I have a No-name 17-LED three D-Cell flashlight and only 7 of the LEDs are
lit, and another one goes on and off. I tried new batteries but no change.
I assume the flashlight is kaput. Anything I may be overlooking I could try
before I replace it? Screw-in parts like front end are tight.
You know it's time to clean the refrigerator
when something closes the door from the inside.
Are you sure there aren't two "settings" available:
- 7 LED
- 17 LED
Is there a pattern to the LED's that are failing to illuminate?
I.e., left side, right side, outer ring, inner core, etc.
Is the LED assembly accessible (to inspect)? Or, is it a sealed
assembly (discard when broken)?
On Wednesday, October 21, 2015 at 2:00:09 PM UTC-4, KenK wrote:
I'm not saying you are wrong, just that an "on/off push button" can do more
than just turn the device "on/off". It can indeed control 2 modes.
I bought some no-name LED lights with a single "on/off push button".
One push turns a set of 15(?) LEDs on the face of the unit as a work light.
The next push turns the work light off.
The next push turns on 5(?) LEDs on the end of the unit as a flash light.
The next push turns the flash light off.
Wash, rinse, repeat.
I wanted to use them in a linen closet, with one attached to the underside
of each shelf to provide light for the shelf below, but I didn't want to
have to push the button 3 times to get to the work light function.
I opened each one up and moved a wire so the switch is now an "on/off push
button" that controls only the 15 LED's on the face. Internally, the switch
is still toggling between the 2 modes (flash light and work light) but
externally, the work light just turns on or off with each push.
Maybe yours is supposed to be a 2 function device but the switch is screwed
up and not making the correct contact any more.
Often, the button is multipurpose: press once for "low", again
for "high" a third time for "off". Lather, rinse, repeat.
Or, other variations on that theme:
- 7 LED's (low)
- the other 10 LED's (medium)
- all 17 LED's (high)
Note that if the connection from the switch (or its internal
mechanism) to those "other 10 LEDs" is defective, the behavior
would end up as:
- 7 LED's (low)
- the other DISCONNECTED 10 LED's (off)
- all 17 LED's -- of which 10 are disconnected (low)
"Can't tell the players without a scorecard"
And, in truth, the arrangement of the 17 LEDs is probably somewhat
"random" -- it's not like a nice square grid of 4x4 LED's, etc. :>
Apologies if I'm oversimplifying in what follows -- or, not
simplifying enough! And, freely mixing metaphors/analogies... :<
Internally, a 3C (alkaline) flashlight looks like a 4.5V battery
feeding a bunch of LEDs through a switch. (duh!) The battery
represents the available "pressure" in the system. The switch
is a valve (thinking in terms of a water analogy, here).
LED's are current (flow rate) driven -- the amount of electricity
flowing THROUGH the device determines how brightly it lights. Sort
of like the rate of flow over a water wheel determines how quickly
it will rotate.
But, there is a voltage (pressure) component as well. E.g., unless
you can get the water up to the *top* of that water wheel to start
with, doesn't matter how much of it is flowing as it won't be flowing
*over* the water wheel!
LED's typically start to turn on around 2V. But, this voltage varies
depending on how much current is flowing *through* them. As current
increases, voltage "drop" across the device also increases. So, at
a low current, it might "drop" 2V but 2.2V at a higher current, etc.
(at too high a current it simply burns out)
We have 4.5V available and need to LIMIT this to something around 2V,
based on what the LED's actual needs (at a given current) are.
So, there is a ballast resistor in series with the LED's (and the switch)
that "soaks up" (drops) some of this available pressure (voltage).
The pressure (voltage) drop across the resistor varies with the flow
rate (current) through it. So, as the LED wants more current, the
pressure (voltage) dropped by the resistor increases -- which cause
the LED to need less current (because it has less AVAILABLE pressure),
etc. Hence the term "ballast".
[Hopefully that makes *some* sense :< Sorry, it's too early in the morning
for me to be thinking in these terms! :> ]
Now, with 17 LED's, if you stacked them end to end and EACH ONE needed
~2V, you would need a > 34V battery to cause *any* of them to illuminate.
So, instead of wiring them in series, they are wired in parallel. I.e.,
all 17 (or, perhaps a group of 7 and another group of 10?) side-by-side.
So, they *share* the flow (current) from the resistor.
But, not all LEDs are created equally (manufacturing variations called
"process variations"). So, some LEDs might START to light at 2.0V while
others don't start until 2.1. Some might get brighter much more quickly
with increases in available current while others less so. Etc.
It's sort of like putting a bunch of batteries side by side in a charger
and hoping that they all charge at the same rate and to the same final
state. In practice, some batteries will be lower than others and may
need more charge, etc.
[this is a bad analogy as batteries in parallel can charge each *other*;
LED's in parallel can't *light* each other!]
