How to Drive LEDs - DC To Daylight

In this theory-based episode, we’re going to take a look at what makes LEDs tick and how to drive them properly making sure they operate long-term, avoiding premature failure. We’ll also explore the right and wrong way to drive multiple LEDs in series and parallel, including how to drive a string of LEDs in series using a transistor.
Engage with the element14 presents team on the element14 Community - suggest builds, find project files and behind the scenes video: bit.ly/3tmdewv
Visit the element14 Community for more great activities and free hardware:
Tech spotlights: bit.ly/3qPrDhM
RoadTest and Reviews: bit.ly/3pV5Bux
Project14: bit.ly/31wbnJY
#0:00 Welcome to DC to Daylight
#1:10 LEDs
#5:10 Theory
#10:35 Breadboard
#13:08 Give Your Feedback
#LED, #lightemittingdiode, #resistor, LED current source

[@KevinDC5]

This is gonna sound crazy, but this series(DC2DayL) has helped me more than anything/anyone else in the years of reading, research, conceptualizing and trying to teach myself. I can’t explain why, but your concise and logical approach has me finally “understanding”. I hope to find a more direct communication channel in the links/ forums because I’d really like to hear more on the deeper subjects/circuits! This comment has turned into a rant, but I hope you see it and realize just how transformational they have been for me. RoCkOn!🤘🏼🤘🏼

[@vysakhak191]

I should have seen this video two-three years back. Very useful👍

[@AmRadPodcast]

Thanks! Glad you found it useful.

[@colonelbarker]

I really, really enjoyed this. Really clear and concise, but you didn't feel like you skipped over anything. There's nothing worse than someone cutting from a very simple idea and then throwing in a bunch of components without explaining what they are for. Personally I'd like to see a lot more on transistors, when to use what kind and how. Maybe even making and / or / nor etc gates with them. Keep up the good work!

[@AmRadPodcast]

Great idea Aaron. Transistors are definitely on the list. Maybe a Mega-Transistor episode? -Derek

[@colonelbarker]

@@AmRadPodcast That's a good idea. I tend to stick to TTL logic chips when I need something, when I *know* that a lot of the time a transistor would actually work perfectly fine, depending on the need. So it would be good to have a few examples of nice practical uses where I can break the buggers out, rather than say, stuffing a microcontroller into something.

[@rampr4225]

Loved it 👍

[@tagalogelectronics6615]

nice .....sir..😊

[@jazekerxx7535]

interesting video! i would like to see the more complicated and stable constant current/ led driver circuits..

[@AmRadPodcast]

Stay tuned... I just wrapped up a video using a diode reference / BJT driver for the next project coming out in a couple of weeks.

[@andreaswolfesberger6140]

nice to see the basics - thank you. It would be nice to see something about energy harvesting.....

[@AmRadPodcast]

Hmmm. Interesting. Thank you for the suggestion. You can add more detailed input for video ideas in the e14 community- links are up in the description. I’ll check into it!

[@thegeek3295]

I'd really like to hear / see the difference between smoothing capacitor and decoupling capacitors. They sort of look the same but there must some difference I don't understand.

[@haashiliiran.ahwazcity2506]

very nice and interesting video. sir it would be perfect if you break down a car led headlight bulb driver circuit. good job thank you very much sir

[@yummyklown9226]

I would love to see some cool projects involving the RGB LED's. Maybe controlling them with an Arduino?

[@AmRadPodcast]

It is definitely a possibility as a future project. Thanks for sharing! -Derek

[@Andratos95]

Could I use this to drive a NUBM08 laser diode at 3A? What are the disadvantages and advantages of this setup compared to a linear regulator (such as using an LM317)?

[@vikalpchaudhary]

gr8 video!! could you pls help me understand why would 4 digit 7 segment display run on low brightness when multiplexed via arduino?

[@vinayakonimani7628]

Maybe the LEDs in the setup aren't recieving enough voltage. Or so i guess

[@GarrettH1]

How would I shift an led in the AVR processor?

[@GangstaGranksta]

Excellent video to start messing up with LEDs.

[@luisfernandolebrando7677]

helli, would you mind a O10 LED MODULE feature ..

