Strange Flashlight performance with Low battery

[kosPap]

note to mederators: i am placing it here cos it is more relevant

I was sent an Ultrafire A30B flashlight (Akoray Clone) to troubleshoot.

the owner said that after a while it acts up and dims or gets off when taped/moved.

I tried both protected and unprotected batteries and it does act up, but every time and ONLy when a battery goes low, like 3.8V.
Bang it forward or backwards and it does it. Add a charged battery and you can kick it around.

I have already excludee bad contact causes. Fully dissasembled and retightened everything, added more solder on the switch PCB ring (it had only one on a side), added a contact spring on the driver PCB, and the batteries tested with were of various lengths.

It is definetely an electrical issue, but what?

 

[VidPro]

You did everything a person could think of. test the switch?

 

[THE_dAY]

That light has an XRE and driven @1000mA, the XRE has ~3.7Vf @1000mA.
Maybe the light has a linear driver where the light will drop out of regulation when the battery's voltage under load is below the LED's Vf + driver overhead which would be ~3.8V.

 

[kosPap]

yes but why the light either goes very dim or even goes out completely then when taped it will go on to full power?

Since I added a spring on the driver I noticed that it behaves normally with UNprotected cells which are shorter.
But no luck with the protected ones.

Somehow the added length pushes into something (rear switch?) and crates bad contact or parasitic resistance.
I plan on adding more sodler on the switch PCB ring to take any possible slacl and the chance of backing out of contact.

It does have a linear driver (justs PWMs the output) but can parasitic resistance make the difference? Say create a 0.5 volt drop?

 

[VidPro]

"" I noticed that it behaves normally with UNprotected cells which are shorter.
But no luck with the protected ones.""

Originally you said both of them, so now it is only the protected?
DX indicates it is boost with 2.5-4.0 style input and 1 x 3.0V/3.6V CR123A battery.
Then http://club.dx.com/forums/forums.dx/threadid.340247 <-- this (shows no driver at all :-) . as usual it might be different for different peoples. Ya never know what you gonna get from many of the china sellers, they might not have the same exact one to send you, so they send whatever.

The one you have is a linear? you got it apart, mabey a picture would help?
also please define Exactly what you mean by "acting-up" , what does it act like other than the tapping on it changing it?

This (from the link above)
""Now that I've had mine for a while and run it I have discovered what appear to be regulator problems with RCR123A batteries. After a few minutes of runtime the light flickers and/or dims. Turning off and on may reset it, but it will eventually start all over. I tested rechargeables on a second unit and it did the same thing. The batteries work fine in all my other lights, just not this one. This happens with all 6 of my rechargeable batteries, no matter the charge level.""

^ that Is probably the boost curcuit getting fried, from using the higher voltage battery on what probably was a boost curcuit, going "direct" sort of and cooking it. so at some time in its history 2009 it probably was boost?

 

[kosPap]

sorry for taking long to reply but I spent soemdays feeding it different batteries and monitoring it.

Since i installed a spring on the driver PCB it practically stoped acting up. That means stepping way down when the battery is low, then coming back up again when tapped.
Of the batteries used most were AW protected ones. And one of them will make the light "act up" veen when full.
So probably it is a shot out battery.

I also tried a CR123 and the light is lower than RCR as expected for such drivers.

You make a good point on the boost part being damaged and workig intermittenly.

Unfortunately I do not know if I should remove the driver.

 

[VidPro]

So now you have springs on both sides? Are you beginning to run out of space in there :-)

it also could be the battery protection kicking in, IF it is a "no-driver" kind of PWM 3-5 mode thing. If that was occuring in mere miliseconds the power returns. You would have to figure out if that is what it is acting like.

 

[kosPap]

the flashlight seems trouble free now.
the one battery that gave trouble turned out to have a faulty protection PCB one.
Now i will ahve to wait one the owner input