http://londonstransport.proboards.com/

Sidelight / Speakerlight light faults



The information and descriptions given here are based on a sidelight PCB. “All of one colour” or “whole group” means what it says - i.e. all the red LEDs, or all green LEDs, or all the blue LEDs. “One group” means one of the two groups of that colour - i.e. only half that colour LEDs. The red, green and blue circuits operation are identical. I will use the red circuit as an example. If you have a problem with the green or blue LEDs, then follow the description for the red circuit, but substitute and component reference with the relevant colour. E.g. the pot in the red circuit is R49. In the green circuit it is R47 and in the blue circuit it is R44.

When describing the possible causes for a fault, I am assuming that the PCB itself is not faulty - there are no broken tracks (it’s unlikely that the PCB would be at fault, unless work has been done on the PCB and potentially caused any damage). Another cause of a fault could be a dry joint between the component and the PCB. Unless specifically mentioned, it is assumed that there are no dry joints.

The test points are a useful means of checking whether everything is as it should be and help to pinpoint where something might be wrong.


The sidelight PCB gets quite warm - around 52 degrees on the LED side and around 45 degrees on the component side. Most of this heat comes from the LEDs. Too much heat can cause a dry joint. This is when there is poor contact between the component and the PCB through the solder. Because of the constant change in temperature, the PCB expands and shrinks. This puts pressure on the contacts and can cause intermittent problems. One of the potential signs of a dry joint in these circumstances is that the operation changes as the heat increases. For example, there might be contact when the PCB is cold, but as the PCB warms up and expands, the contact is lost and a failure occurs. A similar thing can occur within a component and eventually the component can fail.


Potentiometers can be another source of problems (see calibrating for potentiometer details). The pot track can get worn or dirty. The pots used are a standard type, but they are “cheap and nasty” and I’m certain they’re not really suitable. The current passing through R49 from group 1 red LEDs is 0.014mA (0.25w) and the current passing through R44 from group 1 blue LEDs is 0.016mA (0.3w). This is a lot of current and I suspect that the track on the pot eventually gets damaged / burnt out at the point where the wiper makes contact. This can give an intermittent contact (flickering light) or an open circuit (no light). If a pot needs to be replaced, replace it with one of a better quality, higher wattage type.



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

Refer to and follow the circuit diagram.


At present, I’ve only had a faulty LED, so some of the information is guesswork. However, I think it is correct.



Working on the PCB

When working on the PCB, remember that it is “live” if it sidelight is plugged into the wallwasher and the wallwasher is switched on. Even if the light has been switched off in the DirectControl panel etc. The VCC supply is still present. Because of the closeness of the components it is very easy to short circuit something and potentially damage one or more of the components. Ensure there is nothing metal on the work surface you are using that can potentially touch any of the components. Always keep screws, probes etc. Away from where the PCB might be resting. Remember, there are contacts on either side of the PCB.


Always unplug the light from the wallwasher before removing or inserting the plug that goes into connector J5. Be careful when moving the PCB. If the attached cables are flexed too much, eventually they could break. When removing or inserting the plug, take care not to damage or bend any of the pins. The plug can be quite stiff. Never try and remove it by tugging on the wires. Use a small flat blade screwdriver between the base of the plug and the connector and gently lever it up a bit. Go to the other end and do the same. Repeat this process until the plug is fully removed.   



Where is the fault?

There are various reasons why a sidelight doesn’t work, including:

faulty component

faulty connection

faulty wallwasher


It’s necessary to try and identify the cause, usually by a process of elimination

First of all, go to the DirectControl panel and make sure the light(s) are switched on, enabled  and the height is set to None.  Make sure the lights are set for white light and are set at maximum brightness. White light means that all three colours, Red, Green and Blue, should be lit.



Check the plugs

If the sidelight plug is not fully in place, it’s possible that some contacts could still be made. Depending on what are making, you could have no colours, one colour or two colours lit. A loose plug could also cause an intermittent light operation.

Check that the light is fully plugged into the wallwasher and tighten the plug’s thumbscrews to hold the plug in place.



No lights working at all

Suspect the wallwasher



Important

See Faults and repairs for precautions you should take before taking the light apart.


