If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Display Modes |
#11
|
|||
|
|||
green led
T. Ment wrote:
On Wed, 26 Jun 2019 17:44:20 -0400, Paul wrote: Obiwan, use the maths. Plug in the 2.2V instead of the 1.6V in my example. The resistor needs to be series-connected to limit the current flow and reduce LED intensity. I get the math. But the terminology confuses me. Like forward voltage drop. What about maximum input voltage. There must be a limit before you burn it up. Not with the limiting resistor. Reverse PIV is another matter (can go avalanche at high enough potential). But I already mentioned these LEDs are "perfect" hobby LEDs, because the PIV set by the designer is... 5V. That means if the LED is reversed, the reverse potential is not a challenge to the junction. In the forward direction, it just conducts and tries to draw as much current as it is allowed. Since we did our maths, it's "on a leash now" and has to obey us. Even if I used 100V instead of 5V, and as long as I don't reverse it by accident, it's all good. At 100V, I do the maths, and the resistor is probably 20x higher resistance. Since the motherboard uses 5V (clever choice), it doesn't matter whether it's forward or reverse. If it doesn't light, reverse the leads. However, if you powered the front panel with 12V, then there *could* be a reverse PIV issue and damage to the LED. If you work at voltages above 5V with dime store LEDs like those, take some care. Take even more care with the powerful LEDs that cost $20 a pop, as some of those have no reverse protection. A few have taken to placing an "antistatic" device inside the LED, and I suspect that covers the reverse PIV issue. But never assume anything without a spec sheet, which has notes about "issues" with usage. For example, some powerful LEDs have limited soldering temperatures, and the requirements are tough enough to meet, you want the LEDs "mounted" for you on a separate substrate, and then you solder to the substrate and not the LED. This is why some of the SMT LEDs are such a pain in the ass. They're not meant for humans to work with. Only machines like them. The ones with the little legs you bought, are perfect for hobbyists. Paul |
#12
|
|||
|
|||
green led
T. Ment wrote:
On Wed, 26 Jun 2019 21:58:03 +0000, T. Ment wrote: I get the math. But the terminology confuses me. No doubt the circuit is simple to you. But my electronics knowledge is very sparse. I know a resistor "resists" but what does it reduce? Current, voltage, or both? Where is the voltage dropping? At the resistor? At the LED? Or both? We use Ohms law for resistors and Kirchhoffs circuit laws (KCL, KVL) to write equations and solve for unknowns. https://isaacphysics.org/concepts/cp_kirchhoffs_laws Yes, a resistor resists. V --- = I R The larger R is, the smaller the current flow would be. This even works to our advantage with the LED, because the LED is "very greedy". The diode equation shows "how greedy". The diode equation is half way down this page. I didn't show you this, on purpose, because it has nothing to do with getting our computer LED to work. The LED diode behavior is invariant enough, we don't have to solve anything related to this. We just select an "operating point", like 2.2V at 20mA. https://www.allaboutcircuits.com/tex...nd-rectifiers/ Knowing that equation is important if the temperature of the circuit changes. The LED output varies a bit with temperature. The resistor we're using, is not a "perfect" implementation of a current source. It's only an approximation to a current source. The behavior doesn't vary enough, for us to care. That's why, to a first order approximation, I wrote the equation I did. If we spent the evening pondering the actual physics at the atomic level, we'd probably go crazy :-) If you want exact solutions for the circuit behavior, you could try some sort of Newton-Raphson iterative method. Assuming you can make it converge. The LED is a non-linear circuit. And strange things can happen working with those. If you want a component that will blow your mind, that would be the tunnel diode. Too bad it isn't all that useful for hobbyists (I don't think I have one of those in my parts tray, but we did one or two labs with them in school). I seem to remember they're used in Time Domain Reflectometry (TDR), because of the near vertical edges they can produce. In the second article, they make a radio transmitter with nothing but the tunnel diode to make it work. https://en.wikipedia.org/wiki/Tunnel_diode http://www.rfcafe.com/references/pop...lectronics.htm I would need an entire electronics faculty staff to explain it all, and I'm not that guy. There are people in sci.electronics who have written books and/or are professors, if you need that kind of explaining. I just know enough to write a LED equation for you. Paul |
