14Vf LED question / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

14Vf LED question

previous 1 2 3 4 next

Move Thread

LAN_403

Bored Chemist

Wed Sept 23 2015, 12:47PM

Registered Member #193 Joined: Fri Feb 17 2006, 07:04AM
Location: sheffield
Posts: 1022

OK a couple of points, nice try with the electrolysis experiment there, but completely wrong.
The tables of electrode potentials will tell you that the voltage from a cell where the electrodes are oxygen and hydrogen saturated platinum is 1.7 volts.
But, as you might imagine, that voltage depends on the concentrations of the components (the pressures of the gases are proportional to their concentrations).
So, while it tales 1.7 volts to produce oxygen and hydrogen at 1 atmosphere, even a very small voltage will produce some- and if it can diffuse away into the bulk solution, a current will flow.
So lets be absolutely clear about this.
This statement " Below roughly 1.7v, you cannot split a water molecule"
is not true.

Now, on a related not, back at the topic.
I got bored and measured the voltage / current curve for an LED that happened to be lying around.

It's tricky to measure currents below a few tens of nA and since the voltmeter has about a gigohm input resistance the measurements would be a bit flaky below that. Similarly, if you run more than a few tens of mA through a LED you risk burning it out so that's the range I covered.
Can someone please tell me where the "magic" threshold is?
V micro A
1.40082 0.013
1.4508 0.037
1.49997 0.118
1.50213 0.12
1.55067 0.399
1.55083 0.44
1.60168 1.37
1.65 169 5.54
1.70085 24.47
1.75028 109.79
1.79941 441. 73
1.84977 1413.7
1.90405 3439.77

Bored Chemist

Wed Sept 23 2015, 12:53PM

Registered Member #193 Joined: Fri Feb 17 2006, 07:04AM
Location: sheffield
Posts: 1022

This picture of the plot of log(I) vs V makes my point I think.
The "threshold" voltage is a myth.

Dr. Slack

Wed Sept 23 2015, 01:17PM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

lovely linear plot, but my science teacher always told me to label the axes, and state what was being measured, and under what conditions.

Ash Small

Wed Sept 23 2015, 01:17PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Bored Chemist wrote ...

This picture of the plot of log(I) vs V makes my point I think.
The "threshold" voltage is a myth.

I modelled this once in SPICE, the only time I ever used it, trying to understand how semiconductors actually work. I'll see if I can find a link to the thread, it was years ago.

Bored Chemist

Wed Sept 23 2015, 01:28PM

Registered Member #193 Joined: Fri Feb 17 2006, 07:04AM
Location: sheffield
Posts: 1022

Dr. Slack wrote ...

lovely linear plot, but my science teacher always told me to label the axes, and state what was being measured, and under what conditions.

My science teacher would have told me to look in the posting immediately before..
The point is that it's a plot of log I vs V and it does not show a threshold.
Incidentally, it should be a straight line

and it is.
So, why is there this bizarre myth that LEDs have a "threshold"?

Bored Chemist

Wed Sept 23 2015, 01:39PM

Registered Member #193 Joined: Fri Feb 17 2006, 07:04AM
Location: sheffield
Posts: 1022

One of the good folks here (and I) did some brief research on low currents in LEDs

and it seems that this assertion
"Below a specific voltage, depending on some horrible physics involving band-gaps, you won't excite carriers and produce photons. Below a few volts depending on the colour, a LED is a DED, a dark-emitting-device. Both of these devices may conduct some leakage current by other mechanisms below the threshhold voltage, but neither could be said to be 'working' in this state."
is also questionable.

Even very low currents (and fairly low voltages) give visible amounts of light.

Dr. Slack

Wed Sept 23 2015, 03:14PM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

You were saying?

Measurements from 100nA to 10mA for a range of diode devices. Click on the image to see a full-size readable one.

wrote ...

So lets be absolutely clear about this.
This statement " Below roughly 1.7v, you cannot split a water molecule"
is not true.