One light not working (no lights at all)

Note

The 18v VCC must be present on pin 1 of the wallwasher socket (and thus on pin 1 of the sidelight PCB connector J5 and PCB test point TE9) or the light will not work.

The 18v is only available when pins 9 and 12 are shorted together (see the sidelight pin-out diagram). This occurs when the sidelight plug is plugged in. If you’re testing the wallwasher socket without the light plugged in, make sure pins 9 and 12 are shorted together. See Testing the wallwasher sockets on the Faults and Repairs page.  


Prove whether it is the light or the wallwasher that is defective by trying another sidelight in the same socket. A red plug can go into a white socket and vice versa. (I assume that sidelights with speakers can still be connected this way, although you may lose the speaker / light controls).

If the light lights up in a different socket, then suspect the wallwasher. If the same light still doesn’t light up, then that light may be faulty.


A light not working could be caused by:

faulty plug

defective cable

poor contact on jumper J5 on the light PCB

a fault on the PCB (PCB or component fault)


Check that the plug and cable aren’t damaged. Try moving the plug slightly in the wallwasher socket - if the light comes on / goes off, it may be a faulty plug

With the light plugged in, try flexing the cable up and down - if the light comes on,/ goes off, it may be a faulty cable


Dismantle the sidelight so that you can access the PCB. Ensure that Jumper J5 is fully secure and pushed in straight. J5 is normally a tight fit. Unless you’ve taken the light apart before, it’s unlikely that J5 is the problem.


Using a multimeter, do the following measurements, either between the test point (TE) or J5 pin (J5/X). Pin 1 of J5 (normally the black wire) is on the left (next to J6). VCC should be approximately 18v, the colour inputs should be approximately 16v

TE9 to TE8  (J5/1 to J5/7) = VCC to GND

TE7 to TE8  (J5/2 to J5/7) = Red in to GND

TE6 to TE8  (J5/4 to J5/7) = Green in to GND

TE4 to TE7  (J5/6 to J5/7) = Blue in to GND


If the voltages are not correct

If the voltage is 0v, then there is probably a break in that wire or a poor connection at the plug. If VCC is less than what it should be, this would affect all colours. As long as VCC is correct, a break or poor connection on one of the colour wires will only mean that that colour is lost.

It is very unlikely that a poor connection or a faulty cable would affect all the wires.


Unplug the light. Disconnect J5. Plug the light back in. Do the same tests again, this time testing between the holes on the J5 connector. Use a thin piece of insulated wire to go between the hole and the multimeter probe.

Test between the holes connected to the following wires:

Black to Blue = VCC to GND

Brown to Blue = Red in to GND

Orange to Blue = Green in to GND

Green to Red = Blue in to GND   Note, the Red, Yellow and Blue wires are all GND. Using the green instead of blue wire socket means that the two wires you are putting in the holes will be further apart, thus there will be less chance of them accidentally shorting against it other during the test.


If this test shows the correct voltages, unplug the light, connect J5 back again and plug the light back in. Repeat the tests between the test points / J5 pins. If the voltages are now showing incorrectly again, it’s possible that there might be a short circuit on the PCB or one of its components, such as a capacitor.



If the voltages are correct

If all the voltages are correct when J5 is connected to the PCB, then the connection between the wallwasher and light PCB should be OK and the fault is somewhere on the PCB.


Although the colour circuits are separate on the PCB, they all share VCC GND and IC1. IC1 has four op-amps. Three are used, one for each colour. If one op-amp fails, that colour’s  LEDS will not work. If there is a fault with the whole of the IC, then no LEDs will work.


Refer to the normal readings diagram on the Circuit page.

VCC has been tested and proved OK, so that leaves IC1. Check the three colour outputs of IC1. The places to check are:

Pin 1 (at the junction of R7 and R8 - Red output

Pin 7 (at the junction of R13 and R29) - Green output

Pin 8 (at the junction of R19 and R39) - Blue output.


If the voltage is wrong, check the voltage at the input pin:

Pin 2 (wiper of R49)

Pin 3 (junction of R4, R5, C1)

Pin 5 (junction of R10, R11, C2)

Pin 6 (wiper of R47)

Pin 9 (wiper of R44)

Pin 10 (junction of R16, R17, C3)

If the voltage at the input pins is correct, IC1 is possibly faulty


The voltage at both input pins on IC1 is normally the same for the IC output voltage to be normal.