#13
|
|||
|
|||
green led
On Wed, 26 Jun 2019 19:39:53 -0400, Paul wrote:
If you want exact solutions for the circuit behavior, you could try some sort of Newton-Raphson iterative method. I just want to learn enough so I don't burn the house down. I googled and found questions similar to mine: https://electronics.stackexchange.co...-the-same-time |
#14
|
|||
|
|||
green led
On Wed, 26 Jun 2019 19:13:51 -0400, Paul wrote:
The ones with the little legs you bought, are perfect for hobbyists. After some google and Youtube, I think I got it now. The LED is like a diode, current only flows in one direction, from the anode to the cathode. The LED requires a minimum voltage difference between anode and cathode before it lights up. Mine is rated 2.2v. As Paul said, the input voltage can be higher. Voltage won't kill it. Current will. The LED acts like a short circuit once it lights up. It does not limit the current flowing thru it, so you must add a resistor in series to do that. Otherwise, it burns up. The current rating for mine is 20mA, which gives maximum brightness. Lower current should reduce the brightness. One web site suggested as little as 1mA. Higher resistance, lower current. Just beware of the resistor wattage rating, so you don't burn that up either. |
#15
|
|||
|
|||
green led
T. Ment wrote:
My computer case green led failed. I bought some on Ebay, but they are so bright they light up the room and obscure the other color leds. Any suggestions for a cheap source of low intensity leds in a 3mm size? Maybe Mouser, but then $1 worth of parts costs $10 to ship. If I have to spend that much to fix it, I'll just live without a power led. Can you add a resistor or something to change the brightness? You could go through what Paul mentions. Just remember to get some heatshrink tubing to slide onto the wire, solder the resistor to the cable wire, slide up the tubing, and heat to coat the bare wires. Another possibly is scotch tape which is a bit frosted. I don't know the LED you have fit into the holder in the case. You might be able to layer a dozen pieces of scotch tape, trim it off using an exacto blade to match the diameter of the LED's face, and check it still snaps into the case holder. Else, you could buy some self-stick rubber feet that are translucent and stick them over the LED opening in the case. The idea is to attenuate the output via diffusion rather than via voltage change. |
#16
|
|||
|
|||
green led
T. Ment wrote:
On Wed, 26 Jun 2019 19:39:53 -0400, Paul wrote: If you want exact solutions for the circuit behavior, you could try some sort of Newton-Raphson iterative method. I just want to learn enough so I don't burn the house down. I googled and found questions similar to mine: https://electronics.stackexchange.co...-the-same-time First, let me assure you, that if where those two LED legs plug in, if you short those two together, the 180 ohm resistor limits the current. By shorting the legs, the 180 ohm resistor sees a power of V^2/R or 5*5/180 = 0.138W or less than a quarter watt (0.250W). Usually, an engineer selects a resistor suited to being "shorted out by a customer". There are a few things on a computer which are actually dangerous. The SPKR 1x4, where that plugs in, one of the wires is tied to +5V, and if you short the red wire from SPKR to chassis, the wire smokes. This usually happens when someone closes the side on the computer, the red or orange "SPKR" wire happens to get pinched, and the sharp metal casing slices the insulation so that the wire touches ground. That's about the largest exposure I know of. The USB and Firewire ports probably use Polyfuses. Also, the cooling fan +12V bus, is not fuse protected. If you have an issue with a fan, the copper trace in the motherboard will burn out (without a fire issue), and you lose all cooling fans as a result. I suspect some mental genius does this on purpose, and uses an undersized copper track as a zero dollar cost safety precaution. You can, of course, use wire adapters and run a lead from a Molex to the fans, and