EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: RJSV on January 22, 2021, 09:15:59 am
-
Hi there! I was trying to re-use my battery pack, for other sorts of LEDs. So the 3-cell pack turned out to have about a 16 ohm resistor, and the output probably an NPN transistor, like a 2n2222.
I put a one ohm resistor in the (output) line, and each of the four 'CORN COB' LED strips also connects to the positive (4 volts). Confusing, at first, as the current reading, made by reading the voltage across the temporary one-ohm resistor, was staying 'constant', 60 ma regardless how many LEDs were in the parallel circuit. This is where a Large signal / Small signal view point helps:
In the gross, or Large signal view, it is a 'constant current' source circuit. So within reason, any number of LEDs can be connected, although all will dim, as each gets a smaller share, of the 60 mA current total.
I calculated the internal resistor at 16 ohms, using first subtracting Vsat for any transistor, 200 mV.
Then used the measured output voltage, about 1.19 volts.
I say 'about' because that output goes up, slightly, when all four LEDs are connected, from 1.19v to 1.25v.
That is part of what is meant by small signal: While individual current has quadrupled, device voltage only changed by 90 mV (as measured directly at LED).
-
Leds are rather amazing devices. They have so much better light efficiency, compared to the old fashioned filament bulbs, which didn't last too long on batteries.
It is also surprising how small a current, can still light up a modern Led, even if it is faint, at very low currents. If I remember correctly, less than a microamp, can still light up a modern Led.
So Leds work great on batteries.
-
So sensitive that even tiny leakage current in your house wiring will keep them glowing.
-
So sensitive that even tiny leakage current in your house wiring will keep them glowing.
Yes, I think for that very reason. Better designed/quality ones, have resistor(s) in the right place, to stop that from happening. But poorer quality, cheap, not so well designed ones don't. Hence small leakage currents in light switches (or however the leakage currents are getting through), can be enough to cause slight glow, in the Leds. Which can annoy/distract some people, or make them think something is wrong with their electrics.
I think BigClive's video channel sometimes talks about it and demonstrates it.
-
You are right, about LED efficiency my friend: I couldn't get the brightness down enough for normal camera settings, even using a 1 K ohm resistor. That's a LED current below 400 micro Amps.
This 'CORN COBB' style LED is a bar with 14 dots, in a row, and I suspect each 'dot' is a UV or IR led underneath the little dot of chemical 'coating'. The color is warm, while the bar shape is unusual.
At 2.7 Volts, is the hottest I dared to run them (but better to not use 'VOLTS'). That's at 60 mA through one corn cobb LED. Starts to be dim at 2.6 V, but again, better to use current, 15 mA, was typical, for this Drug Store' yard product (has four giant lights).
-
Leds are rather amazing devices. They have so much better light efficiency, compared to the old fashioned filament bulbs, which didn't last too long on batteries.
It is also surprising how small a current, can still light up a modern Led, even if it is faint, at very low currents. If I remember correctly, less than a microamp, can still light up a modern Led.
So Leds work great on batteries.
Some time ago I pulled a CR2032 coin cell out of a remote control that it was too dead to operate reliably. There was a bag of white LEDs on the table so on a whim I wedged one across the battery and set it there. That was over a month ago and it's still glowing, not particularly bright but bright enough to see it glowing in normal room light. It's really pretty amazing I think, I expected it to last a day or two at most, the battery was pretty low when I started.
-
Some time ago I pulled a CR2032 coin cell out of a remote control that it was too dead to operate reliably. There was a bag of white LEDs on the table so on a whim I wedged one across the battery and set it there. That was over a month ago and it's still glowing, not particularly bright but bright enough to see it glowing in normal room light. It's really pretty amazing I think, I expected it to last a day or two at most, the battery was pretty low when I started.
Despite me knowing/understanding about how little current lights many modern Leds. That is still an amazing story.
