How test LED?

[Alex Coleman]

I have a light consisting of one super-bright white LED which plugs into
the USB port of my PC. It was flickering and now does not illuminate.

The supply leads to the LED are delivering 5v from the USB port but my
meter wouldn't show the sort of intermittency which might cause flicker.

On dismantling the light, I see the leads of the LED are the same length
and are unmarked.

(1) In general, how can I test such an LED using my multimeter and/or
battery? (If it helps, my multimeter has a socket marked E.B.C.E which
I think can somehow test transistors.)

(2) If I apply something like 5 volts or perhaps 8 volts to just the
LED in reverse to its normal polarity then will I cause the LED damage?

 

[[email protected]]

I have a light consisting of one super-bright white LED which plugs into
the USB port of my PC. It was flickering and now does not illuminate.

The supply leads to the LED are delivering 5v from the USB port but my
meter wouldn't show the sort of intermittency which might cause flicker.

On dismantling the light, I see the leads of the LED are the same length
and are unmarked.

(1) In general, how can I test such an LED using my multimeter and/or
battery? (If it helps, my multimeter has a socket marked E.B.C.E which
I think can somehow test transistors.)

(2) If I apply something like 5 volts or perhaps 8 volts to just the
LED in reverse to its normal polarity then will I cause the LED damage?

Check it with a 1000 ohm in series and a 9v battery

 

[Lostgallifreyan]

I have a light consisting of one super-bright white LED which plugs into
the USB port of my PC. It was flickering and now does not illuminate.

The supply leads to the LED are delivering 5v from the USB port but my
meter wouldn't show the sort of intermittency which might cause flicker.

On dismantling the light, I see the leads of the LED are the same length
and are unmarked.

(1) In general, how can I test such an LED using my multimeter and/or
battery? (If it helps, my multimeter has a socket marked E.B.C.E which
I think can somehow test transistors.)

(2) If I apply something like 5 volts or perhaps 8 volts to just the
LED in reverse to its normal polarity then will I cause the LED damage?

Never connect it directly to a battery, even in reverse, Bad habit. An LED
depends on control of current, not of voltage.

The formula for finding the right resistor for an LED is (Vs - Vf) / If.

Vs = Supply voltage in volts.
Vf = LED's rated forward voltage in volts.
If = LED's rated forward current in AMPS. Take care, LED's are rated in mA

Plug those values into the formula to get the series resistance in ohms.

If you need a controlled current and a variable input supply range, look at
the datasheet for the LM317 to see how to use it as a current regulator.
It's the simplest way, but take care to calculate power as well, or a large
supply voltage will burn the regulator.

More info than you want, perhaps, but hopefully good info you can use.

 

[Dorian McIntire]

I have a light consisting of one super-bright white LED which plugs into
the USB port of my PC. It was flickering and now does not illuminate.

The supply leads to the LED are delivering 5v from the USB port but my
meter wouldn't show the sort of intermittency which might cause flicker.

On dismantling the light, I see the leads of the LED are the same length
and are unmarked.

(1) In general, how can I test such an LED using my multimeter and/or
battery? (If it helps, my multimeter has a socket marked E.B.C.E which
I think can somehow test transistors.)

(2) If I apply something like 5 volts or perhaps 8 volts to just the
LED in reverse to its normal polarity then will I cause the LED damage?

Most diode checkers built into multimeters don't provide enough forward bias
voltage to check a LED.

I quick-check LEDs by using a 1k resistor in series with a 9 volt battery as
mentioned by another poster here. I have never damaged a LED using this
method even if I reverse bias a LED.

One way to quickly tell the polarity of the leads is by looking inside the
LED. The lead attached to the chip carrier is the cathode or negative
terminal. The chip carrier is the larger of the two pieces of metal inside
the LED. The other terminal has a fine wire running over to the chip on the
carrier. You may need a powerful magnifying glass to see the wire and the
chip but the size of the cathode in relation to the anode is unmistakable.

I have only seen one or two exceptions for identifying the cathode this way
with some special, high-wattage LEDs.

Dorian

 

[James Thompson]

I have a light consisting of one super-bright white LED which plugs into
the USB port of my PC. It was flickering and now does not illuminate.

The supply leads to the LED are delivering 5v from the USB port but my
meter wouldn't show the sort of intermittency which might cause flicker.

On dismantling the light, I see the leads of the LED are the same length
and are unmarked.

(1) In general, how can I test such an LED using my multimeter and/or
battery? (If it helps, my multimeter has a socket marked E.B.C.E which
I think can somehow test transistors.)

(2) If I apply something like 5 volts or perhaps 8 volts to just the
LED in reverse to its normal polarity then will I cause the LED damage?

Simply use a 3 volt coin cell to test it. Same as your pc clock battery.
They use them in the small keychain lights with a white led.

 

[ian field]

Simply use a 3 volt coin cell to test it. Same as your pc clock battery.
They use them in the small keychain lights with a white led.

The LED keyfobs I've opened used 2x 2016 coin cells for white LED and 1x
2032 for red.

