Defective rechargeable floodlight


A friend of mine has a 100W rechargeable floodlight.
It doesn’t work anymore.

A couple of weeks ago his charge port broke, and I replaced it with a new one. Insulated and sealed everything with hot glue. After this it worked for 2 days and now it stopped working completely.
Charger says battery is full and from what I measure this is correct.
When he turned it one it only flashed for a second. It has done this a couple of times but now it won’t even do that.
He got it to work 2 times by keeping the charger connected and the light stayed on. But this also doesn’t work anymore.
Because of this I figured the batteries are dead but they don’t appear to be so.

I measured voltage on the leads going to the LED to see if there was something wrong in the connection or switch. With the switch off I measure nothing, with the switch on I measured 12.1V. This leads me to believe the batteries are holding the charge and the circuit is working.
I also created a breakout cable and used the charger directly on the led. Led did not turn on.

There are no markings on the LED so I’m not sure what I’m dealing with. But searching for a bit it could be that this LED might need 30-36V, so that would mean something in the circuitry is defective and not stepping up the voltage? Either that or the LED is defective would be my guess.

I’m not that into electronic repair, I know how to solder a wire and measure simple things so not sure if everything I tested is the correct way of testing and that my findings are what they should be.

Here are some pictures of the lamp, click on them for larger pictures. Maybe someone can tell me what would be the issue, and if it’s fixable for me.
He bought the lamp for I believe around 250€ so being able to repair it instead of tossing it would be nice.

Allright, in the meantime I did some more research and tests.
I found a 32v power supply and tried the LED. It worked, so LED is fine.
This also confirms the LED itself. It should be a 30-36v LED with a max of 3A.
I measured the boost converters (?) output again a couple of times but instead of the previous 12v it gives me 44v now. Seeing 36v should be the max for the led I didnt dare trying with the risk of frying it.
I also really dont understand what changed so that it gives a different output.

Its fully encased in what I believe is sillicone so it can’t be a loose contact or short.
To elimitate the batteries I ripper it out of the case, hooked it to a 12V 2.5A powersupply and got the 44v again. Tried to remove the sillicone to see if I can find markings but that’s not going so well. Can share photos if needed for the parts that I could clean.

I discharged the batteries 500mAh with my accucel 6 and after this the charger noticed it could charge again. So charger and batteries are working.

So it would seem I need a new converter? Or is the 44v output normal due to not having a load and is the problem mysteriously solved?
And if a new converter is needed, what do I need? I’ve seen things about constant current and constant voltage, not sure what I need. Seen dc-dc converters with configurable output for a few bucks local, but also for 40/50 euro. The cheap ones are rated for 15 watt so that wouldn’t be enough I guess, or is that not comparable with the LED output?

And when I find something. Do I need to encase it again like this was? Would lots of hot glue work or do I need to buy sillicone?

Hope someone is able to give me some insights.

yes it is normal for higher open voltage circuit, you need to measure it with the load,

I have a suspicion that the battery pack has bitten the dust or, at least one of the cells. Showing proper voltage without a load is completely different than testing it while pulling decent current. You may be losing all voltage when the defective/bad/junk cell is asked to do actual work.

Hooked it on the 12V power supply and on the LED and it works. Nice 30V output and bright light.
Hooked it on the batteries. Batteries give 12.4V. Switched it on. LED flashed again, voltage output on the booster drops from 30 to 12V and stays on a steady 12V.
On the battery side it still kept the 12.4V

So I guess the batteries are empty after all? I assumed seeing as the charger was 12.6V they are connected in 3S
Cant see the wiring due to the shrinkwrap around the batteries. It’s 2 rows of 6 batteries. But with 18650 range of 3.7 empty to 4.2 full, empty should be 11.1V and full 12.6. So with constant measurements above 12V I figured it should be able to power the LED.

Charger also said full batteries all the time, until I drained them. But it’s an 1A charger, and it charged for over an hour, despite draining them only 500mAh. I went to bed after posting my previous message so didn’t want to leave it charging.

