Battery Powered Infrared LED Array - General Assistance

Thread Starter

Nd_

Joined Nov 25, 2014
11
I am relatively experienced with general electronics but the general conduct in regards to LEDs is something that I am not extensively knowledgeable in. I am attempting to create an IRLED array of roughly 60 LEDs (exact number is available to change) that will all be powered by a battery.

My general design consists of a 6 x 10 array of LEDs (6 in series with a total of 10 rows in parallel) and I am not entirely sure which battery to specifically use. I want this array to be portable (so the option of using a wall power source is not available) as well as potentially allowing the array to have power for an extensive period of time (ideally the longest that can be achieved). The array does not have to be continuously on for the entire period of time as ideally it will be controlled to be turned on/off.

The IRLED that I have chosen to use is listed here (http://www.digikey.com/product-detail/en/SIR383/1080-1175-ND/2675666), but the general specs that are important are as follows:
  • Current - DC Forward (MAX) = 100mA
  • Voltage - Forward (Typical) = 1.3 V
Originally, I thought of using a typical 9V battery to power this array, so I had calculated the required resistor per series as R=(8.4V-(1.3V)(6))/100mA = 6 ohms. I also used the LED array wizard calculator to confirm this result, and it provided a slightly different answer of 6.8 ohms (http://led.linear1.org/led.wiz).

I figured that using a Lithium-Ion battery would be the best choice as it would allow for a longer use, but I am unsure of the voltage and type of battery interactions in regards to powering an array of IRLEDs. I have searched these forums as well as the general internet and I have unable to find a direct answer to my question.

Any assistance or guidance on this project would be greatly appreciated.
Thanks
 

#12

Joined Nov 30, 2010
18,224
Several things to discuss here. First, http://batteryuniversity.com/learn/article/lithium_based_batteries
seems to say there is no such thing as an 8.4 volt lithium battery. Could you clear this up for me?

Second, you're cutting it too close to the bone with the series resistor. You should calculate for the highest battery voltage and expect the brightness to dim as the battery gets used up, but more serious is that you can't trust LEDs to be all that accurate about their forward voltage. You either measure after you buy them, or calculate the series resistor at the highest battery voltage and the lowest possible Vf.

If that's enough information, fine. If not, this is a conversation.
 

#12

Joined Nov 30, 2010
18,224
ps, I think I saw an ad for a UV flashlight at Harbor Freight 3 days ago. You might want to check there quickly to see if it will do the job and save you dozens of hours designing and building.
 

wayneh

Joined Sep 9, 2010
17,496
You'll need a LOT more juice. Each LED needs about 100mW, so in total you need ~6W. If you want, say, 4 hours you'll need 24 watt-hours. A 12V, 2 Amp-hour battery would do it. Using the higher voltage would also allow longer strings and fewer resistors.

Your project might be big enough to warrant an LED controller instead of resistors for current limiting. This could gain a bit of efficiency and better hold constant brightness as the battery fades.
 

#12

Joined Nov 30, 2010
18,224
Now you got me. I'm not sure. I probably ignored it because it was IR and that means nothing to the work I do. So, yeah, I think it was IR.
 

Thread Starter

Nd_

Joined Nov 25, 2014
11
Several things to discuss here. First, http://batteryuniversity.com/learn/article/lithium_based_batteries
seems to say there is no such thing as an 8.4 volt lithium battery. Could you clear this up for me?

Second, you're cutting it too close to the bone with the series resistor. You should calculate for the highest battery voltage and expect the brightness to dim as the battery gets used up, but more serious is that you can't trust LEDs to be all that accurate about their forward voltage. You either measure after you buy them, or calculate the series resistor at the highest battery voltage and the lowest possible Vf.

If that's enough information, fine. If not, this is a conversation.
The 8.4 V battery I was referring to in the original post was just a general 9V (I believe alkaline) battery that I have used in the past for school related projects. I was referencing this battery (http://data.energizer.com/SearchResult.aspx) that I found on the energizer website. I don't believe that this type of battery would produce the power that I am looking for. As far as being completely accurate with the forward voltage, I will have to manually measure them once they arrive but for now I am hoping to grasp a purely conceptual idea of how to go about it.

