Defective rechargeable floodlight

Yes you are right, the driver works in protection mode.
Turn the “CC”, “UVP” resistor fully counterclockwise until it clicks. Connect to a 12V 6A power source or lead battery and try increasing the output current again.
And do not short-circuit the driver output.
With an input voltage of 12V and an output power of 35 watts, the input current of the driver is about 3.20A. This should not be a problem for a good lead battery or 12V 6A power supply.
The minimum value for undervoltage protection is 9.80V, you cannot set it lower, but you can set it higher using the “CV” resistor.

To measure the cells in the battery you need to remove the blue film. In this case, it is not necessary to tear off the battery from the cover, you can cut the film gently and remove the cells.

It’s strange.
I began to doubt, maybe the old driver is working.

Just tried again. Same results.

12V 6A power supply.
Dialed everything CCW until clicks.
Connected PSU, ramped up voltage to 32V, measured voltage at input 12.1V.
Connected light. Slightly turned up CC, lights goes on at 0.4A, immediately voltage on input side drops to 9.8V.

Am I unlucky? Do I have ordered a defective driver? Should I just give up :confounded:

This is really weird.
Can you show the photo:
new driver (on both sides),
connection photo
12V power supply
Lead battery.

It’s too early to give up

@RD, your English is perfect. Most of us don’t speak 2 or more languages, so you are doing great.

@johnny, you have done a great job at troubleshooting. Considering all the strange results i would recommend a simple approach to start at the input and work to the output.

0. Remove all the goo and glue around the connectors and switch and any other place in which the wiring solder joints are covered. Inspect and touch up any cold or broken solder joint. Replace wires if necessary to ensure good path and eliminate the wiring or joints as a culprit.

1. Slice the blue film so you can verify the configuration of the cells, to measure and compare the individual cell voltages (looking for laggards), to load test and check capacity of individual cells, and to look for some sort of BMS board for the cells.

2. Rebuild the pack (replace defective cells) and BMS as necessary to eliminate that as a culprit of the problems.

3. Charge the pack and measure/verify the voltage and current while charging, if possible. Measure the pack voltage after charging to verify that the charger is working properly, no over or under charge, cuts off at limit, etc. Eliminate the charger as a culprit. [and like RD said, don’t try to run any load from a charger, not designed to work that way and possible to damage something.]

4. Test the driver and LED with this known-to-be-good pack. If the results are flaky with good solder joints, wiring and pack, then the driver is suspected as a culprit. Repair or replace.

5. Now the light has been rebuilt from front to end, and should give good service, especially considering the price.

Here is just an example of some various drivers: open boards, enclosed, and waterproof.

Good Luck and Enjoy.

The wiring looks like a mess but I know where everything is going. Don’t have any better way of attaching everything and being able to switch to the different parts and power supplies easily.
The 12V lead acid photo will have to wait if necessary. Put it back in the storage downstairs of my appartment building. It’s an 12V 12AH battery used in a mobility scooter. I believe from powercell.

@Kenny
Right now I’m not dealing with any solder joints. I have taken everything apart as far as possible to have quicker and more room to work with.
In this I have fully eliminated the switch and charge port.

Don’t know if I am able to rebuild the battery pack. I can try and slice it open, but there is not much room on one side to measure things. Sure can’t measure the bottom row without touching the top row without taking the whole battery pack apart.
But since it’s not working properly with the power supply I think I can hold of messing with the batteries until that issue is resolved?

Your mistake is that you use wires that are too long on the driver input.
Connect the driver cable directly to the connector as shown in the picture, and try to adjust the driver again.

The battery in your current meter is indicating low and should be replaced.

The red lead on the current meter is cracked and broken right at the strain relief. This could be a high resistance junction or go open intermittently causing additional issues.

I forgot to add, the driver output wires to the LED should be as short as possible. I recommend to remove the factory probes (wires) of the multimeter, and connect with short good wires as shown in the photo.

I see some soldering defect, I recommend cleaning it.

A photo of a lead battery is not necessary, if it is used for a scooter it can give a lot of current. I wanted to see why this did not help you.

And for the future.
When measuring current, it is recommended to use as short wires as possible and short connections. Otherwise, you may get a big measurement error. Factory probes (wires) of the multimeter are not always suitable for measuring high currents
For a better current measurement, I recommend buying a “banana” plug and soldering short thick wires to them.

It will also allow you to measure high currents of flashlights.

And as @ Kenny said it’s better to replace the multimeter’s battery later. With a low charge, the multimeter may give incorrect values.

I have a question. How did you connect the old driver? If you connected it through long wires, maybe check it again?

Oops. I figured the quality of my test materials would influence the readings, all cheap chinese stuff. Didn’t figure it could have this much impact.
I do have a lot thicker wire, even thicker than the regular pieces of wire I stripped out of a 230V power cord, and a LOT thicker than the wires inside the alligator clamp leads.
I’ll make some better equipment.
I have some banana plugs lying around used for a stereo. It’s also cheap chinese stuff. Would it work and be good enough if I soldered the thick wire to it? Otherwise I will order real multimeter plugs


Also found these, if it works these might be better quality?

The second ones look good. They also have a screw, that can fast connect the wire without soldering. You can try using thick wires with a length of about 300mm. In any case, it will be better than the factory probes.
But if the covers are metal, you should be careful with high voltage measurements.