And today I noticed that when I twist and turn the plug (already tried that before opening everything, but didn’t do anything back then) the charger goes from full to charging. So it looks like I got a bad connection, or messed something up replacing it. Although I can’t figure why it worked for 2 days. Plus when I replaced the connector, I also did a full charge to see if it worked.
Also replaced connector in a second light he had and that one is still working as it should.

Any insights on where my thinking process about the voltage goes wrong? Or am I missing something else?

Sorry friend, I wanted to help you earlier, but I forgot.

Maybe 3 Amps or maybe not.
All COB 100W LED’s that I’ve seen look different.

Typically, a COB 100W LED has 10 serial connected chips (32V-36V) and 10 parallel lines of 300mA each (300 x 10 = 3Amp).

Your LED has 10 serial connected chips (if 3.2V-3.6V x 10 = 32V-36V) and 5 parallel lines (if one line is 300mA x 5 = 1500mA). And maybe the power of the LED is 32V-36V x 1.5A = 48W-54W.

Let’s move on to the driver.
Drivers in LED lamps operate in “constant current” (CC) mode. Therefore, it has 44V output, this is normal. To check the driver, you need to connect it to the LED through the ammeter according to the scheme as in the picture.

And connect the battery pack directly to the driver for 3-4 seconds.
If it works, you can measure the current and find out the power of the driver and LED. (You can tell us, I’m curious).
If it does not work, you need to connect any other power source 12V 10A (As in the picture).

If it still doesn’t work, you need a new driver.

Battery Pack:
To check the battery pack you need to load a current of 4-6 Amp.
You may need to measure the capacity of the battery pack.
And I do not see the BMS Board. Perhaps it is on the underside of the battery pack.
I’m wondering what is the runtime of this spotlight on a charged battery - 30 minutes?

The problem may also be in the three-pin charging connector.

If the negative wire of the battery is connected to pin 3 and the negative wire of the Driver is connected to pin 2, contact may be lost between them.
I do not see how connected, a lot of hot glue.

You also need to check the power switch.
We are waiting for the results.

Sorry if there are a lot of mistakes and bad english.

Thanks for the help!
I’ll check in on how to measure the Amps. I’m not that into electronics so I’ll have to reread your steps a couple of times and figure out a way of putting a load on the batteries as using the converter and hooking it up to the LED doesn’t seem to work

Here is a photo of the original connector.

Not much to see again. But it is a 3 pin connector, but only 2 of the 3 pins are connected (was like that when I opened it and no other loose cables)

The red wire from battery with black wire from switch to 1 pin. The black wire from battery and white wire to converter on a second pin.
And then theres a red wire from switch to the converter.

I charged the battery some more and after half an hour it showed full again. I could twist and turn and push and tinker all I want, it keeps saying full again. Measuring a nice 12.5V

The original runtime of the battery is specced as 4 hours. According to my friend it does have a runtime of about 3-4 hours.
It was sold as 100W, but it could easily be 50W. Although it is bought locally for a hefty price, I’m sure it is just some chinese import.

I don’t have any 12V 10A power supply on hand. Will search through my junk to see if I can find anything.
With a 12V 2.25A powersupply I get a reading of 0.68A (If I did it correct)

With the battery pack, again a quick flash and nothing. Multimeter couldn’t make a reading.

Please note, all my multimeters and cables are cheap chinese stuff. Not sure how accurate they are.
My normal measurements only consist of yes this battery is still full, yes this light socket still has power on it, yes it beeps so the connection is okay :blush:

Also tried my 32V supply directly on the LED again, is an old HP printer supply with 32V and 940mA output.
Measurement 0.98A so I assume my method of testing is correct

Connect the three LI-ION 18650 (Button on top) cells in series.
But be careful, an incorrect connection can burn the driver.

If it works, you need a new battery pack; if not, you need a new driver.

Why am I interested in runtime.
If we have a 100W LED, a driver with a constant current mode (I’m not sure I need to check), and a 18650 3000 mAh cell.
With a 3s2p circuit, we get a 12.6v 6000mAh battery pack.
When consuming 100W from the battery, we get a runtime of 35-40 minutes. Therefore, I doubt that it is 100W.


