Hello everyone. This is my first post on the forum, I am currently working on a seemingly simple circuit. I think a constant current driver is the solution; after wiring everything up last night the LED was nice and bright and circuit ran great. I decided to charge the Lipo over night, and when I turned it on this morning the LED was at 10%-50% power. Certainly not as bright as last night. I’m thinking maybe this is due to the Lipo charging up to 4.2 volts? Anyways here is the data. I am running an LED with a 3.3 forward voltage rated at 20 ma off of a 3.7 v rechargeable Lipo battery. Using ohms law (3.7-3.3=.4 divided by .02 Amps) I thought I needed a 20 ohm resistor. This in theory works fine if the Lipo stayed at 3.7 volts. After fully charging to 4.2v ohms law would say that the 20 ohm resistor I put in was not enough and the LED was getting 45 ma of power. Could this have caused it to loose it’s brightness so fast? I would just put a 45 ohm resistor in to drop the led to 20 ohms @ 4.2 volt supply but then when the Lipo dropped to its normal 3.7 volts the LED would only be pushing about 9 ma. If I am correct? I am a noob electrician. For my project I want the LED at max brightness. And it’s datasheet shows a steep drop off in lumens when under 20 ma so under powering it is not an option. I used thin 30 gauge wire and thought maybe a connection got loose. So to validate that it is the LED I was going to re-soldier/ new wire everything. But if the LED is still dim it’s back to the drawing board. I was planning on purchasing a 20 ma constant current driver and replacing the resistor with that. Would this in theory work to constantly regulate the circuit to provide the LED with 20 ma regardless of if the battery is at max charge (4.2) volts or it’s normal 3.7 volts? I am looking at the constant current driver AL5809q by diodes incorporated. It has a minimum input voltage of 2.5 v. My other question is will the constant current driver hog voltage from my circuit? As I am powering a 3.3 v forward led with 3.7 volts I don’t have much voltage to spare to add other components requiring power. Thank you for your help. Will attach a spec sheet for the constant current driver I am looking at.

 

I don't think that will work as the specs mean there's a minimum of 2.5 volts from input to output you only have at max 4.2 - 3.3 = .9volts
You need a higher voltage battery.

 

Did you say you were pushing 45ma through an LED rated at 20ma for several hours? If I understand that, you might want to try replacing the LED to see if it's bright. But I think you're right, you need to stabilize the current at the rated 20mA or the LED will overheat and lose brightness.

 

If You would provide an explanation of exactly what You are trying to accomplish,
more useful solutions could be given.

It's possible that You have partially damaged your LED.

You may need to use 2-Li-Po-Cells to get the maximum life from a Charge.

How are You charging your Li-Po-Cell ?, you may have damaged the Cell.

A "CL20" Current-Limiter, with 2-Cells, is probably the easiest solution to
get constant Light-Output from your LED, and the longest run-time,
but, You must be very careful not to over-discharge the Cells, or You will damage them.
.
.
.

 

What is a constant current driver versus a constant voltage driver?

In theory, an ideal constant voltage driver has zero internal resistance.
Conversely, an ideal constant current driver has infinite internal resistance. For this to work, the constant current driver has to have infinite supply voltage.

Therefore, your setup is wrong. You need to increase the series resistance, and yes, the resistance has to drop the excess voltage.

If you must use a 3.7-4.2V battery, the only solution I have to offer is to use a boost converter to boost the voltage to at least 6V. 12V would be even better. Calculate the series resistor to suite.

 

Another solution is a 3.3 volt LDO regulator (low voltage dropout )
The MCP1700 will maintain 3.3 volts down to a 3.5 volt input.

 

I don't think that will work as the specs mean there's a minimum of 2.5 volts from input to output you only have at max 4.2 - 3.3 = .9volts
You need a higher voltage battery.

I thought this was the case…

Did you say you were pushing 45ma through an LED rated at 20ma for several hours? If I understand that, you might want to try replacing the LED to see if it's bright. But I think you're right, you need to stabilize the current at the rated 20mA or the LED will overheat and lose brightness.

Not several hours, not sure how long a 3.7 volt lipo at full charge of 4.2 volts takes to drop down to 3.7 volts but it seemed dimmer almost immediately. Maybe I’ll look for an LED rated to withstand the fluctuation in voltage.