I want to personally create this device as I am hoping to implement it in a future design, so I would not be able to necessarily reverse engineer a device that contains IRLEDs. Regarding the website you posted about lithium batteries, which one would be optimal for what I am trying to achieve? How important is the specific energy and loading rate?
 

Thread Starter

Nd_

Joined Nov 25, 2014
11
You'll need a LOT more juice. Each LED needs about 100mW, so in total you need ~6W. If you want, say, 4 hours you'll need 24 watt-hours. A 12V, 2 Amp-hour battery would do it. Using the higher voltage would also allow longer strings and fewer resistors.

Your project might be big enough to warrant an LED controller instead of resistors for current limiting. This could gain a bit of efficiency and better hold constant brightness as the battery fades.
What exactly is an LED controller and how does one implement it within this array? Ideally I would also like to have a microcontroller interact with the array to turn them on/off. Are there any specific 12V, 2 ah batteries that would be optimal for this?
 

#12

Joined Nov 30, 2010
18,224
Here's the datasheet on the 8.4 volt lithium based battery. http://data.energizer.com/PDFs/la522.pdf
That's about a 90 minute run time for 100 ma, each.
Other problems: The 100 ma spec is for a 1% duty cycle. That means they will fry if you go over 20 ma continuous.
http://media.digikey.com/pdf/Data Sheets/Everlight PDFs/SIR383.pdf
Vf is 1.3 to 1.6 @ 20ma.

Good thing you checked here first. You were about to shoot yourself in the foot. :D

Now, let's start over with a design that won't let the smoke out. ;)

ps, I'm getting nothing searching Harbor Freight...UV or IR.
 
Last edited:

Thread Starter

Nd_

Joined Nov 25, 2014
11
Here's the datasheet on the 8.4 volt lithium based battery. http://data.energizer.com/PDFs/la522.pdf
That's about a 90 minute run time for 100 ma, each.
Other problems: The 100 ma spec is for a 1% duty cycle. That means they will fry if you go over 20 ma continuous.
http://media.digikey.com/pdf/Data Sheets/Everlight PDFs/SIR383.pdf
Vf is 1.3 to 1.6 @ 20ma.

Good thing you checked here first. You were about to shoot yourself in the foot. :D

Now, let's start over with a design that won't let the smoke out. ;)

ps, I'm getting nothing searching Harbor Freight...UV or IR.
Wow, thank you very much as this design could've gone south very quickly. Since it seems that I missed a pretty important aspect in regards to LED function, do you have any suggestions on where/how to start over with this design? Would using a 12V 2 ah battery be a more optimal choice? If so, how would we change the LED array?
 

#12

Joined Nov 30, 2010
18,224
Searching for the brightest 20ma IR emitter is difficult because they are almost always advertised at their 100 ma, 1% rating. :mad:

We have to find a good LED before we do a design.
PS, Quit asking about the 12V, 2AH battery that was calculated for a forward current that can't possibly happen.

Let's try the LED you already chose, fed the proper amount of power.
math math math...
 

wayneh

Joined Sep 9, 2010
17,496
I'm getting nothing searching Harbor Freight...UV or IR.
I'm guessing it may have actually been UV, like the ones you see on CSI. I can think of a few applications for that (UV is handy for finding pet stains), but I can't think of anything I could use an IR flashlight for.
 

Thread Starter

Nd_

Joined Nov 25, 2014
11
Searching for the brightest 20ma IR emitter is difficult because they are almost always advertised at their 100 ma, 1% rating. :mad:

We have to find a good LED before we do a design.
PS, Quit asking about the 12V, 2AH battery that was calculated for a forward current that can't possibly happen.

Let's try the LED you already chose, fed the proper amount of power.
math math math...