get power for them that way. It doesn't mean the computer is "finished" - it just means you will spend hours poring over catalogs looking for the right wire set(s) to bring the fan power back :-) This is some kind of punishment, I would guess. I have a few bags of connectors here, so that if this happened to me, I have all the parts I need to fix it :-) That doesn't say I would appreciate having to do that though. I've never lost a fan track, and I've taken my share of chances with goofy wiring jobs too. It's not like I'm a boy scout on the wiring :-) ******* So no, for a fan job, the odds are good there isn't a fire risk. The mobo resistor is on the +5V side, and ground is ground. By adding your additional resistor, the LED is going to get dimmer. Even if you screw up and short the pins on the LED, that should not harm it. Paul |
#17
|
|||
|
|||
green led
VanguardLH wrote:
T. Ment wrote: My computer case green led failed. I bought some on Ebay, but they are so bright they light up the room and obscure the other color leds. Any suggestions for a cheap source of low intensity leds in a 3mm size? Maybe Mouser, but then $1 worth of parts costs $10 to ship. If I have to spend that much to fix it, I'll just live without a power led. Can you add a resistor or something to change the brightness? You could go through what Paul mentions. Just remember to get some heatshrink tubing to slide onto the wire, solder the resistor to the cable wire, slide up the tubing, and heat to coat the bare wires. Another possibly is scotch tape which is a bit frosted. I don't know the LED you have fit into the holder in the case. You might be able to layer a dozen pieces of scotch tape, trim it off using an exacto blade to match the diameter of the LED's face, and check it still snaps into the case holder. Else, you could buy some self-stick rubber feet that are translucent and stick them over the LED opening in the case. The idea is to attenuate the output via diffusion rather than via voltage change. There might be a grommet or LED holder in the case, to hold the LED in place. The Polyolefin tubing is still a good idea, to make a neat job of insulating. There aren't many tape products I trust for any long term stability, and I'd much rather use tubing, and slide it down into place after some soldering is done. The tubing diameter should be about 2x the size of the solder joint (which would be a bit bigger than the wires being soldered). Then, when you apply a bit of hot air from the soldering iron tip (the air stream off the tip), that can be enough to shrink the tubing. Paul |
#18
|
|||
|
|||
green led
On Wed, 26 Jun 2019 23:16:24 -0500, VanguardLH wrote:
Another possibly is scotch tape which is a bit frosted. I don't know the LED you have fit into the holder in the case. You might be able to layer a dozen pieces of scotch tape, trim it off using an exacto blade to match the diameter of the LED's face, and check it still snaps into the case holder. Else, you could buy some self-stick rubber feet that are translucent and stick them over the LED opening in the case. The idea is to attenuate the output via diffusion rather than via voltage change. Like the black sharpie marker idea, that leaves the LED running at higher power, which might mean shorter life. Soldering a resistor is more work, but seems worthwhile to me. |
#19
|
|||
|
|||
green led
On Thu, 27 Jun 2019 00:46:27 -0400, Paul wrote:
The tubing diameter should be about 2x the size of the solder joint (which would be a bit bigger than the wires being soldered). Then, when you apply a bit of hot air from the soldering iron tip (the air stream off the tip), that can be enough to shrink the tubing. I have a low cost soldering station that includes a hot air gun. I'm equipped for the job. As for the resistor equation, one thing I'm still puzzled about. https://electronics.stackexchange.co...