PDA

View Full Version : LED Brightness Problem



Jasontron
10-01-2015, 09:41 AM
Hi,
So I'm in the process of making an LED string blade using around 80-90 LEDs
wired together.

The LED specifications are:
FV: 3.2~3.4V
FC: 20mA
mcd: 12,000-14,000
Viewing angle: 120-140 degrees

I'm using a 3.7V 5.3Ah Li-ion battery to power this, and am also using a
200ohm 2 watt resistor for the LEDs.

This is all well and good, except for the fact that the LEDs really aren't that bright.
After being put in the blade with a diffuser, I can barely see the light in a lit room.

Is there something I'm doing wrong with the setup? Are the LEDs the wrong kind? If the
LEDs are wrong, is there any you guys know of that would work better?

Thanks a bunch,

Jason

Silver Serpent
10-01-2015, 10:46 AM
How are they wired? Multiple segments, or all of them in parallel? If segments, how many LEDs per segment?

I'm curious how you calculated your resistor value. That's much higher than you'd need for even a single LED at those ratings, let alone 80-90 of them.

Jasontron
10-01-2015, 12:02 PM
Thanks a lot for the quick reply,
First of all, I'm wiring the LED string in parallel.
Second of all, I'm not sure what I was thinking doing the resistor the way I did.
I used the calculator you provided and saw that I would need an 18ohm 1/4watt resistor
For each LED in the string. Is there a way I can just use one resistor for the whole string?

Thanks,

Jason

Silver Serpent
10-01-2015, 12:27 PM
Yes. If you're wiring them in parallel, you add up the mA of all the LEDs in each segment. So, 20mA * however many LEDs = total current for that segment. Treat that segment as a single LED for purposes of resistor calculation, and each segment will get a single resistor.

Please note that most people who build LED string blades will make 5-6 segments, so that they can get the extend/retract scrolling effect for their blades. You'll need an appropriate soundboard or LED controller to handle this.

If you wire up 80 20mA LEDs in parallel, you'll be needing about 1600 mA of current. If you wire them up in segments of 20 LEDs each, then each segment will only need 400 mA of current. The total for the whole strip will be the same, but you're likely to run into heat issues if you're pumping 1600 mA through your wire as opposed to 400 mA.

Jasontron
10-01-2015, 02:12 PM
Alright, so I entered those values in the calculator (this time for 1000 mA because I'm thinking of doing two segments of 50) and the calculator said I'd need a 1ohm 2 watt resistor. That's great, but if I wire the two segments in series I'll need 6.8V not 3.4, correct? If that's the case, I'll need another 3.7V li-Ion battery. Is this all right? Or am I going crazy.

Thanks for all the help!

Jason

Silver Serpent
10-01-2015, 07:10 PM
Or wire the two segments in parallel. You'll end up with a total drain on the battery of 2000mA, perfectly fine for an 18650 sized li-ion. Each segment can be considered as a single LED, 1000mA each.

Jasontron
10-01-2015, 07:37 PM
Thanks!

You've been a big help!

Jason

Jasontron
10-01-2015, 08:20 PM
Oh, and one more quick thing,
I was looking up what people said about wiring LEDs in parallel with one resistor, and there was a lot of talk about it being a bad idea.
Here is the link to the this thread: http://electronics.stackexchange.com/questions/22291/why-exactly-cant-a-single-resistor-be-used-for-many-parallel-leds
Is this a problem applicable to me?

Thanks for all the help

Jason

Silver Serpent
10-02-2015, 04:40 AM
It's something you need to be aware of, but it's a bigger problem with the higher-current LEDs.

Obi1
10-02-2015, 05:35 AM
(Jedi mind trick: "after reading this, you will forget who wrote it. You will not curse my name if the idea I present does not work for you - disclaimer end.")

Actually I've built my share of LED string blades and also modified some toy blades for more brightness.

In case of blue LEDs with max 3.4V, I would not even consider a resistor with a 18650. These LEDs will definitely survive even the 4.2V max voltage this Akku can throw at them. I have a blue one with 2-serial set-up, 2x18650 and so far no issue. Same with a 3-serial red one (3x2.4Vmax=7.2V, but 2x18650 can go up to 8.4 max).
For some Hasbro toy sabers I bypasses all the resistors in the blade. They work perfectly and so much brighter than with the supressed set-up.

Of course all on your own risk...

P.s.: if you have a CF LS, you can even set the drive strength, so no resistor is needed.

Silver Serpent
10-02-2015, 07:25 AM
My advice on string blades is all theory. Obi1's advice is from experience.

Take his advice under strong advisement, you should. :)

Weaver
10-06-2015, 05:08 PM
I am actually having this problem as well, now that I've begun building a CF LS saber. I read the section about drive adjustments, did Class III Seriallel, etc, etc, and bench-tested one segment of LEDs. It's 5 red LEDs per section (2.1v @20mA), three sections in series. The segment instantly heats up to a temperature you can't touch without pain.

When I was wiring the diodes, I measured the forward voltage across the leads of the LEDs themselves. I noticed that the initial forward voltage began dropping almost immediately, then became more stable around 6.1v. I set the drive adjustment to ~6v equivalent and tested the segment. I do not believe I had the heat issue while the segment was connected to my PSU.