A very long time ago, I bought a tiny Led torch, I can't remember the price, but it was something like £0.99 or 3 such torches for £1.99 etc (you get the idea). Not much money, anyway.
At some point, it broke or ran down or I got curious and opened it up.
All it had inside was one or two CR2032 cells (possibly a thinner variety, with lower part numbers), with a through hole white or green Led, with the leads (no PCB or even additional wires/soldering), folded over the battery(s). So when you pressed the thin plastic case, the slightly bent lead, completed the circuit and lit the torch.
Amazingly minimalist.
The outgoing tungsten filament bulbs, usually don't even bother to light up at all, at something like 10% or even 20% of their rated voltage. It (presumably) gets hot inside, but not enough to give out red/white hot light, so just appears 'off'.
-
Tungsten bulbs are particularly poor for battery applications because as the voltage is reduced the efficiency is reduced dramatically. Filament resistance decreases so it drags the voltage down really quickly. LEDs don't behave like that at all, the efficiency rises as the current falls. It's really remarkable how little current is necessary to create a visible glow.
-
Tungsten bulbs are particularly poor for battery applications because as the voltage is reduced the efficiency is reduced dramatically. Filament resistance decreases so it drags the voltage down really quickly. LEDs don't behave like that at all, the efficiency rises as the current falls. It's really remarkable how little current is necessary to create a visible glow.
It is amazing. One would have thought, that at such low current/power levels, our human eyes, wouldn't even be able to detect light, at such low power levels. Yet they can. Even though they don't even have amplifiers in them (that I know of), as such, like electronics devices could/would have. I.e. Photodiode plus op-amp or similar, to get the intensity level.
The old fashioned (nowadays), Neon bulbs (small types), also seem to glow, with rather low current levels (compared to filament bulbs, but still tons more than Leds need). But they need an initial highish (e.g. 90 volts) voltage, to initially strike, and still need a relatively high voltage to stay on (compared to Leds).
-
The amount of current needed to get visible light from an LED is so low that it often catches people off-guard: https://www.eevblog.com/forum/chat/led_s-light-when-soldering/ (https://www.eevblog.com/forum/chat/led_s-light-when-soldering/)
-
You are right, about LED efficiency my friend: I couldn't get the brightness down enough for normal camera settings, even using a 1 K ohm resistor. That's a LED current below 400 micro Amps.
This 'CORN COBB' style LED is a bar with 14 dots, in a row, and I suspect each 'dot' is a UV or IR led underneath the little dot of chemical 'coating'. The color is warm, while the bar shape is unusual.
At 2.7 Volts, is the hottest I dared to run them (but better to not use 'VOLTS'). That's at 60 mA through one corn cobb LED. Starts to be dim at 2.6 V, but again, better to use current, 15 mA, was typical, for this Drug Store' yard product (has four giant lights).
I have sometimes missed the old tungsten filament bulbs/lights (manufacturing/new-sales mostly/fully banned in the UK/EU). So, I'm pleased there is such a thing as Led filament lights, like you are describing and showing in the picture (if I understand things correctly).
They give me options, to go back to something, partly like the old filament bulbs, and yet still have the benefits of modern Led lights.
BigClives Youtube channel, also likes them.
Example:
https://www.youtube.com/watch?v=GD_BLc3R4N8 (https://www.youtube.com/watch?v=GD_BLc3R4N8)
https://www.youtube.com/watch?v=NffhdAz9pc4 (https://www.youtube.com/watch?v=NffhdAz9pc4)
-
The amount of current needed to get visible light from an LED is so low that it often catches people off-guard: https://www.eevblog.com/forum/chat/led_s-light-when-soldering/ (https://www.eevblog.com/forum/chat/led_s-light-when-soldering/)
That is amazing!
A recent thread here, had a similar problem, where their Led Driver Fets, 25 microamp (approx) leakage current was enough, to flicker/light the Leds. Which worried them.