Tell you what, you make an electrolyser, with inert electrodes, that produces a sustainable stream of oxygen and hydrogen gas from water with less than 1.7v across it, and I'll invest in your resulting perpetual motion machine.

Ash Small

Wed Sept 23 2015, 03:55PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

I found a link:-

This is a plot showing how using inductors instead of resistors affects current:-

Note: using resistors current reaches 23mA max, using inductors current reaches 25-26 mA max. Ten percent more current.

There are various other plots in the thread showing what happens when the mains switch is switched mid-cycle, etc.

Bored Chemist

Wed Sept 23 2015, 04:12PM

Registered Member #193 Joined: Fri Feb 17 2006, 07:04AM
Location: sheffield
Posts: 1022

Pretty nearly snap!
The bottom right hand corner gives the log of the current vs the voltage much the same as I did, (albeit with the axes transposed) and, apart from the Zener (which is a weird animal anyway) they all give more or less linear plots with no obvious kinks in (though they "round off" towards the top end- I'm not sure if that's the effect of heating or lead resistance or what.
I replotted the data for the green one (the one I had was yellow so that's the most similar) and the data I got on the same scales (Log of i in micro amps vs voltage).
Yours are the data at slightly higher voltage.
There are two obvious differences- the rounding at the top which as I said, I suspect is thermal.
There's also the flattening off at the bottom where the points for about 1.7 and 1.9 volts have rather higher currents than you would expect by extrapolating the linear part.
The equivalent resistance (from R=V/I) of the LED under those conditions are about 6 M Ohm and 17M Ohm
Did you measure those voltages by putting an ordinary voltmeter across them?
If so, the loading by the relatively low resistance of the meter (typically 1M Ohm or 10M Ohm) will have upset the results.
Are you in a position to use a higher resistance meter or to rig up a buffer amp? That would be helpful in establishing why your measurements don't tally with the mathematical model.
It's also interesting that the "typical" voltage for a green LED is about 2.0 to 2.1 V (data on-line vary) and that's pretty much in the linear Log I vs V part of the curve. the area of the graph where people use LEDs is roughly where they don't have a of threshold.

And finally, you seem to have missed the point.
nobody said that you could generate commercial supplies of hydrogen that way.
That particular logical fallacy is called a straw man.
What I said was that a current would flow at a much lower voltage than 1.7.
Would you like me to set up the experiment and plot out the results for you?

Dr. Slack

Wed Sept 23 2015, 04:48PM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

Bored Chemist wrote ...

What I said was that a current would flow at a much lower voltage than 1.7.
Would you like me to set up the experiment and plot out the results for you?

Yes, check my long post and that's what I said too, by mechanisms other than the 'wanted' behaviour of splitting water molecules.

By the same token, dragging the subject back to the topic of LEDs, light photons need a certain minimum energy, and in a LED, that needs to be provided by a single electron. The excitation voltage must be above that energy threshold, or we're back in perpetual motion territory again. However, below that voltage there is still current flowing by other leakage unwanted mechanisms, just no light production.

Loading the LEDs with a 10M meter as I measure them tends to reduce the measured voltage at low currents, hence suggest that the string current is flowing at a lower device voltage. Measurement with a high impedance meter would sharpen the knee in the voltage/current curve. Thus my mid-impedance measurement conditions favour your 'no threshold' position.

I think if we haul the argument all the way back to the OP, he was talking about what conditions would give him visible light, not whether his LED had sub-threshold leakage currents. I don't think he'd be looking for remarks on his costume like 'Dude, I really like the leakage currents in your non light emitting things!'

Is it really the case that the only difference between our arguments is that you say there's no threshold for current flow, which I agree with, there isn't, due to leakage currents, where I say there *is* a threshold for useful function, light output or molecule splitting, which there is as it depends on the finite energy needed to make one quantum of stuff do its useful stuff?

previous 1 2 3 4 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob