http://londonstransport.proboards.com/

Sidelight / Speakerlight light PCB circuit description



Sidelight

The sidelight PCB is double-sided, with though-plated holes making connections between the two sides. All components are Surface Mounted Devices, except the electrolytic capacitors, the preset potentiometers and jumper connector J5, which are mounted on the component side and are soldered on both sides.


There are fifteen test points on the PCB, marked TE1 to TE15. These are larger, through-plated holes and can be accessed from both sides..


The sidelight PCB can be thought of as being in four parts:

Red LED driver circuit

Green LED driver circuit

Blue LED driver circuit

Sensor circuit

The three driver circuits are the same, so I’ll just describe the red one. Voltages are DC and approximate, and measured with respect to GND. Refer to the circuit diagram.


The red signal from the wallwasher comes in at pin 2 of connector J5 on the PCB. The signal is Pulse Width Modulated and varies between 0v when the red is off and around 16v when the red is at maximum brightness. Test point TE7 allows the red signal input to be monitored.

Diode D3 is a double ESD protection diode (only one of the two is used) and protects the red input from ElectroStatic Discharge and surge pulses.


The signal passes to the non-inverting input (pin 3) of op-amp IC1. The inverting input (pin 2) is connected to GND via the wiper of preset potentiometer R49. This allows the brightness of all the red LEDs to be adjusted when calibrating the light.


The output of the op-amp is used to drive the two groups of LEDs - group 1 (LD4, LD3, LD2, LD1 - LD1) and group 2 (LD15, LD14, LD13, LD16 - yes, that is the correct order!)

R7 is the bias resistor for T1. T1 acts as a switch and closes to connect LD4 to GND via R49. Current can then flow between VCC and GND via LD1 - LD4 and the LEDs are illuminated.

The circuit for group 2 is the same, except that R14 is a fixed resistor, whereas R49 is variable.


As already mentioned, R49 is used to control the brightness of all the Red LEDs. R49, in conjunction with R47 in the green circuit and R44 in the blue circuit, are used to calibrate the light - getting the right amount of red, green and blue to make a uniform white light. Pin 1 of R49 is not connected. Pin 2 is connected to ground and pin 3, the wiper, provides the inverted supply to the op-amp and the path to GND for LD4 - LD1.


Sensor circuit

The sensor circuit has its own 5v supply from IC3, a 5v regulator. The circuit is built around photo transistor U1 which is mounted in the centre of the LEDs. The output from U1 (pin 4) goes to the wallwasher via pin 8 on connector J5. Diode DA3 is a bi-directional double ESD protection diode (only one of the two is used) and protects the sensor circuit from ElectroStatic Discharge and surge pulses. TE2 and TE5 are test points for the output of U1. These two test points are connected together for some reason, so both test points will always show the same reading. The output is 0v when no LEDs are on and around 4v when when all LEDs are on at maximum brightness (i.e. full white light). The voltage reading varies between these two voltages depending on the brightness and colour of the light that is given out.



Grounds

The circuit / PCB is divided into two separate grounds. I have labelled them as GND and GNDs.

GND joins all the driver circuits. Red, Green and Blue have their own ground wire. This is twisted with the corresponding colour input wire when the wires leave the shielded cable. These grounds go to the wallwasher ground and are joined together on the light PCB. This means that J5 pins 3, 5 and 7 are effectively joined together.

The sensor circuit has its own ground (GNDs) which is connected to J5 pin 9. This is twisted with the sensor wire when the wires leave the shielded cable. There is no physical connection on the PCB between GND and GNDs.

GNDs is joined to the wallwasher ground. If the sidelight is unplugged or the sidelight is plugged in but connector J5 removed, then there is no connection between GND and GNDs. This is something that should be remembered during testing.

The sidelight cable has individual wires with an overall screen shield. The screen is connected to the shell of the sidelight plug. There is no connection to the screen within the sidelight. When the sidelight is plugged into the wallwasher, the screen is  then connected to the wallwasher ground.





















Fig. SC1:  Sidelight Ground connections




Component locations




















Fig. SC2:  Sidelight PCB - overview of component areas



See the Sidelight faults \ Testing LEDs page for the LED PCB overview



Test points

There are a total of fifteen test points (TE) on the PCB. These are marked on the component side. They are also accessible on the LED side, but not marked - see the component identification picture below.

See the main circuit diagram for where the test points connect to in the circuit


The test points are listed below, together with a description of what they are measuring and their expected meter readings when the colours are full on (max white - normal) and the colours are off (light disabled)


TE1 5v VCCS for the sensor circuit

TE2 Sensor output - connects with J5 pin 8  Also connects to TE5

TE3 Sensor GNDs - connection at J5 pin 9

TE4 Blue input - connects with J5 pin 6

TE5 Sensor output to J5 pin 8  Also connects to TE2

TE6 Green input - connects with J5 pin 4

TE7 Red input - connects with J5 pin 2

TE8 Main ground - connects to J5 pins 3, 5, and 7

TE9 VCC in

TE10 T3 connection to blue LEDs group 1

TE11 T4 connection to red LEDs group 2

TE12 T9 connection to green LEDs group 2

TE13 T14 connection to blue LEDs group 2

TE14 T2 connection to green LEDs group 1

TE15 T1 connection to red LEDs group 1


































Fig. SC3:  Sidelight PCB component side test points and main components




































Fig. SC4:  Sidelight PCB LED side test points and through-hole component locations

























Fig. SC5:  J5 connector (partially inserted)



Circuit notes

In the voltage readings, I have replaced the decimal point with a ‘v’ to make it easier to read. E.g. :-

17v12 = 17.12v,     0v66 = 0.66v = 660mv

I’ve also used a v instead of V in all the diagrams and text because it looks clearer.

I have used the more common way of representing a resistor value with an R or with the multiplier. E.g. :-

47R = 47 ohms,  6K8 = 6.8K


The red LED group is shown the correct sequence - LD15, 14, 13, 16.


R5 is shown correct at 390R. This is different to its equivalent R11 in the green and R17 in the blue circuits, which are 510R

The red LEDs have a lower forward voltage than the green and blue LEDs. This is the reason for the smaller difference in the voltage readings between the red LEDs in the group. E.g. The voltage at TE11 is 10v74. This is a drop of 7v2 from the 19.94 VCC. Green and blue have a voltage drop of 12v34 and 12v08.


All the electrolytic capacitors, except C8 are connected between VCC and GND. For convenience, I have shown these separate on the circuit diagram. C14, 15, 16, 18, 19 and 21 are associated with the six groups of LEDs. C4 and C12 provide smoothing for VIN of IC3. C8 provides smoothing at VOUT of IC3.


Sensor U1 (Avago APDS-9002-021) is shown on the data sheet as pin 4 being No Connection. On both PCBs I tested, pin 4 is connected to pins 3 and 2, as is shown on the diagram.



Working on the PCB

When the sidelight is switched off via the DirectControl panel or similar, although there is no light showing, the VCC voltage is still present - see the note on the Sidelight faults page.



Measuring the resistance

If you want to measure the resistance of any part of the circuit or a component, the light or wallwasher must be disconnected. Do not try to measure the resistance with the light on, or with the light just switched off via DirectControl etc. (VCC is still present), otherwise you may get a false reading, or potentially damage something.



Voltage readings

The circuit diagrams with voltage readings at the test points and various other places are at the bottom of this page. Unless stated otherwise, all measurements are between the measuring point and GND (i.e., with respect to earth). GND being TE8 or J5 pins 3, 5 or 7.

The readings have been taken using a digital multimeter as this is what most people will use. An oscilloscope would have given an exact reading for any point in time of the PWM voltage, whereas a meter tends to give a more average reading, depending on the actual PWM wave, but it doesn’t matter here. I compared the meter readings with the scope readings and they are near enough the same. See the bottom of the page for PWM information.


Normal readings circuit diagram

The light was set up on white at maximum level. Only the one sidelight was connected to the wall washer. There was no rumbler or fan connected. The three wallwasher lights were also on white at the maximum level. However, whatever you have connected and running shouldn’t make much difference. Depending on how much current is being drawn, there may be a slight reduction in VCC. Likewise, if you only have one colour showing instead of white, VCC may be very slightly higher.


Light off readings circuit diagram

The sidelight was turned off from the DirectControl panel and there was no light showing. The PCB is still live because VCC is still present. Whilst the colour inputs and IC1 input and outputs are 0v, it will be seen that the voltage across the LEDs actually increases. This is due to the fact that the transistor switch is open and so there is no path to GND through the transistor that would normally drain (for want of a better description!) some of the voltage away.


Taking your readings

These readings are approximate and may vary slightly between lights. You will probably find that some of the readings are slightly different at different times when you take measurements after the light has been tuned off and on again, even though the settings haven’t been changed. The readings are a guideline to what to expect. When faultfinding, any significant change in one or more voltage readings may give a clue as to where the fault lies. Some of the readings are very low and it’s possible they may change slightly depending on if you’re touching anything on the PCB. Your body acts like a conductor and it may be that at times you could be creating a high resistance path to earth or between the components you’re touching. Depending on the sensitivity of the light circuit, this can sometimes be seen by causing the LEDs to glow faintly when testing the LEDs. Likewise, connecting the multimeter leads to components can cause a similar effect, especially when testing the LEDs.


When taking readings, especially if putting the probe directly onto the component’s solder connection, you may find that you don’t get a reading or the reading seems way out. Sometimes the contact between the probe tip and the solder surface can be poor and you may need to move the tip around the joint a bit or put SLIGHT pressure on it.


It is not very easy to try and take readings from the PCB with two probes.  Don’t attempt to clip a probe to a capacitor “leg” - you will end up with a short circuit. The simplest way that I’ve used (although not for the readings shown here) is to connect the COMmon probe to a crocodile clip and attach this to the shield of the USB wallwasher plug at the PC. This will be the multimeter GND connection and you only need one probe to touch to the PCB to get the measurements. The only problem with this is that the extra long path from the meter to GND means a higher resistance in the meter connection which can affect the readings. However, this difference is slight and in most cases won’t make any difference. If you want an accurate reading, connect the multimeter probes to the PCB.





































































Fig. SC6:  Sidelight PCB circuit diagram






































































Fig SC7:  Sidelight PCB circuit diagram with normal voltage readings

















































  




















Fig. SC8:  Sidelight PCB circuit diagram - voltage readings with the light off



Pulse Width Modulation

The voltage to the LEDs is Pulse Width Modulated. PWM is a common way that is used to dim lights etc. and extends the life of the bulb or LED. Put very simply, instead of using a continuous voltage for a given level of brightness. PWM works by sending pulses of electricity. The top of the pulse is normally at the maximum voltage (around 16.2v for the sidelight), but the duration (Width) of the pulse is varied (or Modulated) according to the brightness.

The charts below are from the red input and show an example of the pulse wave. The vertical grid is in steps of 1v, from 0v to 17v. The horizontal grid is in steps of 1 millisecond, from 0 - 17mS.


An oscilloscope will show the voltage of the part of the pulse that is being measured at that exact moment. If you were to draw a vertical line down the chart below, the voltage measurement would be where the line touches the blue line of the pulse. If a multimeter is used to read a PWM voltage, it will tend to display the average voltage of the pulse which is normally sufficient.


The PWM wave can be seen at different stages at the LED driver circuit, including the input and output of IC1



Fig. SC9a:  The red light is at maximum brightness - input voltage 16v15



Fig. SC9a shows the red on maximum brightness. The top of the pulse is 11mS wide, with only a very narrow drop to 0v between pulses. When measured with a meter, the red input has a voltage of 16v15.






Fig. SC9b:  The red light is at mid brightness - input voltage 5v31



Fig. SC9b shows the red on medium brightness. The top of the pulse is 0.25mS wide, with a much bigger gap at 0v between pulses. When measured with a meter, the red input has a voltage of 5v31.




Fig. SC9c:  The red light is at minimum brightness - input voltage 1v1





Fig. SC9c shows the red on minimum brightness. The top of the pulse is 0.05mS wide, hardly any pulse width at all. Nearly all of the voltage is at 0v. When measured with a meter, the red input has a voltage of 1v1.


VCC  in
Blue LED driver
VCC smoothing
Green LED driver
Red LED driver
Sensor circuit