Overdriving Nichia white - Any Actual Problems?

[Quickbeam]

Hey, all. I know this topic has come up many times, but all the answers have been a bit wishy-washy, so for the record: Has anyone actually fried Nichia white LEDs by driving them with 4.5V each? Has anyone had the LEDs visibly damaged/dimmed as a result of driving at 4.5V? If so (for either), please provide the type of battery cells used and any other relevant information. If you've never had any problems in this configuration, what precautions did you actually take? Heatsinking? Nothing?

I would like to drive some Nichia White LEDs at 4.5V, but I can't find any difinitive answers in the forums as to actual damage/destruction occuring - just speculation. I hope we might be able to avoid speculation completely here

and instead share everyone's actual experiences. Thanks !

(BTW: Referring to the 5mm 5600mcd LEDs)

 

[**DONOTDELETE**]

hey,
I have tried driving white leds directly from 4.5 volts, i only left them connected for about 5 seconds though as they got very hot and if left like that would have surely self destructed in a few minutes.

i am going to explain how i drive my 10 led bank from 4.5 volts in another topic if you would like to know.

 

[PeLu]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Quickbeam:
Hey, all. I know this topic has come up many times,... : Has anyone actually fried Nichia white LEDs by driving them with 4.5V each?<HR></BLOCKQUOTE>

You should mention the desired current, not the voltage.
At 4.5 V you may have 80mA with one LED and 60mA or 100mA with another. One time I had one with some 60mA at 5V (ok, this was a very early one back almost 5 years).

And at 100mA you will get about 41% of it's rated efficiency, at 70mA about 50%.
And thermal runaway is possible at constant 4.5V.
Damage may be slow, the phospor will degrade in 100 hours or less.
If you use more than one LED try to heatsink them in a proper way and try to keep them at the same temperature.
And do not forget: the LEDs vary not only in voltage at nominal current, they also do so in dynamic resistance.
From where do you get constant 4.5V?

 

[Quickbeam]

GADGET !!! I think you may be able to help me with this one, but anyone else can certainly help out. I put 3 nichia whites in a pr bulb base with no resistors to run in a 4 aa light with a dummy battery (so 4.5v total) with either 0 resistance or 10 ohm resistance built in to the dummy cell. When I run a single nichia directly on the 4.5v it draws 75 mA. When I drive the 3 bulbs in the base the entire contraption draws about 54 mA.

But it's really bright and the pr base (soldered to the negative leads and serving as a makeshift heatsink) warms right up! I take the mA readings right at the positive end of the base. What the heck is going on?

I matched the LEDs for brightness by driving them at 3V. I have 2 that only draw 60 mA at 4.5v and were significantly dimmer than the others when compared at 3v. All the others in the batch were about the same brightness so I know the 3 in the base are very close. I would expect the 3 in the bulb base to draw about 75x3=225 mA...

Now just to test that it's not the meter, I put a single nichia in a pr base to run off 6v with a 100ohm resistor and the meter reads 23mA - almost exactly what it should be per ohm's law equation (24 mA)... AARRRRGGGGGG!

 

[Go Go Gadget Flashlight]

Doug,
I'm stumped.
EDIT: Not stumped anymore.

Read all the info below in case this doesn't work, but...

I just took my 3 LED bulb, 3 AA's, and my meter and did a test.

Meter set on 400mA range, 77mA thru bulb.
Meter set on 20A range, 260mA thru bulb!

So, right there may be the problem.

The meter has to have some internal resistance to measure the V drop internally. I bet this is throwing it off significantly. You meter may be like mine and have a pretty big internal resistance for its "mA" range.


Let me paraphrase so I know I understand the situation:
You have 3 Nichais that are close in their normal current draws.
They are in parallel, mounted in a PR base.
You have your meter set on the mA range (200 or 400mA max).
You're driving them with 3 AA alkalines (4.5v @ no load).
Your meter is telling you you only have 54mA going thru the circuit.

Correct any of these points that are wrong or add anything I've missed.

In the mean time, measure the voltage across the PR base while it's running.
And, try your meter on a high amp range. I've found that my expensive Craftsman meter has a fairly high resistance on the mA range. So to get accurate amp readings, I have to put it on the 20 amp range.

 

[Quickbeam]

AHHHHHHHH..... All of the information you listed "below" (above) was correct

. I'll try the 20A range on the meter later tonight. Good Idea! We might just have a winner!

 

[PeLu]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Go, Go Gadget Flashlight:
I've found that my expensive Craftsman meter has a fairly high resistance on the mA range. So to get accurate amp readings, I have to put it on the 20 amp range.<HR></BLOCKQUOTE>

Usually Ammeters are not rated for their internal resistance, they give you the voltage drop for full scale reading. OK, it is easy to calculate it. And I cannot find significant differences in between cheap and expensive meters. If you really need a very low drop, it is best to use an external shunt.
Beware: if you use your 20A range, accuracy will be pretty low (check manual).

Or to use a regulated power supply which compensates for the drop at the Ammeter.

 

[PeLu]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Quickbeam:
I put 3 nichia whites in a pr bulb base with no resistors to run in a 4 aa light with a dummy battery (so 4.5v total) <HR></BLOCKQUOTE>

You always assume that you have 4.5V with 3 alkalines. It would be better to state the real voltage (under load of course). Open circuit voltage does not say too much. If you use two DMMs you may use one for the current and one for the voltage.