The ideal solution is to select specific LEDs for EACH FLASHLIGHT such that
they are identical to each other. Then, they will "share fairly" because
their individual characteristics will track regardless of operating conditions
Ain't gonna happen! :>
A more practical solution would be to associate a single "ballast" resistor
with *EACH* LED. Then, tweek the value of that resistor to compensate
for the specific characteristics of the LED that it supplies! But,
that adds lots of "unnecessary" resistors for the sole purpose of
getting the same amount of light out of each LED. Does the (price driven)
user really care if LED #8 emits less light than LED #12? The user is
just interested in the TOTAL light out of the flashlight -- regardless of
which LEDs happen to be carrying the load!
Take thisi a step further and you end up with the "consumer solution" -- put
all the LEDs in parallel with each other and use *one* resistor to save
the 16 micropennies that the other 16 LEDs would have cost. And, live with
That gives you an idea as to why there is LED-to-LED variation.
The fact that some are dark suggests a broken foil (if all were supposed to
be on one circuit -- not high/medium/low) or a wire. Or, a bad switch as
described above. Or, a different configuration (perhaps TWO resistors
used: one for this group of 7 and another for a group of 10).
Individual LEDs "flickering" may be caused by a poor ("cold") solder
joint and/or mechanical flexing of the circuit board.
On a lot of the cheap "no name" stuff they don't even use a ballast
resistor. They simply wire pairs of LEDs in series, and then parallel
them. Depending on the LED forward voltage, they may wire triplets in
series instead of pairs.. So if one LED fails, 2 or 3 go out. If one
shorts, another 1 or 2 get over-bright and shortly after also fail.
The GOOD ones will use current regulator chips for each LED, and will
continue to put out consistent light until the battery is totally dead
(or at least until the battery voltage drops below the forward voltage
of the LED. There are also "regulators" that will run a 3.5V LED off a
1.2 volt battery, at full brightness. These are called buck/boost
On 10/21/2015 2:13 PM, email@example.com wrote:
A typical way is to use the internal series resistance of the battery
to limit the current. Works great if you use crap "heavy-duty" AAA
By the time you get up to D-Cells, that ain't
I modified one of the free Harbor Freight 9-led 3-AAA lights to accept
one 18650 lithium ion battery to see how long it would last.
4.2V is less than 4.5V, right?
Well, it was quite bright for a while. Even when used less than a
minute at a time, the lights began to flicker and die. There's zero
I did another one with 1.5 ohms in series. That one is still going
Cross-threaded ends would not make some LEDs work and others not. The
LEDs in a 17 LED array At least in most "cheap" arrays) are wired in a
series-parallel matrix, and if one fails open, all in the series
section shared by that LED will fail. If one LED fails shorted, the
other LEDS in that series string will light brighter - and will
eventially fail from overcurrent.
I have had really bad luck with some Chinese LED lamps failing both
Glad to know that I'm not the only one who destructs broken stuff
before tossing it !
I like the term " educational destructive autopsy " thanks !
I agree - poor solder connections are common
< on the little 3 x AAA 9 / 21 LED flashlights >
The other common problem was a bad battery -
that didn't test bad at first .. but putting in 3 good brand-name
batteries - made the difference ..
Not often the switch - unless it was obviously broken.
There is probably a break in the printed circuit board they are mounted
to. If you got nothing better to do, remove the board and start looking
with a magnifying glass, or jumper them if you know the basic pattern of
the board. It's easier to just buy another flashlight, but I too have
torn stuff like that apart and sometimes can fix it.
A local rock band has all LED colored stage lights. Each color has an
individual controller. There are hundreds of LEDs in each light. One of
the lights has a bad segment, whereas about 10% of the lights dont light
up. All of them are in one section, and all colors are affected. That
tells me that the neutral is not going to that section. Probably a break
on the board. I know the guys in the band and asked the guy that runs
the lights about it. He said he dont work on that stuff, and was told it
would be expensive to repair by the store that sells that stuff. He said
that eventually he will just replace that light if it gets worse. But he
said they are costly, so he will just uses it as-is, until it gets real
I agree. Probably a cracked ciruit board or cold-solder joint. Or shorted
ot open LED as someone else suggested. I'm just trying to decide if it's
worth the time to tear the bulb assembly all apart. I'm retired so I have
more time than money, but I'm incredibly lazy.
You know it's time to clean the refrigerator
when something closes the door from the inside.
Get off your lazy ass and consider it an adventure!
If you can get the board out, it should not be that hard to find the
crack or open circuit board trace or had solder joint. An open LED will
only affect one LED. A short will not allow ANY of the LEDs to light.
On Thu, 22 Oct 2015 14:05:04 -0500, firstname.lastname@example.org wrote:
An open LED will affect all LEDs in the series string, if they are
wired series parallel - which virtually all COB (Chip On Board) Arrays
are. If one is shorted it will over-drive the rest of the LEDs in that
string, causing them to fail in short order.
HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.