-the-same-time Says: So you subtract the LED's forward voltage from the supply voltage, since those are both fixed voltages, and the result will be the amount of voltage that must be dropped across the resistor for the whole to total 9V. So 9V - 2.2V is 6.8V. That is a fixed voltage. The current you want is fixed too - you have decided on 30mA. Why does he want to drop all 6.8v? Why not leave a little extra voltage in the circuit, to be sure enough remains to light the LED? The concept makes more sense to me in terms of limiting current. In my case, say I want only 2 - 4 mA thru the LED. Based on your explanations, I understand how to calculate the resistance for that. But tell me if I'm understanding this: The cathode of the LED is connected to ground, so I will see 0 volts there. The LED needs 2.2v to light up, so the anode will measure 2.2v, which is the sum of LED forward voltage drop and 0v ground That's also the output side of the resistor. The supply side is 5v. In my circuit. those two will never change, no matter what resistor value I choose. The only thing that can change is the current flowing through the LED. So I don't understand why they are talking about resistor voltage drop above. That seems irrelevant. Or maybe I still don't get it. |
#20
|
|||
|
|||
green led
T. Ment wrote:
On Thu, 27 Jun 2019 00:46:27 -0400, Paul wrote: The tubing diameter should be about 2x the size of the solder joint (which would be a bit bigger than the wires being soldered). Then, when you apply a bit of hot air from the soldering iron tip (the air stream off the tip), that can be enough to shrink the tubing. I have a low cost soldering station that includes a hot air gun. I'm equipped for the job. As for the resistor equation, one thing I'm still puzzled about. https://electronics.stackexchange.co...-the-same-time Says: So you subtract the LED's forward voltage from the supply voltage, since those are both fixed voltages, and the result will be the amount of voltage that must be dropped across the resistor for the whole to total 9V. So 9V - 2.2V is 6.8V. That is a fixed voltage. The current you want is fixed too - you have decided on 30mA. Why does he want to drop all 6.8v? Why not leave a little extra voltage in the circuit, to be sure enough remains to light the LED? The concept makes more sense to me in terms of limiting current. In my case, say I want only 2 - 4 mA thru the LED. Based on your explanations, I understand how to calculate the resistance for that. But tell me if I'm understanding this: The cathode of the LED is connected to ground, so I will see 0 volts there. The LED needs 2.2v to light up, so the anode will measure 2.2v, which is the sum of LED forward voltage drop and 0v ground That's also the output side of the resistor. The supply side is 5v. In my circuit. those two will never change, no matter what resistor value I choose. The only thing that can change is the current flowing through the LED. So I don't understand why they are talking about resistor voltage drop above. That seems irrelevant. Or maybe I still don't get it. I wasn't joking about the Newton Raphson. The two components, the LED is "non-linear". The LED "does not follow Ohms law". Attempts to use some sort of static voltage divider behavior, will only lead to hair loss. The best way to think of it, is there is a "tug of war", between the diode and the resistor. The resistor has the upper hand, because you cannot have more than 5V / Rseries amperes of current flowing. The resistor ensures that no matter what voltage appears across the LED, not more than that number of amperes (or milliamperes) can flow. Now, what can the diode do ? This is a strawman diode curve I'll just make up. If we used a "stiff" voltage source, the diode burns brightly with tiny increases in voltage. The slope of this curve is the "dynamic impedance". 2.0V 1mA 2.1V 10mA 2.2V 100mA See how non-linear that is ? Say that if I asked the resistor for the entire 5V/180ohms = 28mA by shorting the end of the resistor, what would the LED do with that much current ? It would hop up to 2.13 volts. OK, now that this has happened, the ends of the resistor have (5 - 2.13) volts across it. The current drops to 2.87V / 180ohms = 16mA. Alright, what does the LED do now ? The voltage drops to maybe 2.11V. Notice how we're asymptotically converging to a "final answer". If we could solve the equations in closed form, I could give a rather ugly formula that would be "accurate" but meaningless to a computer builder. Suffice to say, that the resistor performs a "current limiting" function. The diode presents a "relatively low impedance". The impedance is dynamic (changes with current flow), but is still present, and in my description of the LED behavior, you can see that "stepping along the dynamic diode curve causes hardly any voltage change". Using my Newton-like method, I got 2.13, then 2.11, and after a few iterations, it'll be somewhere between say 2.10 and 2.11. The situation is also