It is difficult to test the voltage on the pins of IC1 without the risk of shorting an adjacent pin. The photo below shows other components that the pin is connected to where you can use them to take a reading from instead of the pin. All the same colour dots go to that pin. Make sure you use the correct end of the resistor, as shown by the position of the dot.







































Fig. SF1:  Places that can be used to test the voltage at IC1 pins




Replacing the plug / cable

See extending the sidelight cable for further information



One light partly working

Normally this would be all of one colour not working, or half of one colour not working - see below.



All of one colour not working

The red circuit will be used as an example. Refer to the circuit diagram for the equivalent components in the other colour circuits.

No red LEDs

Swap the lights over and try them in different sockets - red plug in white socket, white plug in red socket. If the other light now doesn’t work and the first one does, then there might be a fault in the wallwasher. If the first light still doesn’t work in the other socket, then the fault is somewhere on that light.


Confirm their is a red input voltage at TE7 on the light PCB. The reading should be about 16v at maximum brightness.

No voltage?

Disconnect J5 and measure the voltage between holes 2 and 3 (normally the brown and red wires).  If there is no voltage, the fault could be in the cable or the plug connection. Check them out and replace as necessary - see extending the cable on the sidelight page.


If there is a voltage on the PCB when you checked TE7, the fault is somewhere on the PCB.


Try adjusting pot R49 slightly to the left or right. (See the Calibrating page for details of how the pot works). If the red LEDs now light, R49 may be faulty - poor contact between wiper and resistor. Try the light on test for a while. If there red remains lit, the fault may be cured for now. If the fault comes back as a result of the faulty pot, then replace it.

Warning

Do not turn the pot to the maximum brightness - this may damage the LEDs, especially if a LED is already staring to fail


If there is still no red after adjusting R49, it’s possible that R49 may be open circuit. R49 can be temporarily by-passed by connecting a low value (between 33R and 68R) resistor between GND and the wiper of R49. Pin 2 of R49 is GND, or use TE8 which is GND. If the red LEDs now light, then R49 is faulty - replace it.

You can also test to see if R49 is open circuit by measuring the resistance between pin 1 and pin 2. This should be around 100R. If the resistance is higher than 200R, then R49 is faulty.


If the red LEDs are still not lit after testing R49, then

Test the voltage between the wiper of R49 and GND

Test the voltage at the junction of R4/R5/C1 - this is the red input to IC1 (pin 3)

the two voltages should be the same

If the reading at pin 2 and pin 3 are the same, then test voltage at the junction of R7/R8 - this is the red output from IC1 (pin 1)

If the reading at IC1 (pin 1) is wrong, an the readings at pins 2 and 3 are correct, it’s possible that the op-amp in IC1 associated with the red LEDs is faulty and IC1 will need to be replaced.



The diagram below shows a real example of the readings when the pot - R49 - is open circuit.

The symptoms are all of one colour not working.




Fig. SF2:  R49 open circuit




R49 is open circuit, or the wiper of R49 is not making contact. The input at pin 3 of IC1 is normal. Because there is no path to GND, the voltage at pin 2 of IC1 is higher than normal, and higher than that at pin 3. As a consequence, there is no output from IC1 and so T1 and T4 are switched off. Both groups of LEDs are off. The voltage at the LEDs will be the same as when the light is switched off.

R49 can be tested for an open circuit by measuring the resistance between pins 1 and 2. The pots have a wide tolerance, and the resistance may vary a bit from the nominal 100R. Of six tried, the resistance varied between 91R3 and 103R6.

One other test could be by temporarily connecting an ordinary 47R or 56R resistor between terminal 2 and the wiper of R49. If all the LEDs now light, then R49 is probably faulty.



If R14 was open circuit instead of R49, this would not affect the output of IC1 as R14 does not go to IC1. With the light switched on, the voltage readings would be the same as normal, except that the voltage across group 2 LEDs and TE11 will be about the same as those when the light is off. Only group 2 LEDs would be affected. Although T4 is switched on, there is no return path to GND for LEDs 15 - 16 and so group 2 LEDs will be off. Group 1 LEDs will be on.