 

[Hosé]

I have a light consisting of one super-bright white LED which plugs into
the USB port of my PC. It was flickering and now does not illuminate.

The supply leads to the LED are delivering 5v from the USB port but my
meter wouldn't show the sort of intermittency which might cause flicker.

On dismantling the light, I see the leads of the LED are the same length
and are unmarked.

(1) In general, how can I test such an LED using my multimeter and/or
battery? (If it helps, my multimeter has a socket marked E.B.C.E which
I think can somehow test transistors.)

(2) If I apply something like 5 volts or perhaps 8 volts to just the
LED in reverse to its normal polarity then will I cause the LED damage?

dumb ist.

 

[Adam Aglionby]

I have a light consisting of one super-bright white LED which plugs into
the USB port of my PC. It was flickering and now does not illuminate.

Cheap white LEDS are a bit failure prone, flickering before end of life
not unknown.

The supply leads to the LED are delivering 5v from the USB port but my
meter wouldn't show the sort of intermittency which might cause flicker.

On dismantling the light, I see the leads of the LED are the same length
and are unmarked.

(1) In general, how can I test such an LED using my multimeter and/or
battery? (If it helps, my multimeter has a socket marked E.B.C.E which
I think can somehow test transistors.)

As suggested battery and resistor, white blue and true green LEDs need
at least 3.something V to light, so a single 1.5V cell wont work , and
even 2 may not be quite enough to reach minimum voltage hence most LED
torches 3 cell, 9V and 1K would be fine, drop it to 390R if you want
bright.

(2) If I apply something like 5 volts or perhaps 8 volts to just the
LED in reverse to its normal polarity then will I cause the LED damage?

Stick with the 1K until you work out polarity.

Adam

 

[Simoc]

(2) If I apply something like 5 volts or perhaps 8 volts to just the
LED in reverse to its normal polarity then will I cause the LED damage?

Like already mentioned, never connect to a constant voltage without a
series resistor. But a comment on connecting in reverse polarity:

In the basic theory, in normal use, where you limit the current by a
resistor, there is no "the highest possible _supply voltage_", as long
as you equip it with a resistor calculated correctly according to the
voltage (as it gets dropped on the correctly calculated resistor). But
notice that if the LED is accidentally connected in reverse polarity,
there will be no current, which results in no drop in the resistor, so
there will be the full supply voltage across the LED, regardlessly to
the resistance of the resistor.

So then if the supply voltage is higher than the rated max reverse
voltage, connecting in wrong polarity will destroy the led even if
there is a correct-resistance series resistor.

So short answer: don't connect in reverse polarity.

 

[Spehro Pefhany]

Like already mentioned, never connect to a constant voltage without a
series resistor. But a comment on connecting in reverse polarity:

In the basic theory, in normal use, where you limit the current by a
resistor, there is no "the highest possible _supply voltage_", as long
as you equip it with a resistor calculated correctly according to the
voltage (as it gets dropped on the correctly calculated resistor). But
notice that if the LED is accidentally connected in reverse polarity,
there will be no current, which results in no drop in the resistor, so
there will be the full supply voltage across the LED, regardlessly to
the resistance of the resistor.

So then if the supply voltage is higher than the rated max reverse
voltage, connecting in wrong polarity will destroy the led even if
there is a correct-resistance series resistor.

That's not entirely correct. Many LEDs have a typical breakdown far in
excess of the max rated reverse voltage, so it's more correct to say
that it "might" or "could" damage the LED. Many 5V rated LEDs have a
typical reverse breakdown in excess of 30V. I also think that it's not
likely to damage the LED unless the power dissipation is too high,
even if the LED conducts some reverse current.

So short answer: don't connect in reverse polarity.

I have yet to see an LED or LED array s.t. probing it with a regulated
5V source in series with 1K could cause any damage. Some are
apparently sensitive to ESD, so take care with that as well.


Best regards,
Spehro Pefhany

 

[Lostgallifreyan]

Like already mentioned, never connect to a constant voltage without a
series resistor. But a comment on connecting in reverse polarity:

In the basic theory, in normal use, where you limit the current by a
resistor, there is no "the highest possible _supply voltage_", as long
as you equip it with a resistor calculated correctly according to the
voltage (as it gets dropped on the correctly calculated resistor). But
notice that if the LED is accidentally connected in reverse polarity,
there will be no current, which results in no drop in the resistor, so
there will be the full supply voltage across the LED, regardlessly to
the resistance of the resistor.

So then if the supply voltage is higher than the rated max reverse
voltage, connecting in wrong polarity will destroy the led even if
there is a correct-resistance series resistor.

So short answer: don't connect in reverse polarity.

No. Volts are EMF, they move the force that is represented by current.
While the voltage appears to be high across the LED in reverse, it's the
current that does the actual damage, if any be done, and we already know
that there is no current.

Even ESD is only dangerous because it instantly and momentarily causes high
current.