Disconnected everything. Directly hooked the batterypack to the driver -> Single flash of light, voltage on battery end 12.7V, voltage on driver output: 12V

Couldn’t find a 10A power supply but found a 6A. Hooked that to the driver. LED works, drawing again 0.68A.

Freshly charged 3 18650 batteries from myself, connected them in serie -> Single flash of light, multimeter not fast enough to display any current reading.

So defective driver? But why does it work with a power supply?

Some more photo’s in case it might help.

If the driver is defective, what should I buy to replace?

How confident are you in your batteries?
Are these the ones that came with the light?
If it works on the power supply, it should work on the batteries.
Pull the pack apart and see if one cell is different than the others, when trying to power the light.
A cheap light might not have any balancing and one of the cells in the pack may be bad or way out of balance with the rest.
All the Best,

Oops, just saw that you tried some of your own cells.
No clue why it works with a power supply but not batteries.
Very odd.
PS try it with a 12v lead acid battery, like a car battery if you don’t have any smaller ones.
See if that has juice to light things up.
PPS, you might try it with 4 cells in series. Perhaps the boost driver is not quite getting enough voltage to make thing happen.

There are all sorts of adjustable boost driver on Amazon and ebay. just look for one that has enough output voltage and current.
For example, and I actually know nothing about this brand or seller.

Thanks for the input.

Tried 4 cells, it took a second, like it needed to charge but it turned on!
Tried 3 cells and only the single flash again.

Tried a 12V lead acid battery (12AH, not for a car)
Bright flash and nothing. Battery was at 13.2 volt, output of the driver was 12.75.

So can we conclude the driver is defective?
Still don’t understand why a 12V power supply is able to drive the light but 12/13V of batteries does not.

On the amazon link I see they have a 250W and a 600W. Seeing as the light is 100W, or as found out probably even just 50W. Would the 250W be enough or do I need to look for the 600W?
And if enough, would this one be good?
Input voltage: 10V-32V
Output voltage: 12V-35V
Maximum input: 16A
Maximum output: 10A, No-load stroom 25mA
output power: 100W (passive cooling in 20°C air), 150W with a fan on the heatsink(active cooling in 20°C air)
Conversio -efficiëntie: 94%

And as a follow up question. This one was encased in some sort of sillicone. Do I need to do that again? Would a load of hot glue work?

there is nothing wrong with driver, at least not in a sense that it is defective, it works off 4 cell, it just needs more input voltage for given output, you can replace the driver if you want, ot rewire cells to 4S it would be a good idea to include balance board\pcb, it will probably work some time with no balance board, and no pcb, but i would not rick it, especially with unknown brand cells.

as far as why power supply worked, probably because it has more than 12v, idk what ps you used, but most ps today will give you rated voltage under rated load, if load is lower it may give you more voltage

If it works on 4 cells and doesn’t on 3 it must be the batteries.
Probably a cheap poor quality battery pack.

The power supply I tried was 12V 6A, just measured voltage without load, 12.4V

The battery pack fully charged gives 12.7V output.
The lead acid battery measured 13.2V under load.

Rewiring the batteries doesnt sound like a good option.
The included charger wouldn’t work anymore and my friend isn’t going to know how to handle things I guess.
Don’t know if I’m confident enough either to do that.

Sorry to be absent for a long time, a lot of problems lately.
Perhaps the driver was overheated, and some components left their nominal values, so it does not work now from 12v.
To find out, you need to remove the compound and take some measurements, but it will be difficult and I’m not sure if this will help you.

It looks like in this photo we see a driver power of 35 watts.
This was the AD-DC version of the driver. The photo shows that there are no 2 diodes out of 4, and the remaining two diodes protect against reverse polarity.

And I was mistaken, I thought the battery pack consists of 6 cells (3s2p), but it looks like there are 12 cells (3s4p).

I took some measurements of boost converters.