[@paulsine5190]

Newbie here, still have lots to learn, I really enjoyed this video, thanks so much (I ended up replicating what you did, it was fun) but my multimeter is currently down and I’m curious to know about the volts, the 1.8 volts from the base to the emitter, and the 12 volts -0.7v you supply from the collector-base, would that mean there’s 13 volts being supplied across the emitter collector junction?

[@AmRadPodcast]

If you head over to the demo at the end of the video… add the 1.8v emitter voltage to the collector-emitter voltage at ~3v (on the left-hand meter) you’ll have a total of 4.8v at the collector itself. Take the supply voltage of 12v and subtract that 4.8v and you’ll end up with 7.2v across the string of LEDs. If you divide that 7.2v by 4, then you’ll get 1.8v across each LED. Hope that makes sense! -Derek

[@paulsine5190]

Hey Derek is there a way I can email you a specific question regarding a project of mine?

[@markday3145]

Thanks! It hadn't occurred to me to think about the voltage drops along the base-emitter path. Now I also have a better understanding of using an NPN transistor as a switch (emitter connected to ground, and a current limiting resistor between the input signal and the base). The current limiting resistor would have to drop the input voltage minus the base-emitter drop (around 0.7 volts). Thinking of a BJT as a current amplifier, could you also control collector current by limiting the current into the base? For example, 5 volts connected to a 50K resistor to the base, emitter to ground. Current into the base would be 0.1 mA. If the transistor has a beta of 200, the collector current would be 20mA. I'm guessing this is less stable/reliable. I imagine the beta varies between transistors (even in the same manufacturing batch). And does it vary with other factors (like temperature)?

[@AmRadPodcast]

You can use the transistor's beta to 'ballpark' the current, but as you mention, beta varies wildly in manufacturing.. it's also temperature dependent and changes with the amount of current you put through the transistor. Using a voltage divider, or reference at the base, and an emitter resistor to set the current, things are more stable. In the next video, I use a more robust approach - so come on back! -Derek

[@michaelgregson5165]

Can I drive 16 leds from the same setup just by calculating the resistors? Thanks.

[@rellevartns9415]

You comment at 7:46 about the 3 LED and resistor series possibly having different voltage drops across each LED (assuming manufacturing tolerances) as a reason for utilizing the transistor circuit and then at 11:34 comment that each of the 4 series LEDs in the transistor circuit will have the same voltage drop. In both circuits the current through the respective LED in series will be the same, so what principle makes a consistent voltage drop on the transistor LED series that is not achievable in the non-transistor series? Not following the logic of a more complicated circuit requiring 2 voltage sources (5V and 12V) to achieve the same result of illuminating the LED. The transistor should have no impact on manufacturing tolerances of the LED, should it?

[@AmRadPodcast]

The LED's Vf will never be exactly the same as their neighbors' even if manufactured from the same wafer. Variation comes from non-uniformity during processing, whether it be vapor deposition of metals or non-metals across the wafer's surface, inconsistent plasma (temperature, RF distribution, chemistry) during etch, optical aberration in the lithography processes.. and the list goes on. This is something process engineers, and designers have to live with unfortunately - Hopefully this clarifies my comment at around 7:46. To clarify my comment at 11:34, I'm generalizing and rounding up to 2 volts for each LED in order to explain why *compliance voltage* is important; these are two independent topics. Back to manufacturing tolerance, and why the need for a more complicated circuit: Given that all of the LEDs will have a slightly different Vf, one will run hotter than the rest, also a small change in Vf causes a large change in If. This leads to self-heating, and something called current bunching or hot spots at the PN junction. This changes Vf further, which increases current draw and power dissipation further - this is called thermal runaway, and leads to device failure. Thermal runaway occurs when you are driving a string of LEDs with a constant voltage source, and an LED pulls as much current as it wants (or is available). With a constant current source, this doesn't happen as the transistor circuit dictates how much current is available to the LEDs. Now to be fair, this circuit is a current source and stability is dependent upon supply voltage regulation, so isn't technically a *constant* current source, though sets up the groundwork for the next video in the series, in which I use diodes as a stable reference at the transistor base, providing a constant current, to PWM-drive strings of LEDs. -Derek

[@rellevartns9415]

@@AmRadPodcast : Thank you for the further details and explanation. I appreciate it. Cheers!