Alright, this is my new test equipment. Will need to redo it. The alligator clamps lost their clamping power. Probably have overheated it. Also replaced battery in multimeter.

Also replaced the wires going to the LED with this thick wire. Might try to rewire the battery pack but don’t know if I’m confident enough to do that.

With this new short leads, the light works! Can amp it up a lot more. Way brighter. Only thing I still noticed is the output seems to hang at 28V, might be my multimeters? But tried both.
With the 12V 6A power supply the light has an awful amount of flickering. Almost nauseating. This does not happen with the battery pack attached.
Also tried the old driver. It works………Draws 0.68A, output of 28.8V, nothing to adjust here. Didn’t replace any more wires here so still the very flimsy thin wires on the output.
So…. Don’t know what to think of it now. Battery pack is the culprit? And with the newer thicker wires it is able to draw more and keep up?

Good. It looks like the Alligators are a bit overheated.
Voltage 32V-36V for COB LED is for reference only. It does not matter what voltage you get on the LED, it matters what current you give to the LED. It’s hard to explain. You can see the electrical characteristics of the CREE XML2 LED, other LEDs behave in a similar way.

What current did you install on the driver output, and to which battery did you connect the driver?

Johndoe

What power supply are you using? is it AC or DC? Ac will result in a flicker of 25 Hz of course

It’s an AC/DC adapter.

I stopped at 1A because I did not know if I could trust the voltage output and didn’t want to break anything.
I connected it to it’s original battery pack. So in this result with the old driver everything essentially is connected as it was in the first place, but with some better cables and without switch and charge port.

Next step is to hook it up to its original connectors again, with the added better wiring as far as possible. Maybe my out of the device testing with the bad long wires has thrown all issues to another point and the switch is defective in the original issue.

You have done a lot of work with this light.
The old driver gives out only 20W of power. Original wires may be enough, but if you want you can change.
If the old driver will work and only the switch is faulty, you have a few steps left.

- Fully charge the battery, and check the runtime.

  • Compare this light with others that your friend has.

But we learned that this light was sold as 100W, it looks like 50W, and works on 20W. Marketing.

For now it looks like the switch is the culprit.
Wired everything in as it was supposed to be, nothing. Measured output of the battery with the switch on, no driver, 1.3V.
Replaced the switch with an old inline light switch. Working. So the switch is having a lot of resistance I suppose?
What could be the issue? Water? Light should have an IP65 rating, switch had a rubber cover and couldn’t see waterstains inside that, but it get’s wet all the time.
Tried to resolder the wires on the switch to see if that would work, but probably went too hot. Prongs got too hot, melted the plastic and shifted out of place, doesn’t work at all anymore :person_facepalming: You see why I’m reluctant at messing with the batteries :smiley:

But it doesn’t match my first results, getting 12.1V output with the switch on.
I guess the old driver was able to step the voltage up to 12V with the little bit of power it got?

Right now I’m running it, put it at 1.1A, will see how long it will run, how hot everything gets, if the Amps stay within range.
If I’m satisfied will have to order a new switch, and will have too see if I can waterproof it again.

So, if the switch is the culprit, my stupidity of not understanding the outcome of the values, and using wrong test equipment made everything a lot worse. OOPS.
At least I’ve learned a lot from this thanks to everyone.

EDIT:
Hmm, half an hour in and the light start’s to flicker more and more and the amps are ramping up, dialing it back lowers it again but the flickering stays. You can hear the driver buzz. The indicator light on the driver has some red coming through and if I pick up the driver and twist it a bit the indicator light goes to full red and the flicker stops.
Voltage from the batteries coming in was 10.5V, output voltage 30.1V.

I stopped the test, it was going to 1.4A pretty fast. Right now I’m running it from its original driver, see what that does. Right now the same 0.68A as before.

You need to solder the switch quickly, otherwise the plastic will melt quickly.
In the first results, you got 12V at the switch output, but it looks like the contact had a high resistance and the driver could not start.
Unstable operation of the driver can be slightly reduced by tuning it as I said in post # 20.
It seems that the battery is not in good shape.
I recommend using the old driver. This will give you less power (20 watts), longer runtime, more stable operation at low voltage, and it will possibly drain the battery completely, which will increase the runtime even further.

I have order a new switch.
Locally I can find them only for high prices (mostly due to 4-7€ shipping costs) or not the correct measurements.
From china they are cheap but don’t know when they will arrive so found a seller in france that sold them with shipping for an acceptable price.

Will update this thread again when everything is back in place.

btw those clip leads are usually garbage.
wire too thin and it is simply folded back and crimped to the clip.
they are often bad right out of the package.
as you found out unreliable tools=wild goose chase.
as for the original driver unless it is blown up the only thing i see fail are the input and output caps.
seen plenty of those cheap switches go bad.
they dont stand up to inrush current.
the input cap gives plenty of that if it is good.

A little late but here’s the update.
Despite ordering from within europe it took a while for the new switch to be delivered.
And when I got it I didn’t find the time to work on it again.

But I have installed the new switch and did 2 rundowns and everything looks to work as it should!

I’m going to deliver the light on monday and see what my friend thinks of it.
If he thinks it works and is as bright as always, now we know they mislead him.

In the end a lot of work for what looks like a defective switch, but at least I’ve learned a lot!

Thanks to everybody that contributed to help! Very much appreciated!