Using the LED (with the typical forward voltage of 1.3V), and a new array of 5 x 10, I've come up with this:

The new voltage of the LED array would be (1.3V) x (5) = 6.5V
Vsupply = 8.4V
Using a forward current (If) = 20mA
R=(Vsupply-Vled)/(If) = (8.4V - 6.5V)/20mA = 95 ohm resistor per series

I'm a bit confused as far as determining the correct amount of power in accordance with the 20mA forward current. Am I missing something other than just changing the array and resistor value for accommodation? The new array would require less voltage which would give more space for the array.
 

#12

Joined Nov 30, 2010
18,224
Internal resistance problem.

12 strings of 5 LEDs = 60 LEDs.
12 strings x .02A = .24A
Vsupply at that load for the 9V LiMnO2 battery is 7.85 volts.
R = 68 ohms.

Every way I twist that 9V battery, it's less than 4 hours per charge.
Is that good enough, Nd?

ps, your math is making good sense except you didn't consider every little detail of the battery behavior.

PS, My real life calls. I'm going to have to disappear for a while.
 
Last edited:

Thread Starter

Nd_

Joined Nov 25, 2014
11
Internal resistance problem.

12 strings of 5 LEDs = 60 LEDs.
12 strings x .02A = .24A
Vsupply at that load for the 9V LiMnO2 battery is 7.85 volts.
R = 68 ohms.

Every way I twist that 9V battery, it's less than 4 hours per charge.
Is that good enough, Nd?

ps, your math is making good sense except you didn't consider every little detail of the battery behavior.
A 2-3 hour battery is definitely sufficient, but if I wanted to increase that life to maybe 4-5 hours would I just need to include a higher voltage LiMnO2 battery?

Also, if you don't mind me asking, how did you calculate the Vsupply at that load (where it is equal to 7.85V)?
 

wayneh

Joined Sep 9, 2010
17,496
Uh oh, this time I read the data sheet. Look again. This LED actually DOES use 100mA continuous, and this is much more in line with typical IR LEDs. The peak is 1A for the 1% duty cycle pulses.

I was going to suggest to the OP that he use fewer LEDs rated for a higher power, but that's where he is already. I suppose there is probably a 500mA or 1A IR LED out there but this would have heat dissipation issues.
 

#12

Joined Nov 30, 2010
18,224
Post #9, Energizer datasheet, graphs.

Want more run time? Put (2) 9V batteries in parallel and recalculate the starting voltage based on the graph for 100ma and that is 8.2V average starting voltage. 1.7V / .02A = 85 ohms. Buy 91 ohms.
 

Thread Starter

Nd_

Joined Nov 25, 2014
11
Looking back at the IRLED datasheet (http://media.digikey.com/pdf/Data Sheets/Everlight PDFs/SIR383.pdf), it does show the continuous forward current is rated at 100mA, while the peak forward current is 1 A with a duty cycle of 1%.
Does this 100mA forward current mean that a lower value resistor is needing per string? Or should the 20mA forward current still be used in this setting with this IRLED?

I also have a purely hypothetical question regarding the peak forward current and the duty cycle. If I were to utilize the fact that the peak forward current is noted to be 1A at a duty cycle of 1%, would that mean I would be able to connect many more IRLEDs together if they were programmed to be turned on for 10 ms (0.01 s) every 1 second? Would this allow for the LED array life to be much longer as opposed to being run continuously for set hours at a time? Given this new duty cycle of the LEDs being turned on for 10ms, if I was able to then connect a significantly larger array of them together would the effect be negligible in the long run of infrared light emittance?
 

wayneh

Joined Sep 9, 2010
17,496
When LEDs flash fast enough to look continuous, the perceived brightness is roughly proportional to the average current. So pulses don't gain you anything in efficiency. In fact, the curve of brightness against current lays over slightly at higher current, so you actually lose efficiency by driving a 2X current for 1/2 the time, for instance.

You could design your system to run each LED at 20mA, but you'll need a lot more of them - to get the same amount of light - than if you target, say, 90mA. I wouldn't target the rated maximum of 100mA - they'll last longer if you don't.
 
Top