LED with no current limiting resistor?

[WillHuang]

Hello everyone,

I have a very cheap digital watch (picture 1 below) that I took apart. I thing I immediately noticed was that there wasn't any current limiting resistors on the LED (picture 1 and 2). I was hoping to get your guys' as to why this is the case? Would this not damage the microcontroller (which I am assuming is under the black blob in picture 3).

Thanks!

[theatrus]

It's quite possible the ports of that epoxy blob IC are already limited, especially if it's designed to drive LEDs directly (likely).

You can find current limited IOs in some MCU families (more so in  CortexM parts).

[Zero999]

What voltage does it run off?

[sokoloff]

It may use a constant current LED driver chip.

[james_s]

That would be my guess, limiting built into the IC. Especially for something like a watch this makes sense, PCB space is at a premium, no sense in having separate resistors if you can avoid it.

[mariush]

The batteries have so high internal resistance that it's like you always have a resistor in series with the battery. 
Basically such tiny batteries can only output a few mA at best so there's no need for current limiting resistors.

Here's for example a datasheet for the CR2032 style battery : http://data.energizer.com/pdfs/cr2032.pdf

Notice there it says Typical IR (internal resistance) and a value between 10000 mOhm and 40000mOhm .. so basically between 10 ohm and 40 ohm

So let's say we go with 25 ohm internal resistance then a red led with 1.7v forward voltage would be limited to  Vin - Vforward = I x R  => 3v - 1.7v = I x 25 ohm => I = 1.3v / 25 = 0.052 A or 52mA  ... the batteries that thing uses are much smaller so the internal resistance is much higher.

Bigger batteries have much lower internal resistance, so you'd need to limit current.

[rdl]

You see current limited LED drive circuits in lots of things. PC motherboards do this for the front panel LEDs such as power and HDD activity. You can also buy standard LEDs with the current limiting resistor built in.

[WillHuang]

These watches run on a pair of AG10 button cells that give ~3V total. I couldn't find a datasheet on these batteries, so I don't have any information on it's internal resistance.

[mtdoc]

In general, current limiting resistors are not required when powering LEDs with button or coin batteries at 3V due to their internal resistance.  There are tons of "maker" blinky type projects out there that take advantage of this.  Google "LED throwies" or "paper circuits" for example.  I've recently been teaching paper circuits to a bunch of 4th and 5th graders and they have yet to see or learn about resistors. LEDs, copper tape, and coin batteries are all they use.

[james_s]

It works ok with one LED, but if you try to run multiple LEDs you're going to have problems with some of them hogging all the current, unless you have very well matched LEDs.

[mariush]

These watches run on a pair of AG10 button cells that give ~3V total. I couldn't find a datasheet on these batteries, so I don't have any information on it's internal resistance.

Just go to a store that sells electronic components including batteries and you'll find plenty of datasheets there.

For example, here's Farnell's battery selection : http://uk.farnell.com/c/batteries-chargers/batteries-non-rechargeable

LR = alkaline , SR = silver oxide ... or see https://en.wikipedia.org/wiki/List_of_battery_sizes  .. if you search for AG10 you'll find they're usually listed under following codes:


So here's three of them : http://uk.farnell.com/webapp/wcs/stores/servlet/Search?catalogId=15001&langId=44&storeId=10151&categoryId=700000004383&pageSize=25&showResults=true&pf=110119509,110130045



and here's the datasheets for those batteries (Annsman , Varta and Renata ) :

http://www.farnell.com/datasheets/89170.pdf
http://www.farnell.com/datasheets/856378.pdf
http://www.farnell.com/datasheets/1497204.pdf

The varta one is rated for 5.6kOhm load continuously ... so 1.5v / 5600 ohm = ~ 0.00026 A  So basically even with 0.5mA you're pushing them.

[mtdoc]

It works ok with one LED, but if you try to run multiple LEDs you're going to have problems with some of them hogging all the current, unless you have very well matched LEDs.

For the paper circuits, multiple LEDs work fine as long as unequal brightness is not an issue - which at this point it is not (and it may not be in a cheap watch either). Initially it's about building very simple circuits, learning the difference between series and parallel, and combining art work with electronics.  Once they've got that down - then, hopefully they will ask "why are they different brightness?" or "why can I only get x number of LEDs to light up with one battery?".  Hopefully that curiosity then leads them to a place where learning more about voltage, resistance, and current is easy and painless.  The approach to this elementary school "maker space" is hands on discovery and learning by experimentation and failure.

[james_s]

I remember getting some LEDs when I was about 6, connecting them to a 9V battery and wondering why they got so hot and didn't last long. Some of them would turn interesting colors and at least once I had one explode.

[BrianHG]

I remember getting some LEDs when I was about 6, connecting them to a 9V battery and wondering why they got so hot and didn't last long. Some of them would turn interesting colors and at least once I had one explode.

You are not alone.
I had a red LED 7 segment display actually shine purple then blue back in 1976, though it burned out in around 10 seconds.

[GeorgeOfTheJungle]

Perhaps the leds are being pwm'd?

[danadak]

Another possibility is the sizing of the output NMOS/PMOS transistors greatly
affects Rdson and ability to drive current. Especially true in custom die designs.

That coupled with earlier comments on battery equivalent resistance, and ability
of wide range in CMOS operating Vdd, could explain the lack of Rleds.

Note if driving the LED high onchip crude current source can be realized by
circuit design.


Regards, Dana.

[Audioguru]

The light from the display is white and the LEDs have a yellowish phosphor so they are 3.2V white LEDs. The two battery cells will produce only 2V when used for a while so the circuit must boost the voltage.
The current will be very high for the tiny button battery cells so either the display is dim and only momentary or the battery will need to be replaced often if you use the display frequently. 

[Zero999]

I can't see any inductors or capacitors (for a boost or charge pump) on the board.

The forward voltage of the LEDs depends on the current. At low currents, the voltage can be as low as a couple of volts. My nephews have some glow balloons which run off two 1.5V button cells and have white LEDs. There is no boost converter, just a the LEDs and batteries.

I've done a quick search and there are LEDs specified with a V_F of under 3V, at low currents, see link below for one with a V_F of 2.9V @5mA.
http://docs-europe.electrocomponents.com/webdocs/14a6/0900766b814a61b8.pdf

[Audioguru]

But the battery is alkaline and its voltage continuously drops as it is used.

[Zero999]

But the battery is alkaline and its voltage continuously drops as it is used.

The same is true for my nephew's glow balloons but it's not as bad as you expect. The LED is bright enough to well light the balloon for around 8 hours, a dim light for a few days and the LED just being bright enough to see in a dark room with the dark adapted eye for a few months.