Simplest way to drive four white/blue LEDs from a 3V coin cell

Thread Starter

menandore

Joined Sep 10, 2017
28
What I'm trying to achieve: miniaturized light-up wearables, using coin cell batteries (so that I don't have to run wires to a battery pack) and LEDs. I've actually gotten this to work with LEDs that require 3+ volts before (i.e., white and blue), but from what I can tell, that shouldn't have worked for more than five minutes (since the battery only puts out maybe 3.1 volts, when it's fresh), and I'd like to know how to do this the optimal way.

I've seen boost converter ICs that look like they might kinda do something to boost up the battery voltage to suit the LED? But most of them seem to require me to do differential equations and/or add about five more things to my board in addition to the IC itself.

I understand that I can't expect much in the way of battery life, since a CR2032 or similar battery isn't intended for anything remotely close to this use case. However, it only needs to stay lit for a few hours.

Any suggestions?

For reference, this is what I've achieved so far - just four LEDs in parallel, a 1025 coin cell battery, and a switch.

 

k7elp60

Joined Nov 4, 2008
562
I recently built a number of cross's with 12 blue 5mm LED's in parallel. I use a 2032 battery to power them and there is a 220 ohm resistor in series. The battery lasts for a long time. I have worn the cross multiple times for over 2 hours at a time. One thing I have found is to match the forward voltage of the LED's. I have had no luck getting white LED's to work of 3 volts. There is a 8 pin dip voltage pump that will double
the voltage. In addition to the IC it uses 2 diodes and 3 capacitors. It is a LT1054.
 

Thread Starter

menandore

Joined Sep 10, 2017
28
I did pick up some of these (it only has three pins, so it can't be too complicated to use, right???), but I'm not entirely clear on how to use it; I'm confused about the "switch" input. Is that the same as the input voltage from the battery that I want to boost, or is it something else?

https://www.digikey.com/product-detail/en/XC6372A331PR-G/893-1164-1-ND/2771912/?itemSeq=271361745

It's entirely possible that this IC isn't suited at all, but it looked feasible from the specs.
 

ScottWang

Joined Aug 23, 2012
7,397
The In/Out voltage shown in the datasheet of XC6371-XC6372 and the circuit really quite simple, that is suitable for your small device.
Operation Start Voltage Range : 0.9V~10V
Output Voltage Range : 2.0V~7.0V (0.1V increments)
 

Thread Starter

menandore

Joined Sep 10, 2017
28
So this would work, then? (In general. I might use a different 3V coin cell, and I probably have some polarities wrong. I usually just test the LEDs as I go to make sure I've got them the right way around.)



I really appreciate all the help!
 

ScottWang

Joined Aug 23, 2012
7,397
@menandore
The battery was reversed, so you have to change it and the pin 2 and pin 3 of XC6372 need to exchange.

If the LED is 3V/20mA then the R can be 18Ω for 3.3V Vout from DC-DC.
 

Thread Starter

menandore

Joined Sep 10, 2017
28
That's a pretty easy fix, so looks like I'm set. I'll let you know how it works out. Thanks again for all the help!
 

neonstrobe

Joined May 15, 2009
190
The simplest method of operating LEDs is to use a small battery and often cheap torches etc do not use resistors as they rely on the internal impedance of the battery to limit the current, and often run two or more LEDs in parallel. That is not my preferred method. I use a small ferrite transformer in a simple switching circuit which needs two or three windings (pri+feedback or pri+sec+feedback) typically in the ratio of Vin:Vout:Vin but does not have to be exact; the primary is designed to run at the input voltage and required power level, the feedback is generally the same as the primary while the secondary is able to drive the LEDs in series, so if they have 3.5V forward that is not a problem.
The circuit only needs a resistor to limit the base current, a transistor capable of switching the power (e.g. BC337) and if you want a smooth output a rectifier diode and capacitor (typ. 1N4148 and 1uF for 10-20mA LEDs), and takes very little room on an SMD board. The advantage is that with such a circuit, though needing a small ferrite transformer, does not overpower the LEDs with a new battery, and with LEDs in series all LEDs light without current sharing resistors.
 

RichardO

Joined May 4, 2013
2,270
Are you sure about the LT1054, it seems needs the power 3.5V~15V and the output is negative, do you have the examples of positive output?
The LT1044 and ICL7660 will work down to 1.5 volts.

You don't need a positive output. You connect the LED between the positive input voltage and the negative output voltage.
 

