I had several 2032's that were past their usefulness. I wanted to toss them into my recycle bin but didn't want a fire. So I decided to throw the 2032's and LED's together. No resistors, just let the batteries power the LED's directly. One 2032 and one LED. These already weakened batteries lit those LED's for weeks. As their voltage diminished the brightness went down.

As others have already said - you don't need a resistor with an LED when powered from a single CR2032. Now, I can't speak to three 2032's in series. Still, it's likely you won't need a resistor in circuit. But your assumption is correct that it's the correct approach to have a single LED with a single resistor when designing multiple lit LED's. Any time your power source can exceed 20mA - you're going to need a resistor to drop some voltage. Remember, current is the dividend of voltage divided by resistance. ( I = E/R).

But you don't need to concern yourself with the 2032's.

 

Just tried a fresh CR2032 with a typical white 20mA LED. The battery measured 3.2V no load and dropped to 2.9V when lighting the LED dimly.

Bob

 

Just went went back and checked the current. Same setup, fresh CR2032, no resistor. 5 mA.

Using a power supply:

It reaches 20 mA at 3.3V
It is first visible at 2.3V drawing 1uA.

Bob

 

Those solar-powered garden lights have a jewel thief and charging circuitry under the blob.

The TS concern with space makes using a joule thief impractical. A simple approach would be to use another battery and a dropping resistor. But I make no claims as to it being the simplest.

 

Some of my solar garden lights have a blob on board IC but most use a QX5252F IC.
The circuit is simple and is small:

 

I just tested (using the pinch method) a standard 20mA white LED rated 3 to 3.6 forward voltage with a 2032 battery (fresh) and it was very bright.

I have no idea how long it would stay that way, because I need to keep the battery for my use.

That was a TopBright brand LED.

PS: I wasn't trying to prove anything...just curious to see if a random white LED would light.

I think I am going to use a tiny 5v booster and then put a 100ohm resistor on the LEDs just to be sure.

 

That is a good idea if minimizing battery life is the goal.

Bob

 

if minimizing battery life is the goal.

Is it ever?

 

Is it ever?

Only for those with a long position in the battery company. If you post the scheme on reddit you can squeeze the shorts.

 

Is it ever?

My post was meant to be sarcastic.

If a booster should succeed in maintaining the 5V output, the output current will be 20mA. Which means the input current would be 33.3 mA, assuming the input is still 3.0V. But a data sheet fror the CR2032 shows an immediate drop to about 2.6V at 1/10th o that current. So we are looking at drawing 40mA, which I doubt that it could do into a short circuit.

The TS is expecting way too much from this battery.

Bob

 

CR1632 used battery
5mm Super Bright White LED
LED Vf @5V 20mA = 2.97Vf
Open voltage on the battery is 3.2V
Loaded voltage is 2.68 and falling.

 

We observe the current rating and forward voltage for example an led can have a rating of 20mA and forward voltage of 3V
An ammeter in series with an led will measure the current.
The purpose of using a resistor in series with an led is to limit the current.
For a 20mA led no resistance would be needed with a cr2032.
The datasheet shows you can expect short battery life which means not enough brightness over time but useful in an emegency.
http://www.cr2032.co/cms/prodimages/panasonic_cr2032_datasheet.pdf