So...I'm kind of curious as to what I'm doing wrong. The LEDs were tested at an overdriven 40mA per (200 for the parallel sections). I did use rigid wire for the common anode, as suggested by the CF manual, but it didn't seem to help.

EDIT: I measured 120F with a non-contact thermometer, after about 30 seconds of run time. I am now noticing some dim spots in the segment. I'll try another segment at a lower amperage and see if that helps.

Obi1
10-08-2015, 02:53 AM
Did you check that all LED's in the segment light up? Maybe some of them carry more current due to others not having a proper contact.

Weaver
10-08-2015, 09:51 PM
Well...

Originally, the segment did light up evenly. I drove them at 40mA, under the advisement of the CF LS 6.5 manual. But then I encountered the heat problem, shortly before the dimming occurred. I decided that lowering the Drive Adjust calculation was a prudent move. During my study of the board, I discovered that the battery pack doesn't always output 7.4v. I measured ~8v from its leads, so I used that number for the new calculation.

The segment now reads correct current and voltage when I use my multimeter, and I no longer get the heat issue, but the LEDs in that segment are lost. Once you overheat a diode, it's done. I have to accept that. Of course...that's only 15 LEDs out of a batch of 250. I anticipated losses when I started this project, so I came prepared.

The new segment will be driven using 8v as the pack voltage, and 30mA as the desired current. If that doesn't work...then I know something's very, very wrong.

Side Note: The CF LS 6.5 really is a sweet board. I don't regret my purchase at all. It'll be a hard road, but so worth it.

EDIT: Hmm...you said something about improper contact. One of the LEDs (the last in line, actually) did have a broken lead that I bridged with solder. I didn't think much about it, since it was still conducting, but...could that have caused overheating in the other LEDs?

EDIT (Again): I've been thinking about this. It makes sense to me that the bridged connection might have a different resistance than the others. If that resistance is any higher, it could lead to overheating and overloading of the other diodes. So...with that in mind, should I try 40mA/LED again? Data sheet says 20, CF manual says 40 should be fine. I do have extra LEDs, but I'm not keen on blowing them out any more often than I have to.

SpaceWindu
11-10-2015, 05:32 PM
Well...

Originally, the segment did light up evenly. I drove them at 40mA, under the advisement of the CF LS 6.5 manual. But then I encountered the heat problem, shortly before the dimming occurred. I decided that lowering the Drive Adjust calculation was a prudent move. During my study of the board, I discovered that the battery pack doesn't always output 7.4v. I measured ~8v from its leads, so I used that number for the new calculation.

The segment now reads correct current and voltage when I use my multimeter, and I no longer get the heat issue, but the LEDs in that segment are lost. Once you overheat a diode, it's done. I have to accept that. Of course...that's only 15 LEDs out of a batch of 250. I anticipated losses when I started this project, so I came prepared.

The new segment will be driven using 8v as the pack voltage, and 30mA as the desired current. If that doesn't work...then I know something's very, very wrong.

Side Note: The CF LS 6.5 really is a sweet board. I don't regret my purchase at all. It'll be a hard road, but so worth it.

EDIT: Hmm...you said something about improper contact. One of the LEDs (the last in line, actually) did have a broken lead that I bridged with solder. I didn't think much about it, since it was still conducting, but...could that have caused overheating in the other LEDs?

EDIT (Again): I've been thinking about this. It makes sense to me that the bridged connection might have a different resistance than the others. If that resistance is any higher, it could lead to overheating and overloading of the other diodes. So...with that in mind, should I try 40mA/LED again? Data sheet says 20, CF manual says 40 should be fine. I do have extra LEDs, but I'm not keen on blowing them out any more often than I have to.


If you can get them, try better red LEDs with at least 50ma capacity maximum. Then, when 40-50 passes them there are no problems at all. 20ma is too low for use with CF LS without resistors IMO

Obi1
11-16-2015, 09:33 AM
Solder bridge is metallic contact and therefore low resistive, so should be no issue. But overdrive of the LED's is. I usually go by the voltage rating, the reds I used for my Class III are rated at 2.4V max, that gives you 7.2V for 3 in series. Close enough to the 7.4V of the supplies we usually use for the CF sabers.
But alas you are right, these batteries can be loaded up all the way to 8V, and that will be ~2.7V on the LED's. On the long run it can overstress the LEDs. So the drive adjustment shall be used to bring that down to an acceptable level. The drive adjustment works with voltage settings, so just ensure you assume 8.2V (be a bit pessimistic) and calculate the drive adjustment accordingly. The LEDs will be still bringt enough.
Actually when I started with the LS sabers, I tried the 5mm LED's (different ones, green, red, blue, UV, you name it). For short time they can all withstand higher voltages than the max rating, but interestingly driven slightly above the max they still gain a little brightness, but afterwards the light dims rapidly (until destruction). I do not have a physic explanation for this, but it shows that it does make sense to drive the LEDs in their safe operation area (around nominal condition, so 2V for reds).