Thread Starter

menandore

Joined Sep 10, 2017
28
The simplest method of operating LEDs is to use a small battery and often cheap torches etc do not use resistors as they rely on the internal impedance of the battery to limit the current, and often run two or more LEDs in parallel. That is not my preferred method. I use a small ferrite transformer in a simple switching circuit which needs two or three windings (pri+feedback or pri+sec+feedback) typically in the ratio of Vin:Vout:Vin but does not have to be exact; the primary is designed to run at the input voltage and required power level, the feedback is generally the same as the primary while the secondary is able to drive the LEDs in series, so if they have 3.5V forward that is not a problem.
The circuit only needs a resistor to limit the base current, a transistor capable of switching the power (e.g. BC337) and if you want a smooth output a rectifier diode and capacitor (typ. 1N4148 and 1uF for 10-20mA LEDs), and takes very little room on an SMD board. The advantage is that with such a circuit, though needing a small ferrite transformer, does not overpower the LEDs with a new battery, and with LEDs in series all LEDs light without current sharing resistors.
I'm not quite seeing how this fits together - can you draw a rough schematic?
 

ScottWang

Joined Aug 23, 2012
7,397
The LT1044 and ICL7660 will work down to 1.5 volts.
Thanks.
I saw many applications circuits used the ICL7600 to generates the -5V output, but I didn't check the datasheet, so I thought that the input should be over 5V, I saw you mentioned that it can be used down to 1.5V and I just checked the datasheet, it really could use in the low power, so I can use it in the 3V(maybe 2032 or two 1.5V) low power, but for one 1.5V battery, the XC6372 is better, but I think it can't be easy to buy it in the local store.

You don't need a positive output. You connect the LED between the positive input voltage and the negative output voltage.
After I posted the message and I'm still thinking that why you mentioned the IC with the negative output and I realized that the two output pins just two independent pins, so they should be ok, but I always prefer the positive in and positive out.
 

Thread Starter

menandore

Joined Sep 10, 2017
28
To clarify - this is kind of a starter project for me; I got a hacky version working (see the picture in my first post; it's basically the same as the schematic I posted, but without the step up IC), but I'd like to upgrade it slightly to something that will work more reliably for LEDs with high forward voltages. But, preferably, with a circuit that I can still understand. It's been so long since my EE classes that I don't remember much other than KVL, unfortunately. (Also I know that a capacitor is basically a high-pass filter, and an inductor can be used as a low-pass filter).

I know I'll be sacrificing some efficiency and not getting optimal results if I choose one of the very simple solutions; I can live with that. It's still going to be better than what I have now!

Also, for my application, I'm shooting for a small footprint - so if a bit of ripple in the output won't make the LEDs obviously flicker, and lets me leave out half a dozen components, I can live with the ripple.

Don't get me wrong - I really appreciate the advice for super awesome upgraded circuits - but it's a bit beyond my current level. :-/
 
Last edited:

neonstrobe

Joined May 15, 2009
190
Here's my preferred LED driver circuit, based on a very old ringing choke circuit known since the 1950's (or possibly before even with tubes),
The load resistor is not a resistor but represents the LED loads - 4 LEDs at 3V (approx) and drives up to 10mA.
Input power is limited by the base current and hence base bias resistor, so this may need to be adjusted, which is a consequence of this basic simplicity.
Typical ferrite cores for this need only be small units approx. 1cm cube or thereabouts, primary and feedback about 20 turns, secondary about 100 turns thin wire and a small air gap usually with a sheet of 100gsm paper between the core sections. Not usually critical, only have to check core does not saturate at max. collector current, (typ. 200mA peak)
 

Attachments

Last edited:

k7elp60

Joined Nov 4, 2008
562
Are you sure about the LT1054, it seems needs the power 3.5V~15V and the output is negative, do you have the examples of positive output?
You are right about the data sheets on the LT1054, I have used ones by Linear Technology, and Texas instruments. I have had some work at 3V, I don't recall which manufacturer it was.
 

k7elp60

Joined Nov 4, 2008
562
I have found that present day blue LED's can be very bright with low current flow. Even though the LED may have a maxim IF of 20mA, they can be real bright with current of 1.5mA. I just tested a white LED. It was very bright at 1mA of current and the forward voltage was 2.5 volts.

As I stated earlier, if one is to match the forward voltage at a specified current, LED's can be connected in parallel and one current resistor will work.
 

Audioguru

Joined Dec 20, 2007
11,248
A cheap and simple solar garden light has a 4-pins IC with a simple inductor to boost the voltage from a 1V to 1.4V Ni-MH rechargeable battery cell high enough to light a few white or blue LEDs. The value of the inductor sets the LED current. A transformer is not needed.

The voltage boost will work without a solar panel and without a rechargeable battery.
 

Thread Starter

menandore

Joined Sep 10, 2017
28
A cheap and simple solar garden light has a 4-pins IC with a simple inductor to boost the voltage from a 1V to 1.4V Ni-MH rechargeable battery cell high enough to light a few white or blue LEDs. The value of the inductor sets the LED current. A transformer is not needed.

The voltage boost will work without a solar panel and without a rechargeable battery.
Oh, that's a neat idea. I might have to find one and take it apart to see how it works. Garden lights never occurred to me as an example!
 

Kjeldgaard

Joined Apr 7, 2016
476
Driver ICs for these garden lamps can often be purchased on eBay.

A quick search on Solar Garden Light Driver provides many (cheap) hits.
 
Top