I like the fact that Leds do that. Because it means that if you take a White Led torch (probably has to be the old/cheap, directly connected to the battery via a resistor ones, rather than ones with a switching circuit). Although they get dimmer and dimmer, when the battery runs out, they still give a reasonable/usable amount of light. If you get caught out, in an emergency or something.
Whereas the old filament bulb torches, really would eventually turn off completely, when the batteries are exhausted.
I.e. If a filament bulb torch batteries, last 3 hours, it probably will give off little or no light, after 10 hours.
But an Led Torch, which is suppose to last (brightly) for 3 hours, may still have usable brightness, even 30 hours later (unless it is the type with switchers, e.g. 1 AA powers it, in which case, I'm not sure, how long those can carry on for, it could be a very long time (dimly), or they could need too much minimum current for the switcher, and go off completely).
-
I have a cheap LED shop light style fixture above my workbench and I noticed a while back that when it's plugged in but switched off the tubes glow visibly, I assume just from capacitive coupling. Not very bright, but it's visible in a darkened room.
-
I have sometimes missed the old tungsten filament bulbs/lights (manufacturing/new-sales mostly/fully banned in the UK/EU). So, I'm pleased there is such a thing as Led filament lights, like you are describing and showing in the picture (if I understand things correctly).
Bulb bans create interesting problems. At work we have a lot of instruments that use various filament light bulbs, some designed decades ago. Many of these bulbs can not be purchased anymore. This leads to interesting challenges in retrofitting equipment. LED's and filament bulbs are different enough that it can lead to scrapping some very expensive instruments, odd retrofits and other struggles. Worse the bulbs we can purchase have really fallen in quality control.
-
Hi there! I was trying to re-use my battery pack, for other sorts of LEDs. So the 3-cell pack turned out to have about a 16 ohm resistor, and the output probably an NPN transistor, like a 2n2222.
I put a one ohm resistor in the (output) line, and each of the four 'CORN COB' LED strips also connects to the positive (4 volts).
Just a little FYI, I think you're confounding "corn cob" LED bulbs and COB LEDs. "Corn cob" bulbs refers to the LED replacements that resemble corn on the cob because of having rows of individual LEDs (which of course are yellow when off). COB, on the other hand, means "chip on board", where the LED dice are mounted directly on an aluminum PCB, then covered in a continuous yellow phosphor gel. Your pictures are awful, but looks like it's COB LEDs.
-
I have sometimes missed the old tungsten filament bulbs/lights (manufacturing/new-sales mostly/fully banned in the UK/EU). So, I'm pleased there is such a thing as Led filament lights, like you are describing and showing in the picture (if I understand things correctly).
Bulb bans create interesting problems. At work we have a lot of instruments that use various filament light bulbs, some designed decades ago. Many of these bulbs can not be purchased anymore. This leads to interesting challenges in retrofitting equipment. LED's and filament bulbs are different enough that it can lead to scrapping some very expensive instruments, odd retrofits and other struggles. Worse the bulbs we can purchase have really fallen in quality control.
But the bulb bans only restrict ordinary "household"-style lighting lightbulbs. They never banned incandescent bulbs for any other purposes, including decoration, indication, or specialty uses. Truth is, incandescent indicator lamps were obsolete decades ago, long before any incandescent bulb bans were enacted, and that's why they've become scarce. Industrial indicators and auto headlamps were probably the last holdouts.
And since all the quality bulb manufacturers stopped making indicator bulbs ages ago (since their markets had already shrunk to next to nothing, thanks to indicator LEDs), all that's left are the companies in developing nations who make them with less stringent quality control.
-
Bulb bans create interesting problems. At work we have a lot of instruments that use various filament light bulbs, some designed decades ago. Many of these bulbs can not be purchased anymore. This leads to interesting challenges in retrofitting equipment. LED's and filament bulbs are different enough that it can lead to scrapping some very expensive instruments, odd retrofits and other struggles. Worse the bulbs we can purchase have really fallen in quality control.