[@justinpuzak2558]

how come you dont have to account for the .20mA for the other LEDs, when you had one led in the circuit it had a draw of 20mA, but when you had 4 it only had 20mA draw as well, im guessing that has to do with it being in series and that current stays the same, but still a little confusing to me.

[@AmRadPodcast]

Correct. Forcing 20mA through the “pipeline” of LEDs, they all experience the same amount of current. -Derek

[@chrisallington1638]

Can you talk about commercial LED lighting, such as under-cabinet lighting equipped with a dimmer? What is the dimmer doing to the wave-form, how does it accomplish this, and why doesn't it get warm? (While the power supply and lights are getting noticeably warmer during use?)

[@Otakunopodcast]

Usually those are driven via PWM (pulse width modulation.) In PWM, the LEDs are NOT constantly "on", rather they are turned on an off very rapidly with rapid pulses (think square waves.) To us humans, the LEDs look like they are on solid, because of Persistence of Vision. (If you take a camera with a frame rate that is synchronized with the pulse wave, you can see the LEDs turning on and off.) The amount of time the square wave is "high" vs. "low" controls the brightness (known as "duty cycle.") What the dimmer is doing is it is varying the amount of time that the square wave is "high" (LEDs are on) vs "low" (LEDs are off.)

[@chrisallington1638]

@@Otakunopodcast But, how?

[@Otakunopodcast]

@@chrisallington1638 Nowadays, usually with some sort of microcontroller, like an Arduino. Google "Arduino led pwm" and you will see plenty of examples. You could probably do it with a 555 timer with a variable resistor/capacitor to control the pulse frequency as well. Google "555 timer pwm" and you should see some examples of that.

[@AmRadPodcast]

@@chrisallington1638 come back for the next episode. Spoiler alert 🚨 I do exactly this!

[@kamleshchavan7451]

Can you make a video on transmitters and receiver module circuit diagram used for RC car or toys , reverse engineering. Second thing I'm taking the liberty to ask your help, I'm a beginner in electronics, having basic knowledge, suppose want to design own transmitters and receiver module circuit, for RC toys, what are the topics I should study📚✏

[@AmRadPodcast]

That's a complicated task... However, I'm working on a prototype RF oscillator that an RC control circuit could use as a building block. Keep coming back!

[@DennisKenneybees]

With the prospect of an SHTF given our present and deteriorating economic outlook causing power failures;; I'm wondering how possible it would be to extract enough atmospheric electricity to light enough flashlight LED'S for home elimination and recharge a bank of batteries that would power other household appliances. Or would it be more practicable, all be it more expensive to just use solar panels,

[@objection_your_honor]

This is why they have 2 transistor current mirror circuits.

[@AmRadPodcast]

Yes, you can use a more stable reference at the base, you could also use a two transistor current mirror, or even a cascode mirror. It all depends on what kind of standard you’re after. -Derek

[@objection_your_honor]

@@AmRadPodcast 2 transistors and 2 resistors, and you'll have constant current no matter the voltage (to the rating of the transistor). I think 2N3904 can go up to 40V. A single transistor works if the source voltage doesn't change. Like in a USB power.

[@tyrewald9083]

[@AmRadPodcast]

🤗 -Derek

[@leestewart7558]

Dark corners? Hummmm. 😉😎

[@AmRadPodcast]

There are so many…

[@JanCumps]

So you change your name from currectsource, then talk about currentsource? ;)

[@chrisw1462]

You really don't 'want to be' _several volts_ above that required by the LED string. Anything at or above that value is just fine. If all you have is 12 volts, or that voltage is already in your design, it would be a waste to buy or add another supply voltage just for the LEDS, but a good 9 volt supply would have worked just as well.

[@AmRadPodcast]

True, you can be *at* the required voltage.. I don't mean to imply that you should have a separate supply voltage for this purpose; I hope it wasn't taken that way. But as a general rule, I personally like to stay a couple of volts above the requirement for additional headroom / voltage sag, or event when plugged into a non-ideal power source. -Derek