Why LED no longer bright?

[cykoed]

Hi,

I brough SKU 7999 (Romisen RC-F4 GITD Cree Flashlight Gray (3V~8V) ) from DX and was extremely happy with it in the first 3 days. :thumbsup:

BUT....the light suddenly became dim. Thinking it was the RCRs, I recharged them - no change. Swapped to used CR123s, and even new CR123s - no change. Took off the tailcap and used a wire to 'direct drive' to see whether it was a problem with the tailcap...sigh...still dim!! :sigh::sigh:

• What could have caused the LED to become dim?
• Anything I can do to fix it?

Thanks in advance!

 

[mdocod]

my first inkling on cheaper lights is that either you have a bad solder or connection somewhere, or the output of the driver is somehow making contact with metal it isn't supposed to and current is flowing somewhere it isn't supposed to be. I've had this happen on my Nuwai LuxV light due to a sortof poor design on the retaining ring that holds the star down, I made basically a custom plastic washer to fix the problem.

 

[matrixshaman]

You can try cleaning the contacts with Deoxit or isopropyl alcohol. You might also make sure the light engine is screwed down tight in the head (if it has a couple smalll opposing holes or notches). Make sure everything else is screwed together tight.

 

[cykoed]

if i use RCRs which when fully charged pump out 4.2V each, meaning that 2 RCRS will exceed the rated 8V of the torch by 0.4V...what can i do to cap the RCRS to below 4V? will 3/4 charging prevent each RCR from extending past 4V?

 

[carrot]

You may have burnt out the LED.

Li-ions hot off the charger will be 4.2v. If you let them rest for awhile they will be 3.7v.

 

[MrGman]

you probably burnt out a component in the driver and the LED is fine. If they give a voltage range rating for the driver, you should not exceed it.

Here is a test. Run it with lower voltage, like only 1 3.7V rechargeable battery and use a probe to reach down into the tube and make contact and see if it is brighter than with 6V or 8 plus volts worth of batteries. If it is you definitely damaged the driver circuit. If not, that only means we still don't know.

Under voltage on the driver means direct drive and can't hurt. Over voltage to the driver can damage it, plain and simple, whether you want to believe it or not. Its going to do its best to keep the power to the LED constant and it has to take up the entire difference. That means its trying to hold the LED at whatever current it was set to and the voltage drop that creates. Lets say it was 750mA and that puts it at 3.7V (roughly), at the LED.

Your freshly charged batteries started out at 8.4V. But lets say they sag a little under load but only down to 8.2V because they are in good condition and as we said freshly charged. 3.7V at 750mA at the LED is 2.775 watts and lets not round that off yet If the driver was 100 percent efficient at 8.2V that's 2.775W/8.2V = 338mA to the batteries. That's not much of a load to good batteries. That would also be regardless of efficiency, 4.5V differential on the driver. The driver is not 100 percent efficient. And actually the higher the voltage the less efficient it is for the conversion. So lets say 85% efficiency. Thats actually 3.265 watts for the batteries to deliver to get 2.775 watts at the LED. (So the actual battery draw might be 3.265W/8.2V or 398mA at the batteries, still not much of a draw on a fresh charge. So thats 3.265 watts minus 2.775 watts = 0.49W for the driver to dissipate. Assuming that its within its regulation range and working correctly. The driver board is not that well thermally coupled out to pill as the LED is. Those surface mount component chips in there are tiny. They are cheap components. The pulse width modulator or whatever the actual controlling chip that makes this all happen may be one that was designed to work with a very limited safe operating area of voltage-current that you exceeded. It may have had an actual maximum rating of 0.5w for its own allowed power dissipation that with your very fresh batteries you took over at turn on.

If its not total power of the controller chip than it could be the pulse by pulse peak voltage or charge current that goes up with higher voltage differential to the driver.

May have simply blown a capacitor that sees the pulse voltage to build a charge coming off of the coil as I found in one of my units.

These things have very limited ratings because they are coming with low grade commercial surface mount components with zero burn in and testing is probably 1 turn on with half dead batteries. the Cree LEDs are probably very well tested and good components, but the electronics are all very low budget commercial Chinese parts. At least the vendor's are being honest when they tell you the small voltage rating not to exceed. So if you push your luck, don't be surprised by the results.

I took my dead driver apart, bypassed it made a direct drive out of that LED and it puts out up to 250 lumens at turn on (briefly) as tested. The LED is very robust, compared to the driver components. I just ordered 17670 batteries to run it direct drive and will keep it just as it is.

 

[cykoed]

Mr Gman,

Thanks very much for your indepth reply!

How can I bypass the driver to make it direct drive?

 

[MrGman]

The indepth analysis to tell you how badly you screwed up is free. The simple instructions to move one wire and make it all better will cost you plenty,

Send LED pills and Pizza to I can never have enough.com

Okay I am kidding. In most cases (but not necessarily all), the only thing that needs to change is the red wire. Cut the red wire and stop everything from blowing up, with 1 second left on the timer. Okay I am really kidding. The Red wire that goes to the LED simply gets moved to the same point where the center terminal for the battery comes through the board. This can usually be found with an ohmmeter measuring the positive terminal spring for the battery (it only touches the one so its not for the batteries per se), and then probing the other side of the board where all the chips are and finding where its a dead short. You may actually see the via that comes through the board. You would move the red wire to that point. The black wire for the LED should already be in common with the outer circumference of the pill or the connection to the battery tube that goes to the negative end of the other battery. You can check that with an ohmmeter as well. That is all I had to do with mine. It may help to desolder some of the components and simply get them off the board, but since they are being bypassed they should do nothing. I desoldered and threw away the burnt cap and a diode I found in the middle of the board and left them off just to be sure. The negative wire on mine for the LED was already in common with that for the negative battery terminal.

I however, have never seen your unit's pill or driver card, so I cannot say that this is "the" method to fix your's, but that is generally how it works. The driver circuit is in series with the positive side of the battery to the positive side of the LED. It needs to have a common ground to the negative terminal of the battery and LED in order to power itself up and work right (yet another reason I forgot to mention about why to not exceed the voltage they tell you not to exceed, because parts of the driver circuit do see the full voltage drop of the batteries. I was being distracted at the time).

There are pictures on the intenet of what most of the pills look like for the D26 or replacments for P60's look like, It should be fairly easy to find them and move the wire.

 

[Jarl]

You may have burnt out the LED.

Li-ions hot off the charger will be 4.2v. If you let them rest for awhile they will be 3.7v.

A while in this case is several years. Li-ions don't self discharge much (less than LSD cells), and will stay at 4.2V for a very long time.

 

[cykoed]

thanks mrgman! i'm going to have to slowly and carefully decipher your advice to understand and assess whether i can get it done! might be simply easier to return for a replacement, and just stick with primaries and avoid using rcrs though :thinking:

 

[Fird]

I have caused an LED to fail in the mode described simply by overheating it. Instead of just "poof" it now draws a few ma and produces very little light.

 

[mdocod]

You may have burnt out the LED.

Li-ions hot off the charger will be 4.2v. If you let them rest for awhile they will be 3.7v.

I must also say that this is incorrect, sorry carrot

If you have a li-ion cell that is settling after a full charge to 3.7V, then about 80% of it's capacity is gone after "settling" and the cell should have been disposed of about 60% ago

My OLDEST li-ion cells, (blue label AWs, hundreds of cycles so far), come off the charger at ~4.20V at settle to about 4.1V most of time, indicating to me that they are about half-75% the way through their useful life.

Eric

 

[Lite_me]

I held back. I didn't want to bite the carrot! :laughing: