Help replacing LED on PCB

Thread Starter

john2k

Joined Nov 14, 2019
85
I have a PCB circuit that has 3 LED's on it that are very low brightness and are green. I would like to de-solder them and solder on white LED's. Below is a pic. Can anyone kindly tell me what kind of LED I need to replace this with and are they all the same size and what about brightness etc is that controlled by the PCB anyway? what spec LED should I get?

 

bertus

Joined Apr 5, 2008
20,465
Hello,

Keep in mind that the forward voltage of white leds is much higher as the forward voltage of green leds:

LumexLEDColorGuide.png

IV-curves-all-colours.png

What is the physical size of the leds?

Bertus
 

Thread Starter

john2k

Joined Nov 14, 2019
85
The physical size of the led seems 3.1 x 2.7mm. what about a blue led? If the forward voltage is higher for blue and white compared to the green, then will the PCB already be designed to provide a low voltage to the green LEDs and hence the white and blue ones will be underpowered so should be ok?
 

Thread Starter

john2k

Joined Nov 14, 2019
85
Hello,

The forward voltages of blue and white leds are about the same.
Leds are current controlled devices.
When the forward voltage is much higher, the current will likely be much lower.
It might even be so low that the led will hardly illuminate.

Bertus
So could I de-solder 1 and put a blue 1 and try and see if it illuminates? If it illuminates then it should be ok? What kind of led do I need to order?
 

dl324

Joined Mar 30, 2015
10,536
What kind of led do I need to order?
You're going to be limited to a compatible package.

White LEDs might not be compatible with the circuit. You should trace the circuitry to determine how current is being limited. If you don't have sufficient voltage or current to drive the replacement LEDs, there's no point in even trying.
 

djsfantasi

Joined Apr 11, 2010
6,362
Limited by the package, and current requirement. Vf is not a limiting factor, the current is!

Unless you can change the current for each LED.
 

Thread Starter

john2k

Joined Nov 14, 2019
85
I have a spare PCB to experiment on so worth a go and see if it works. Seems like I need the 1206. What's the best way to de-solder it off the board? My solder iron tip isn't super thin.
 

dl324

Joined Mar 30, 2015
10,536
What's the best way to de-solder it off the board? My solder iron tip isn't super thin.
A hot air tool would be best. If you use an iron, you either need to add some low melting point solder so you have a chance of getting the solder on both sides to be molten at the same time. Or you need to remove enough solder to release one terminal and then melt the solder on the other.
 

ElectricSpidey

Joined Dec 2, 2017
1,070
If it were me…

I would measure the voltage of the LEDs while in the circuit, then unsolder one end and measure the current.

Then I would look for something of a close match IE forward voltage @ given current.

But I would also look around beforehand to be sure the size is available.

But that’s only me.
 

eetech00

Joined Jun 8, 2013
1,869
If it were me…

I would measure the voltage of the LEDs while in the circuit, then unsolder one end and measure the current.

Then I would look for something of a close match IE forward voltage @ given current.

But I would also look around beforehand to be sure the size is available.

But that’s only me.
yes....the TS has not identified the LED driver. That is...if it’s a simple resistor current limiter or constant current/voltage. In either case, the Vf is still important. Since different LEDs have specific Vf/If characteristics and most are not the same.

eT
 

jbeng

Joined Sep 10, 2006
72
From where I sit, it appears that LED1 & LED2 are in series with a 1kΩ current limiter (R6 & R7 - 2 x 2kΩ, in parallel). LED3 is in series with a 900Ω current limiter (R8 & R9 - 2 x 1.8kΩ, in parallel). Probably not much more than that controlling the brightness, unless they're fed from a PWM circuit.

That said, if it's a constant DC feeding them, the potential between test points T8 & T6 should tell the value of that DC level. If it's PWM'd, It would be best to measure it with a scope.
 
Last edited:

Thread Starter

john2k

Joined Nov 14, 2019
85
Here is a full view pic of the board. There is a 6 pin connector soldered with the 6 solder points at the bottom. Is anyone able to tell me which two of these pins provide + and - power going to the 3 LED's on the board? Thanks

 

jbeng

Joined Sep 10, 2006
72
Assuming they LEDs are installed as the silkscreen markings show, the PCB traces show me that the pin third from the left end would be LED+ and the one on the far right end pin would be LED-. You should be easily able to light them up using a 9v transistor battery. LEDs 1&2 will be dimmer than LED3, since LED3 will have more the current flowing through it.

In case you're wondering how I arrived at the LED current values, here's how:
Using data from the LED chart kindly provided by Bertus (thanks!), the forward voltage drop (Vf) for a green LED will be between 2.2v to 2.6v (if it's what I call lemon-lime color, a kind of yellowish-green) or 3.5v to 4v (more of a pure green). The color is dependent on its chemistry, Gallium Phosphide (GaP) or Indium Gallium Nitride (InGaN), respectively.

Let's say the LEDs are lemon-lime color and their Vf=2.2v each, which is typical. Since LEDs 1&2 are in series, those voltage drops add giving 4.4v. Those LEDs are also in series with a parallel pair of 2kΩ resistors, which results in a value of 1kΩ. Using a 9v supply, the total LED voltage drops are subtracted from the 9v. That leaves 4.6v to be dissipated in the resistors. The current flowing through them is calculated by using simple Ohm's Law; divide the value of that remaining voltage (4.6v) by the resistor value (1kΩ) which gives 0.0046 amps or 4.6 milliamps (mA).

In the case of LED3, it's figured the same way, except that there is only one voltage drop instead of two and it's current limiter resistors are 1.8kΩ each (giving a 900Ω result). So, 6.8v would be remaining after subtracting the LED voltage drop. Dividing that voltage by 900Ω, you get 7.5mA.

As far as white LEDs go, they are just blue LEDs with a blob of yellowish phosphor covering the LED chip. The blue light hits the phosphor, the phosphor fluoresces and the resulting color is white.
 

Tonyr1084

Joined Sep 24, 2015
4,391
Assuming they LEDs are installed as the silkscreen markings show, the PCB traces show me that the pin third from the left end would be LED+ and the one on the far right end pin would be LED-. You should be easily able to light them up using a 9v transistor battery. LEDs 1&2 will be dimmer than LED3, since LED3 will have more the current flowing through it.
Assuming we number the connector pins as shown - from left to right, as you said, pin 3 appears to be a positive voltage connection. From there they follow, as you said, through those four resistors. However, at the junction of T6 there's an unspecified resistor (R2, who's resistance we don't yet know), then back to pin 1. It looks to me like there's the potential that all three LED's could be constantly lit by pins 3 and 1. There's also the current pathway back to pin 6 (or 5, hard to tell). When you press the Mode button you jump R2 out of the circuit which will change the voltage (yes, LED's are current controlled but the change in voltages may mean something to whatever the board is connected to). R4 is connected to Seek- and R5 to Vol-. R1 is connected to Seek- and R3 appears to be connected to Vol+. It's not easy to tell for sure. One trick to see the traces better would be to backlight the board (shine light through the board). That appears to be a single layer board and some traces (as it appears) are drilled through after etching to disable a part of the circuit. Those five resistors, R1, 2, 3, 4 & 5 have values that are not marked on them. They can be anything. Pressing the associated button drops those resistors out of the circuit and cause a change in voltage. Or current - I don't know for sure. But changing the LED's (or resistors) may result in the board not functioning properly. For instance you could push seek+ and the unit may switch modes or change the volume.

The only way to tell for sure is to have a manufacturers schematic along with voltages applied to the board and the pin-out of the connector. Without that, changing LED's (or resistors) may cause the board to not function at all.
 

Janis59

Joined Aug 21, 2017
1,169
RE: Berus - nice graph, but sth is wrong about T,V,P types having a wavelength characteral to Roentgen. For sure some figures there are criss-crossed.
 

Thread Starter

john2k

Joined Nov 14, 2019
85
OK so i ran 12v power to the two pins identified and it runs the LED's. So i de-soldered and soldered on 1206 blue LED's and it's working perfectly. The blue LED's are brighter than the green ones too. One thing is that the 1206 is smaller than the ones that were on the board but it fits well. Seems to all be working fine. Left it running for a bit and no issues so far.
 

jbeng

Joined Sep 10, 2006
72
Here's my take on a quick reverse-engineer of the PCB circuit. The connector pins at the bottom are in the the same orientation as the PCB pins in the photo.
Pins 1 & 5 (as we have come to call them) are likely connected to a A/D inputs of a CPU to decode the control switches.
It's not clear from the photo if pins 1 & 2 are connected, they might be. The same goes for pins 3 & 4. So, in my diagram, I used dashed lines to show that they might be connected.
As determined by their markings, R6 & R7 are 2kΩ each and R8 & R9 are 1.8kΩ each. However, as Tonyr1084 said, resistors R1-R5 have no markings, so their values are unknown.
The pads marked with designators T1-T10 are likely test points for automated in-circuit testing.

Control Bd_AAC.jpg
 

MisterBill2

Joined Jan 23, 2018
5,803
Hello,

The forward voltages of blue and white leds are about the same.
Leds are current controlled devices.
When the forward voltage is much higher, the current will likely be much lower.
It might even be so low that the led will hardly illuminate.

Rapid seems to carry 3.2 mm white leds:
https://www.rapidonline.com/Catalogue/Search?Query=smd led white&Attributes={"Length":["3.2mm"]}

Bertus
Look at the chart and graph in post #2 and it will be clear that LEDs are not constant power devices. As the forward voltage rises so does the current, and not in a linear manner like a resistor. They are diodes and perform like diodes, except for the forward drop being a function of the color.
 
Top