ohm resistor above that Wattage would be fine – even an 1/8th Watt one (0.125W), but most normal ones are 0.5W or so anyway.

A quicker rule of thumb for calculating the series resistor is 50 x Voltage to be dropped, so in the above example, 50 x 5.7 V = 285 Ohms. (Use next higher preferred value - 330R).

Or even easier, you can just use one of many online calculators such as:

http://www.hebeiltd.com.cn/?p=zz.led.res...calculator

http://www.quickar.com/noqbestledcalc.htm

I've added a sketch to show how LEDs and series resistors are wired. I should emphasise that the sketch isn't to scale - the LED is montrously large - larger than a PP3 battery in fact!

Hope this info might come in useful to someone sometime.

Happy constructing!

David

[fido]

Thanks for posting this, a useful reference post.

[Retired]

Hi,

Thanks for adding this David it is very useful especially the easy to remember 50x volts to be dropped to give the resistor value.

Kind regards, Col.

[Robjkmannering]

I didn't know about the 50x tip either so thank you.
Worth noting however that not all LEDs drop 3.3v, red ones start at around 1.6v from memory.
Poundshops are great for cheap stuff to rob LEDs from, I picked up a small remote that had 5 blue surface mount ones.
Roberts radios look great with LEDs iluminating the scale, there is plenty of room under the perspex at either end.

Regards,
Rob

[Joe]

The fifty times voltage-drop was new to me as well.

I've just started to use low-current (2mA) ones as it seemed ironic that I was fitting them to things to stop me from leaving them on and thereby draining the battery when in some cases the circuit itself didn't draw 20mA.

Rob's right about the voltages varying and I've always assumed 2.2 which seems close enough for any standard ones. If I've wanted to connect them to 3V I've always got away without a resistor.

These days it's easy to look at the data sheets and see just what they will stand and it's a simple matter to check with an ammeter that you aren't driving them too hard.

It isn't the end of the world if you do pop one. I blew a couple recently when trying to use a capacitor as a mains dropper; I ended up with a good old high-power 10K resistor and the LEDs are still working fine.

- Joe

[ppppenguin]

If the supply voltage is very close to the LED voltage (happens more often if you have a string of LEDs in series) then you end up with a very low resistor value. This gives a variable and rather unpredictable current as the LED voltage drops vary with temperature. As a very rough and ready rule if the drop across the resistor is less than the drop across the LED(s) be careful.

If you connect a LED directly to a voltage supply such as a battery then that's asking for trouble. The current will be utterly unpredictable.

[Joe]

I haven't done much with them for a long time until the last few months. I definitely did connect them to two AAs, when I was young and daft, and don't recall any problems - though that was probably down to luck.

I also put one without a resistor straight across the output of a 3V Walkman power supply I built but that was regulated.

Of course just because you can get away with doing something doesn't mean you should and I'm far more disciplined with the low current ones.

- Joe

EDIT On reflection the AAs were probably Ni-Cads so would only be 2.4V anyway.

[Yorkie]

Thanks for reading my posting and for the comments guys.

Yes, perhaps 3.3V and 20mA were not the best choice of parameters for a worked example - I didn't mean to infer that they were typical. Most ordinary LEDs will be 2 - 2.2V, but ultra bright ones will generally be 3.3V. Obviously it's best if the actual parameters are known for the particular LED, but most of the ones I use are from 'lucky bags' with unknown specs, so I make a few assumptions and if the LED doesn't glow brightly enough, I try a lower value series resistor.

There are now so many varieties of LED - some flash, others will change colour from green to amber to red, depending on voltage, and there are super wide-angled ultra bright pure white ones, which would probably be useful for illuminating dials in place of hard to find bulbs, though they might need a thin coat of amber paint to make them look 'warmer'. I thought I might need to go down that route when I recently restored a Stella 105U as it needed a 19v dial bulb, which I assumed wouldn't be obtainable, but no - they're in plentiful supply!

David

[Mark Hennessy]

The "50 times" comes from 50 being the reciprocal of 0.02. This is a useful numerical result to remember for another reason:

If you're doing power supply work with a 'scope, 50Hz should give 20ms. If you have hum on the output of the power supply, one of the first things to check is the ripple frequency. If, as is most often the case, you are dealing with full-wave rectification, the ripple frequency should be 100Hz (10ms). If you have 50Hz, you have an open-circuit diode in the rectifier...

I'd suggest that "100 times" (10mA) is a better starting point for most modern LEDs. The key words in that sentence, of course, are "starting point". These modern LEDs (blue, white, pure green) with a higher voltage drop are very bright indeed, and if they are being used on a front panel as a directly viewed indicator, 2mA (500 times) might be plenty. Older LEDs tend to need a bit more, but experiment (and enjoy!).

The higher the voltage, the less significant the forward drop of the LED is, so it can be ignored at say ~12 volts and above, as the 20% spacing of typical E12 resistors gives a similar margin. But care is needed at lower voltages.

Oh, and LEDs don't like reverse-voltage. They "zener" at relatively low and poorly defined voltages. If using LEDs to replace dial lamps, it might be as well to use a pair in inverse-parallel so that each one never sees more than the VF of the other.

Like many engineers, I enjoy LEDs just a little more than I perhaps should :-)

Cheers,

Mark

[DODGERN]

Very useful tip David. One to make a note of.

I bought a bag of bright white LED's and fitted them in parallel on my staircase. I had to match them in pairs as the brightness level varied. The total outlay was around £35.
They are running from a Ross regulated 1A power supply on the 3V setting.
They have been running for a number of years now.

Roger.