0
\$\begingroup\$

I'm trying to build a portable LED light and things are happening not the way I wish to.

I have a Sumbulbs 12036 LED COB Strip. This LED works with DC 12-14V voltage, it has 10W power and recommended current of no more than 600mA. Here is the link for this LED.

I mounted an aluminum panel on back to ensure a good heat dissipation. The first thing I did was test the LED using a transformer that output DC 12.8V and current of 520mA. I kept the light turned on a few hours and everything looks ok.

Now as I need it portable, I changed the source to one 18650 battery but it doesn’t work. In addition to the battery I used a MT3608 to boost the voltage but it works only for a short time (few minutes). The battery voltage goes down very soon and the LED's light is very very low. My question is how can I use this LED with batteries? Thank you in advance for any advice.

\$\endgroup\$
2
  • \$\begingroup\$ How long do you expect the light to stay on? \$\endgroup\$
    – jaredwolff
    Oct 9, 2018 at 1:38
  • \$\begingroup\$ I have got one similar to those with a 10 cell AA holder fastened to the back using Ni–MH cells to get 12v. It is very bright. I have just used it as extra lighting for recording a video while potholing. The one I bought claimed to go up to 70W, but at 12v it was drawing less than 3A - ebay.co.uk/itm/… \$\endgroup\$
    – HandyHowie
    Oct 9, 2018 at 7:10

2 Answers 2

Reset to default
0
\$\begingroup\$

how can I use this LED with batteries?

You should use a constant current boost LED driver.

I used a MT3608 to boost the voltage

The MT3608 product you have is a constant voltage regulator which would require a current limiting resistor or CCR.

Example CCR: ON-Semi NSV50350 350 mA CCR
enter image description here

To use the boost DC-DC converter you have,
you should adjust the output voltage to just over 12V (e.g. 15V)
and add a series current limiting resistor of at least 5 Ω and 3 watts (or CCR).
Measure the voltage at the LED terminals.
Lower the voltage of the MT3608 as close as you can to the LED measured forward voltage.
If using resistor, you may want to recalculate a new value based on the forward voltage and temperature. LEDs should not have a case temperature over 50°C.

I changed the source to one 18650 battery but it doesn’t work.

A single 18650 cell should light this LED at max intensity for an hour or so if using a brand name battery. Charlatan vendors love the battery market. Test your battery with a known load (≈ 1 Amp) and monitor the voltage discharge over time.

You can compare your results to the discharge curve in this datasheet of this middle of the road Panasonic NCR18650 energy cell.

I have a Sumbulbs 12036 LED COB Strip.

Your LED has low luminous efficacy of 100 lumens / Watt.
This 100 lm/W is very likely exaggerated as it is not correctly specified by CCT and CRI.
You could get less heat, more lumens, and longer battery life with a more efficient LED.
Current technology LEDs have a luminous efficacy around 180-220 lm/W (mid-powered strips).

This 7 mm x 7 mm high power Cree XLamp XHP70.2 has more lumens, produces less heat per lumen, and is much more efficient that the LED you selected.

12.8V and current of 520mA. I kept the light turned on a few hours and everything looks ok.

At 600 mA this LED is going to get very hot. It looked ok, but how did it feel? Could you touch it without getting burned?

I mounted an aluminum panel on back to ensure a good heat dissipation.

A flat horizontal aluminum panel will only slow down the rising temperature. Unless the panel is much bigger than the LED you will end up with a very hot LED and a very hot aluminum panel.

Even with a decent heatsink, you will likely need to reduce the current to get the LED temperature under 50°C.
This will also give you longer battery life.

I would recommend a heatsink like this: 4.850" Wide Extruded Aluminum Heatsink cut to a few inches. Cost is $1.16 per inch plus shipping.

enter image description here

\$\endgroup\$
1
  • \$\begingroup\$ @@Misunderstood your detailed information really helped me a lot. I appreciate that. \$\endgroup\$
    – Camberra CacaitA
    Oct 10, 2018 at 1:10
0
\$\begingroup\$

What's happening here is that you're boosting from one cell (3.5V nominal to 12V). Due to the conversion of power you're drawing something like 1.68A (likely a bit more due to inefficiencies of the boost) at the battery. It would, with no doubt, kill your single 18650 cell in short order.

Solution? You'd likely have to put two cells in series to get a higher working voltage. But, that will only get you so far. You'd likely have to put several of those pairs in parallel to get any usefulness out of the setup.

Because this LED does draw a boat load of current, I highly suggest moving to a battery that is mean to handle high/sustained currents.

\$\endgroup\$
4
  • \$\begingroup\$ Is 600 mA a "boat load"? One cell should be sufficient for an hour of use at max intensity. Two cells in series will not improve the boost efficiency much but will make the charging more difficult/problematic. When possible, it is best to not put LI-ion in series. In this project two cells in series lacks much, if any benefit. \$\endgroup\$
    – Misunderstood
    Oct 9, 2018 at 2:04
  • \$\begingroup\$ Power in == power out. These batteries are rated to 0.5C sustained currents. If they're 2.2Ah then 0.5C == 1.1A. 1.6A > 1.1A == Boatload. ;) \$\endgroup\$
    – jaredwolff
    Oct 9, 2018 at 2:35
  • \$\begingroup\$ That's a wimpy 18650 @ 2200 mah. Look at the discharge curve for this mediocre middle of the road, 2900 mah NCR18650 "Energy Cell", 5500 mA, 2C, no problem: na.industrial.panasonic.com/sites/default/pidsa/files/… --- UR18650RX Power Cell 5C 10 Amp discharge: na.industrial.panasonic.com/sites/default/pidsa/files/… --- You can get 20-50 Amp out of Li-ion for power tools: batteryuniversity.com/learn/article/… \$\endgroup\$
    – Misunderstood
    Oct 9, 2018 at 4:11
  • \$\begingroup\$ This LED would likely get too hot to run this 10 W LED at 600 mA. I would estimate about 200-300 mA max, without substantial thermal management. An aluminum plate is not substantial unless very large and thick (i.e. expensive). Wired in series and parallel: learn.adafruit.com/li-ion-and-lipoly-batteries/… \$\endgroup\$
    – Misunderstood
    Oct 9, 2018 at 4:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.