I don't like bans, I would much rather tax less efficient products and use that to subsidize more efficient options when it is needed to get the ball rolling. Now LED bulbs are dirt cheap so they can compete directly on cost. I have definitely noticed that the quality of incandescent lamps that are still available has gone way down. I got a few 3-way bulbs for our cabin, the only place I still use incandescent lamps because it maintains the whole 1970s time capsule vibe and the heat is electric resistance anyway. Out of 3 GE branded bulbs I bought brand new, 2 of them had issues on the first evening, bad connections in the base it seems.
I love LEDs for general illumination and was an early adopter, I upgraded most of the lights in my house back when I was paying $30-$50 *each* for ordinary warm white bulbs, but they don't work well for every application.
-
Bulb bans create interesting problems. At work we have a lot of instruments that use various filament light bulbs, some designed decades ago. Many of these bulbs can not be purchased anymore. This leads to interesting challenges in retrofitting equipment. LED's and filament bulbs are different enough that it can lead to scrapping some very expensive instruments, odd retrofits and other struggles. Worse the bulbs we can purchase have really fallen in quality control.
I've noticed that, from opening up various equipment, over the years. Some things actually use the characteristics of a bulb, to help the device in some way.
Some (details from my memory, from long ago, so could wrong or inaccurate) Wien bridge oscillators, use a bulb, to 'balance the loop gain', because of its positive resistance coefficient. Essentially, the bulb is configured, to hold the gain of the loop, just above 1, to keep it 'just oscillating', and hence give the best/lowest distortion Sine wave.
Some valve/tube circuits, especially the ones which had all/many of the heater filaments in series, connected to the mains (to save on bulky/expensive 6.3 V AC transformers), could of included one or more front panel filament bulbs, such as 24 V ones.
There use to be self flashing bulbs available (possibly still available, but probably getting rather rare now), which had a tiny built in bi-metalic strip, and it would self-flash. Useful for flashing torches (especially red), Christmas tree lights (there use to be big chains of say, 50 bulbs, and only one needed to be a special flashing one, too make them all flash, neat/cool) and for making workmans street hazard warning lamps (when road works are being done). Typically Amber in the UK, possibly Green in the US (I'm not sure).
I think some things even use the heat from a bulb to do things, but I'm not sure. Bulbs also make useful current/power/surge limiters.
You use to be able to get, fuses, which were actually bulbs as well, which could light things up. I'm not 100% sure, if the fuse case, was just a convenient way of acting as a bulb holder, or its fusing (current limiting) capabilities, were also needed.
I wonder if you could substitute a suitably valued PTC thermister, with a current limited (resistor in series with) Led, in parallel, to replace a bulb.
-
I think some things even use the heat from a bulb to do things, but I'm not sure.
One of the more recognizable examples is the classic Lava Lamp. In addition to several of those I have a 1980s party lamp that has a little drum balanced on a pin above a light bulb with fan blades stamped into the top of the drum. The sides of the drum are transparent plastic printed with colored patterns and it spins from the heat of the bulb.
The Easy Bake oven is another classic example, these used an incandescent lamp purely as a heater, the light was a byproduct.
-
One of the more recognizable examples is the classic Lava Lamp. In addition to several of those I have a 1980s party lamp that has a little drum balanced on a pin above a light bulb with fan blades stamped into the top of the drop. The sides of the drum are transparent plastic printed with colored patterns and it spins from the heat of the bulb.
The Easy Bake oven is another classic example, these used an incandescent lamp purely as a heater, the light was a byproduct.
Of course, I'd forgotten about Lava Lamps, which can be really fascinating. It's something like a 25 watt pygmy bulb, melting a wax/water mixture, or thereabouts.
Those Easy Bake ovens, I'm not sure if I knew they used incandescent bulbs, or maybe I've forgotten, over the years.
In ancient/dusty electronics books (possibly from the 1930's/40's/50's etc), I think I remember they would have D.I.Y. details on making wooden boxed, artificial Hen Egg incubators, which used 60 or 100 watt bulbs (I can't remember, precisely), to act as a heater.