Input voltage 8.5-40v
Output voltage 10-45v
Output power is 250 watts.
This is a convenient version of the CC CV boost converter; it is assembled on an aluminum plate. It is similar to the MCPCB for the LED and it can be glued to the back cover.
I loaded it at 35 watts, and it works well in the range 8.5-14v. But it turns on very unstably at a voltage below 8.2v and at the same time it gets very hot.
If we turn it on when the voltage on the battery is 8.5-12v, it works fine and discharges the battery to 7.5v. But if we turn it on at a voltage of 7.5-8v, it can fail.


Input voltage 10-30v
Output voltage 10-35v
Output power is 150 watts.
Good CC-CV boost converter with undervoltage protection. With an output power of 35 watts, it does not heat up, it works stably up to a voltage of 9.80v. At a voltage of 9.80v, protection is turned on, and the output power drops to 30-40%. Meanwhile, due to the low load on the battery, the voltage on it rises to 10.3-10.5v and the converter continues to discharge it to 9.80v.
But there are two problems:
1 Fine adjustment of output voltage required. The converter maintains a constant current, but if you set the voltage higher than necessary, it does not work stably.
2 It is also slightly unstable in protection mode.
from undervoltage. There is a slight tremor of light.


The boost converter in your link is identical to the previous one, but it has no current regulation, only voltage regulation.

Most BMS battery boards have undervoltage protection at 2.5v per cell or 7.5v for three cells connected in series.
The first converter can drain the battery completely, and you will have maximum run time, but turning it on below 8.5v can burn it.
The second converter has some problems, and does not completely discharge the battery, up to 9.80v (3.26v per cell), but it is more suitable for you.
The third converter, I recommend you only if you have not found anything.

some links

Thanks for the input.
There are 12 cells indeed.

Going to try the second one.
Rather have the light turn off before completely empty than to burn stuff. And since the driver is enclosed in the battery compartment cooling is an issue.
My friend is using the light while working. Don’t know if he will be attentive enough to notice the light needs to be turned off because he works with multiple lights at once.

What would be a nice voltage to set it on? Since the led runs from 30-36V. Lifetime is more important than brightest output so putting it on the lowest as possible? Or a little higher, at 32V so it has some margin of error?

With an adjustible boost, you might let heat be your guideline.
Crank it up so the light is bright enough without producing too much heat to cook things.

I wonder about the battery pack and charging.
It might be worth a look to see how the cells are aranged.
And if any cells seem out of ballance.
Looks like you are getting close to solving the puzzle.
All the Best,

I think your friend will notice that the light is discharged.
The output power of the driver drops to 40% at a voltage of 9.80V. Then the brightness continues to slowly decline and can reach up to 20 lumens. Such brightness no longer illuminates anything, and it will be clear that it needs to be charged.

A few steps to configure the driver.

The driver has 3 trim resistors on board
“CC” (Voltage Regulation)
“CV” (current control)
“UVP” (Adjustment of undervoltage protection)
First you need to rotate all adjustment resistor counterclockwise 20-25 turns, at the end you can hear clicks.
Connect the driver as shown in the picture.

Turn the “CV” adjustment resistor clockwise and adjust the voltage to 32V.
Then connect the driver as shown in the picture.

Turn the “CC” adjustment resistor clockwise and adjust the current to 1.10A.
Slowly turn the “CV” adjustment resistor counterclockwise until the current drops to 1.08A, then slowly turn clockwise until the current rises to exactly 1.10A. At this point, the adjustment is complete.
Thus, you set the power to about 35 watts.
I think it will be optimal for runtime and brightness. If the brightness is not enough for you, you can increase the power, but I do not recommend increasing it more than 50 watts.

A few points:

  1. I do not recommend using, or turn on the light while charging.
  2. Do not enclose the driver in silicone or hot glue. You need to drill 4 holes and fix the driver to the screws, or wrap it with something and fix it with silicone to the back cover as the batteries are fixed. I recommend checking the silicone seal on the back cover.
  3. In the undervoltage protection mode, the light will not behave stably (nervously), this is normal.

I hope this helps you.
Waiting for the result…