 

[MrAl]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by PeLu:
You always assume that you have 4.5V with 3 alkalines. It would be better to state the real voltage (under load of course). Open circuit voltage does not say too much. If you use two DMMs you may use one for the current and one for the voltage.<HR></BLOCKQUOTE>

Hi again,
Yeah i agree. When you do a test on batteries of any type under load, you have
to measure the ACTUAL battery voltage while
you do the test because the battery has significant
internal resistance, so as soon as you hook
the leds up the battery voltage drops.
If you hook three LED's up and it drops to
3.75 volts, you will calculate very different
values then constantly using 4.5volts in the
calculations. Very important.
Also, another way to measure current in these
kinds of circuits is to connect a 0.1 ohm, 1% resistor
in series with the led and measure the voltage across the 0.1 ohm resistor when
the leds are turned on. The current is then
equal to the voltage reading divided by 0.1:
I=V/R.
I use a 0.2 ohm, 25watt, 1% resistor a lot and
get good results.
Set the digital voltmeter down to a low scale.

Good luck with it,
--Al

 

[Quickbeam]

RE: measuring voltage - what's the best way? I assume you break open the circuit and place the meter in the circuit as you would when reading amps, correct? No? Do you place the meter in parallel? Either way, where are the best connection points? Thanks!

 

[Go Go Gadget Flashlight]

Doug,
for voltage, connect meter in parallel with the bulb (leds). This will tell you the voltage going thru (V drop).

 

[PeLu]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Go, Go Gadget Flashlight:
for voltage, connect meter in parallel with the bulb (leds).<HR></BLOCKQUOTE>

I want to state that checking the voltage with modern DMMs does not influence the circuit in any way we should care about. Unlike checking the current.
(and I'm unhappy with the expression that voltage 'goes thru' .-)

 

[Quickbeam]

Sure enough, I took the reading with the meter on 20A setting and it gave me a .07-.08A (70 to 80mA) reading. Also, with the LEDs running I had a 3.8V reading across the bulbs. This makes more sense. Now the Voltage reading I got, the "voltage drop", is that the voltage being put out by the batteries, or the voltage being used by the leds? I don't understand the term "voltage drop".

 

[Go Go Gadget Flashlight]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by PeLu:
....(and I'm unhappy with the expression that voltage 'goes thru' .-)<HR></BLOCKQUOTE>

Which is why I included the parentheses.

"This will tell you the voltage going thru (V drop)."

PeLu, you have to remember, this is a flashlight forum. Not everyone here is an electrical engineer. Some of these guys don't even have voltmeters and may have only had limited electronics exposure in high school.

Doug,
"the voltage being used by the leds?"
Yep, it's the voltage "dropped across" a component.

It surprises me that 3 LEDs would pull the batteries down to 3.8v. Are they fresh alkalines? Sounds like they have a lot of internal resistance. Or, your wiring from the batteries and/or the switch has resistance.

I did a second test using 3 'almost new' AA alkalines into three parallel LEDs. This time, I got 500 mA! The voltage drop across the LEDs was 3.9v. The leads on the LEDs got hot pretty quick!

A lot depends on the resistance in your wiring. On the second test, I used a battery holder with short wires and a breadboard (500mA). The test from yesterday (260mA) was done using a 3 cell stick from my DB and a different 3 LEDs that are in a PR base.

Is this PR bulb one that you made? There's not a resistor in there, is there?

 

[Quickbeam]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Is this PR bulb one that you made? There's not a resistor in there, is there? <HR></BLOCKQUOTE>
Sorry Gadget! I forgot to mention I was using a 10 ohm resistor in the dummy battery to regulate the voltage a bit. The dummy battery was in the circuit when I took the measurement.

The alkalines are relatively fresh, but not new, and the PR bulb is one I made with no resistors but I was careful to make really good solder connections.

 

[Go Go Gadget Flashlight]

Aha!!

Well, there you go.

Remember this (simplified) formula?

R = (Vbatt - Vled)/Iled

If you use 4.5v batt, 4v LED, and 50 mA you get 10 ohms resistance. Throw in some variation for actual conditions and I could believe 70 mA.

Recalculate this based on numbers you have measured:
Vbatt = 4.5v
Vled = 3.8v
Desired current:
Iled = .120A (40mA * 3)

That equates to 5.8 ohms.
Since these are theoretical numbers, you will have to experiment with values close to 5.8 ohms. I'd start with two 10 ohms in parallel (5 ohms). Measure the current thru the LEDs using the 5 ohm resistor.

If you want to get "good" numbers for the formula, measure the V drop across the LEDs, the voltage accross the batteries, and the current thru the whole thing while using your 10 ohm resistor.
That way you are getting the battery voltage while under load and a more accurate V drop across the LEDs. Also measure the resistor and compare it to what you calculated with the "good" numbers. It should be very close.

 

[PeLu]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Go, Go Gadget Flashlight:
PeLu, you have to remember, this is a flashlight forum.
<HR></BLOCKQUOTE>

That's why I wrote 'I'm unhappy' and added a smiley and not (insert your favorite profane words) .-)


<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>
is a It surprises me that 3 LEDs would pull the batteries down to 3.8v. Are they fresh alkalines? Sounds like they have a lot of internal resistance.
<HR></BLOCKQUOTE>

We should also state here that when you dischage alkalines mainly their internal resistance rises and their open voltage may only drop little.

 

[PeLu]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Quickbeam:
I forgot to mention I was using a 10 ohm resistor in the dummy battery to regulate the voltage a bit. <HR></BLOCKQUOTE>

Try to use an incandescent bulb instead of the resistor. It works a little bit as a regulator. Try one or two (in parallel) #122 bulb(s). A dummy cell should have plenty of space.