Now you mention it, I think there use to be (memory rather vague, as it was a VERY long time ago), household sized bulbs (low wattage, inside the electric fire, maybe 20 watts, very approx), which were both fire coloured (Dark Orange), and had some kind of thin, shaped, metal disk, suspended by tiny wires, inside the top of the bulb. Which, when you turned on the electric fire, would move around (it was circular, maybe it was just above the bulb, rather than inside it).
It would then spin around with the heat flow (air), making fire like patterns, on the front of the electric fire.
Probably from around the 1960's, I would guess.
I guess this is the sort of thing I'm talking about:
https://www.amazon.co.uk/Fire-Effect-Spinner-Electric-Fires/dp/B00ITR4C1E (https://www.amazon.co.uk/Fire-Effect-Spinner-Electric-Fires/dp/B00ITR4C1E)
-
In addition to several of those I have a 1980s party lamp that has a little drum balanced on a pin above a light bulb with fan blades stamped into the top of the drop. The sides of the drum are transparent plastic printed with colored patterns and it spins from the heat of the bulb
Crikey that's brought back memories,i saved up my pocket money for one.The colours were in a spiral down the tube and the sides of the lamp had circles in to break up the light.
making fire like patterns, on the front of the electric fire.
Probably from around the 1960's, I would guess.
not that old. New electric fires use the same technology to give the flame flicker effect over the coals, the lamps (fireglow) are readily available
-
NEXT, I plan to measure two LEDs (parallel and on the 16 ohm resistor of the battery pack.
Obviously share the same voltage, but curious to see how close to 50/50 split of current.
I can tweak a couple of one-ohm resistors, by way of parallel 100 ohm 'trimmers'. (So I can read low currents as identical as possible).
-
Diagram shows how I plan to create two sets of resistors, for two legs of current measurement. Two resistors connect in parallel, total 1/2 ohm, and an optional 100 ohm 'trim', as needed.
I did not like placing much resistor, between each of the two parallel LEDs. I think slapping on a resistor starts turning a 'current' driven device into 'voltage' actuated...
So each LED gets approx. 1/2 ohm, on the positive leg, and the 16 ohms internal to the 3-cell battery pack / blinker.
The diagram shows each LED connects, via blue colored line, to the battery pack (acts to sink the 60 mA total). One or the other side (legs) gets the 'trim' resistor, as needed.
-
I had always wondered about LEDs in direct parallel, so the experimental measuring, on my drug-store bought lawn lights, is curious results.
For measuring one LED leg against other LED leg, I need something like two closely MATCHED Galvanometers, however that's not enough accuracy, but the idea is that each separate leg, A and B would not have any components affecting the DC reading ( and no voltage drop complexity introduced.
I do know that my normal test leads have shown up to two ohms added in to a circuit.
I am even thinking, of how about two digital volt meters that are closely matched.
I suspect, current split between two LEDs of same type, but perhaps from different factory runs, WILL show some decent amount of difference. And, with any differences in LED current, between the two legs A, and B, could end up fairly large (obviously not a 50/50 split.
Consider: suppose LED A voltage is supposed to be higher, than LED B due to manufacturing differences.
Then, the two LEDs will 'conflict', with slight shifts (because voltages have to match, as they are parallel).
So, one of those two LEDs will experience a voltage shift, and my point is, that being overly sensitive to operating voltage, the current shifts by more than just a little, or linear to the voltage shift amount.
Probably a bit muddied, but sometimes just phrasing the questions properly can be elusive !
Just don't use 'voltage',... I got that one
-
Lots of multi-chip LED modules have chips in parallel. Two things are critical when doing this though, matching characteristics, and all chips being at the same temperature, ie bonded to the same substrate. Semiconductors vary slightly from batch to batch and the forward drop of any diode varies with temperature. If you just put two random LEDs in parallel whichever one has a slightly lower forward drop will hog most of the current.