If You would provide an explanation of exactly what You are trying to accomplish,
more useful solutions could be given.

It's possible that You have partially damaged your LED.

You may need to use 2-Li-Po-Cells to get the maximum life from a Charge.

How are You charging your Li-Po-Cell ?, you may have damaged the Cell.

A "CL20" Current-Limiter, with 2-Cells, is probably the easiest solution to
get constant Light-Output from your LED, and the longest run-time,
but, You must be very careful not to over-discharge the Cells, or You will damage them.
.
.
.

will update post with a pic of circuit so you get a better idea. I need this to be as SMALL as possible so I would like to stick to just 1 battery, and it needs to be rechargeable. And as many amps as possible, at least 100.
I am charging via an Ada fruit usb c lipo charger and when all Said and done it will stay connected to the charger and the usb c port will stick out of my device for charging. My circuit is connected to the bat and ground terminals on the adafruit lipo charger. I am using a tiny 105 ma lipo now the size of a quarter which is perfect.

What is a constant current driver versus a constant voltage driver?

In theory, an ideal constant voltage driver has zero internal resistance.
Conversely, an ideal constant current driver has infinite internal resistance. For this to work, the constant current driver has to have infinite supply voltage.

Therefore, your setup is wrong. You need to increase the series resistance, and yes, the resistance has to drop the excess voltage.

If you must use a 3.7-4.2V battery, the only solution I have to offer is to use a boost converter to boost the voltage to at least 6V. 12V would be even better. Calculate the series resistor to suite.

so I was misunderstanding a Constance current driver and they do eat voltage? they are so small I figured they would not consume volts but just impede flow like bottle neck. could I not just use a constant voltage regulator then? Why would that not be a solution?

Would a boost converter push out a steady voltage Regardless of the varying voltage from the lipo? If not I would still need to add a a constant current / voltage regulator no?

thank you for your help

 

It seems as though a voltage regulator would work as a solution.

Another solution is a 3.3 volt LDO regulator (low voltage dropout )
The MCP1700 will maintain 3.3 volts down to a 3.5 volt input.

I see the output current is 250 ma,but I am using a 105 MA lipo battery. would this work?

 

Are You building this project as an entertaining Hobby ?,
just to see if You can do it ?

Miniature Flashlights are less than ~$3.oo at Walmart.

The smaller You make your Flashlight,
the more difficult and expensive this project will become.
.
.
.

 

It seems as though a voltage regulator would work as a solution.



I see the output current is 250 ma,but I am using a 105 MA lipo battery. would this work?

Just means the regulator can handle up to 250 ma.

 

Let us do the math.

Suppose the supply voltage is 13V and the LED forward voltage is 3V.
The voltage across the resistor has to be 13-3= 10V
For 20mA current, the resistor value is 10/0.02 = 500Ω

Suppose the supply voltage drops to 10V.
The LED current will drop to (10-3)/500 = 14mA

The LED is still going to shine brightly.

The trick here is you want the series resistance to be as large as possible.

 

Are You building this project as an entertaining Hobby ?,
just to see if You can do it ?

Miniature Flashlights are less than ~$3.oo at Walmart.

The smaller You make your Flashlight,
the more difficult and expensive this project will become.
.
.
.

cost is not an issue, the circuit being as small as possible Is important.

 

Just means the regulator can handle up to 250 ma.

okay this will work fine then. So no resistor even necessary if its putting out 3.3 v as this is the rating for forward V on led

 

okay this will work fine then. So no resistor even necessary if its putting out 3.3 v as this is the rating for forward V on led

Wrong. You are going to kill the LED.

 

The dynamic resistance of the LED is much lower than its static resistance. Any small increase of the supply voltage (or incorrect specs of the LED forward voltage) will potentially cause a much larger current to flow.

Read my blog here.
https://forum.allaboutcircuits.com/ubs/why-you-need-a-series-resistor-when-driving-an-led.1651/

 

okay this will work fine then. So no resistor even necessary if its putting out 3.3 v as this is the rating for forward V on led

Probably need a small value of resistance depending on the actual LED used and output from the regulator.

 

Just did a test using a 3.3v LDO with a white LED and it required a 15 ohm series resistor. Maintained appx 22ma down to appx 3.4 volts. YMMV.