 

[Don Klipstein]

That's not entirely correct. Many LEDs have a typical breakdown far in
excess of the max rated reverse voltage, so it's more correct to say
that it "might" or "could" damage the LED. Many 5V rated LEDs have a
typical reverse breakdown in excess of 30V. I also think that it's not
likely to damage the LED unless the power dissipation is too high,
even if the LED conducts some reverse current.


I have yet to see an LED or LED array s.t. probing it with a regulated
5V source in series with 1K could cause any damage. Some are
apparently sensitive to ESD, so take care with that as well.

The thing with those ESD-sensitive LEDs and GaN and InGaN ones in
general is some extreme intolerance to conducting any current as a result
of reverse breakdown. Reverse current conduction due to reverse breakdown
apparently causes some destructive elecrolysis effect that forms some sort
of short or partial short.

At least some other LEDs don't like reverse breakdown, due to current
conduction and heating being uneven and localized. But at least it
appears to me unlikely that damage will occur that way from forcing a
milliamp or a fraction of a milliamp through in the wrong direction.
Some LEDs also have issues of ingredients diffusing out of place as a
function of voltage, temperature and time. But I surely don't expect a
few seconds with 9, 12 or 15 volts with theb chip at 30 degrees C to be
any problem in that area. That is mainly a problem with elevated
temperatures combined with long time, although a few LEDs can degrade
a little significantly from normal forward voltage within just hundreds of
hours at higher allowed temperature. Thankfully that one progresses at a
rate slowing greatly with progressing degree of damage - a (formerly) HP
app note appears to tell me that if the LED prone to this degradation
mechanism degraded 10% from that mechanism in the first 100 hours, it will
degrade another 10% in the next 900 hours and another 10% in the following
9,000 hours. That app note advises reducing duty cycle rather than
instantaneous current if you want average current less than 10 milliamps,
and appearing to mainly have that being more important if the LEDs will be
experiencing higher temperatures. This app note was specific to InGaAlP
chemistry. One thing I also suspect is that these LEDs having
wear-and-tear on them could have high resistance partial shorts or
something reducing their ability to work well at lower currents.

- Don Klipstein ([email protected])

 

[Dave M.]

The LED keyfobs I've opened used 2x 2016 coin cells for white LED and
1x 2032 for red.

.... and don't be fooled. The led probably has a built in current source,
otherwise you could not connect the batteries to the LED without a
resistor!

 

[Spehro Pefhany]

... and don't be fooled. The led probably has a built in current source,
otherwise you could not connect the batteries to the LED without a
resistor!

A CR2032 has around 50-100 ohms of internal resistance at some tens of
mA, and that's what they are depending on.


Best regards,
Spehro Pefhany

 

[Dave M.]

A CR2032 has around 50-100 ohms of internal resistance at some tens of
mA, and that's what they are depending on.


Best regards,
Spehro Pefhany

I don't think those figures will establish a resonable current range
,tolerance or predictable battery life. I do know current limiters are
used in those LED's.

 

[Dave M.]

I don't think those figures will establish a resonable current range
,tolerance or predictable battery life. I do know current limiters are
used in those LED's.

....however a current limiter would not work with in this application
because of a barely existing compliance voltage, so you must be right.

 

[Victor Roberts]

I don't think those figures will establish a resonable current range
,tolerance or predictable battery life. I do know current limiters are
used in those LED's.

We're talking about cheap here, not predictable or
reasonable.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.

 

[Don Klipstein]

I don't think those figures will establish a resonable current range
,tolerance or predictable battery life. I do know current limiters are
used in those LED's.

I happen to have disassembled keyfob lights and used different LEDs in
them and interchanged their LEDs in other applications. Keyfob lights in
my experience use LEDs that don't have any components or circuits added
within the LED bodies. Keyfob lights in my experience usually lack
resistors or added circuits.
In my experience keyfob lights do depend on internal resistance of the
battery. With fresh batteries the LED current sometimes (as in often)
exceeds the maximum rating for the LED. My guess is that keyfob light
manufacturers depend on at least one and maybe both of the following being
true:

1) The battery heatsinks the LED

2) The amount of time throughout the product's life spent with the LED
being overpowered does not degrade it to a noticeable extent, or does not
with a high percentage of users.

- Don Klipstein ([email protected])

 

[Simon Waldman]

2) The amount of time throughout the product's life spent with the LED
being overpowered does not degrade it to a noticeable extent, or does not
with a high percentage of users.

This is greatly assisted by the fact that most of these units are so
badly made that they fall apart after a few weeks anyway...

 

[ian field]

This is greatly assisted by the fact that most of these units are so badly
made that they fall apart after a few weeks anyway...

The keyfobs I bought came in 2 versions - one of which was surprisingly well
made! Aside from the ones I've cannibalised for the LEDs and batteries, I do
actually use one for its intended purpose and use it often enough to wear
out the battery.

The better made version has a button that can be pressed or also works as a
slide switch for continuous operation, the less well made version is simply
2 flexible polythene shells that are squeezed to light the LED - these are
the first to get stripped when I want the LED or the battery, the LED on my
heyring is still working after 3 or 4 new batteries.