relatively insensitive. If the temperature changes, that affects the diode equation, but won't measurably affect your perception of what the LED is doing. The current doesn't change much. The light doesn't change much. Nothing burns. The LED lasts for 25000 hours at 16mA or so. Using a much higher resistance, like putting an 820 ohm total in the circuit, just pushes us further down the diode curve. The resistor is the boss (as it's a relatively high impedance and a "pseudo current source"). The diode cannot do much about the situation - it's been whipped into shape by its resistor neighbor. The diode is very greedy. If you connected it directly to +5V, it's life would be very short, and the bond wire near the anvil would burn out. The diode is "stiff". The resistor is like "floating on air", and the resistor is the boss in the circuit. The diode "can hardly move" in this situation. If we were writing a non-linear solver, it would converge in a half dozen iterations. ******* I've seen this "low impedance" behavior of LEDs in a home project here. I made my own LED bicycle lights, which I run from a 3 Watt bottle generator (spins against the back wheel of the bicycle). I can promise you that LEDs "are very stiff". It's so bad in fact, that a length of lamp cord that leads from the generator on the back wheel, to the front light assembly, results in the front light assembly being noticeably dimmer. That's 16ga wire, and you'd think the resistance is close to zero. But the dynamic impedance of the multiple LEDs says that the wire is actually a substantial source of resistance, and makes my front light weaker. Both the front and back lamp assemblies use the same number of LEDs, but as the LEDs are all in parallel they "fight amongst themselves" to see who gets the current. This is referred to as "current hogging". With LEDs, it's a "dog eat dog" world. Every LED is out for itself. It was only by carefully matching the LEDs (examining the diode curve for every one), that I was able to build a nicely working lamp assembly. So I've seen just how stiff LEDs can be. I never even considered the power cable would be a problem. It even affects how much the lights flicker. (Not shown in the circuit, is the schottky bridge rectifier and the filter caps.) By putting two diodes in series, and matching the Vfb sum of every diode pair, you can prevent any one "column" from taking more than its fair share. The LED summations are matched to within about 20mV or so. It took me two hours of measuring LEDs in a jig, and assigning each LED a serial number, to put this idea together. The LED has to thermally stabilize, to get decent (20mV level) accuracy. You watch the meter for 30 seconds to a minute, until it stops changing. 5VDC +---+-- gen ----- 16ga wire------+---+ 4.9 VDC | | | | LED LED LED LED (only enough LEDs | | | | drawn in diagram LED LED LED LED to show the concept) | | | | +---+------------ 16ga wire -----+---+ The circuit has *no* resistors :-) The generator is known to be a current source, which is why resistors are not needed in this case. I actually did a simulation in LTSpice of my bicycle light! There's a model available for "gen" that someone cooked up, and it captures the generator behavior nicely. As the bicycle goes faster (about 30MPH), the light output actually drops. Going faster doesn't actually burn out the LEDs. Unlike when I used to have a sealed beam incandescent, and one night I went down a steep hill, and that same bottle generator burned out my GE sealed beam. And two minutes later a cop pulled me over for "no light on a bicycle" :-/ Some white LEDs are around 3V, while these happen to be 2.5V each. Two in series takes 5V. Each column carries the same (tiny) current. The generator puts out a precise 500mA (at 6V, but the diodes chop that down to 5V and the generator gets a little warm). Using 3V LEDs would have been a better match for the project. Paul |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
When a Green drive is not Green, especially | Flasherly[_2_] | Homebuilt PC's | 4 | March 4th 15 06:27 PM |
P5Q SE2: Green LED Lit, But Nothing Else? | (PeteCresswell) | Asus Motherboards | 5 | February 15th 09 10:48 PM |
MPG's go green? | J. Clarke | Nvidia Videocards | 3 | January 6th 06 01:50 PM |
MPG's go green? | [email protected] | Nvidia Videocards | 2 | January 5th 06 05:22 AM |
MPG's go green? | Dommy | Nvidia Videocards | 2 | December 30th 05 02:02 AM |