R14 can be tested for an open circuit by measuring the resistance across it - it should be 47R.

Another test can be done by piggy-backing an ordinary 47R resistor across the terminals of R47. If the LEDs now light, then R14 is probably faulty.

Although a fairly large current passes through R14, it is less likely to fail than a pot. Eliminate all

other reasons for loss of one group of LEDs before replacing the resistor.



Only one group of LEDs working

This means that one group of LEDs is working, the other group isn’t. The red circuit will be used as an example. We’ll assume that group 1 (LD4 - LD1 is working) and that group 2 (LD15 - LD16) is not

The fact that one group is working proves that there is a red input and IC1 is working correctly. The fault lies with a component within that group. These are: R8, T4, R14, LD15, LD14, LD13, LD16 (in that order). If any of these fail, that group will not light.

The LEDs are connected in series, like fairy lights, so if any one of these becomes open circuit, the path with be broken and none of the LEDs will light.











Fig. SF3a:  All LEDS working                                 Fig. SF3b:  One LED faulty - all LEDs in group will not light




Current flows between VCC and GND, with the transistor acting like a switch. This switch is turned on / off via IC1.








Fig. SF4a:  No input -LEDs are off                                          Fig. SF4b:  Input - transistor switches LEDs on



You need to be able to identify where the fault is. Set up the light so that it is at maximum white. Check the reading at test point TE11. If it is showing 0v, then there is no VCC coming through the LEDs. This means that one or more LEDs are faulty, causing an open circuit. If there is a voltage at TE11, such as the voltage when the light is switched off, it may mean that the transistor hasn’t switched the LEDs on. Check the voltage at the junction of the output from IC1 (pin 1) and R7 and R8. If the voltage is correct, (which it should be if the other group of LEDs are lit), either resistor R8 is faulty (check for a voltage at the junction of R8 and T4) or the transistor is faulty. Test the LEDs as detailed on the Testing LEDs page, just to make sure s faulty LED isn’t giving a false reading. If the resistor test between TE11 and ground lights all the LEDs, then that proves the LEDs are OK


If the reading at TE11 is 0v, then you will need to test the LEDs to identify which one(s) are faulty.

If the junction with T4 and R14 is showing a voltage then R14 might be defective.



The diagram below shows a real example of the readings when one of the LEDs in a group (LD16) is faulty.

The symptoms are no group 2 red LEDs lit.





Fig. SF5:  voltage readings with LD16 faulty




Only one LED not working

If one LED in a group is not lit but the others are, it’s possible that the LED has an internal short circuit and, although it’s not lit, is providing a path for VCC to reach the next LED or transistor. Replace that LED.

Another reason for one LED in a group not lit is if there is a short circuit across its terminals. This is very unlikely, but check the PCB around the LED just in case.



Flickering LEDs - all one colour

If all one colour (both groups of that colour) LEDs flicker, then this could be for the same reason as for all of one colour not working, except that the fault is intermittent. Some people have reported that this only occurs when the lights are full on, but the light works normally if the light is dimmed. From reports, this defect mainly seems to occur on the lights in the wallwasher, but it has been reported on sidelights.


Although LEDs use a lot less current than normal filament bulbs, they do generate some heat. In a wallwasher, there is also the additional heat from other lights and components on the main control board. It’s possible that their may be a dry solder joint on the within that colour circuit on the PCB and the heat is causing the joint to expand slightly and cause a poor contact. However, as with no LEDs, my first suspect would be the pot. When the LEDs are on maximum, they are drawing the maximum current. This passes through the wiper connection and resistor surface on the pot. It’s possible that there is a poor connection between the pot wiper and the resistor surface, or the pot is faulty.

Warning

Do not turn the pot to the maximum brightness - this may damage the LEDs, especially if a LED is already staring to fail


Note - it may look as if all lights of one colour are flickering, when in fact it is only one group that is at fault. Because the faulty group is switching on and off, the other group of that colour also appears to be flickering. In fact is is not actually going on and off, but getting brighter (as the other group goes out) and dimmer (as the other group comes back).  The extra brightness of the good group is because there is a slightly higher voltage to the LEDs when the other group is off. I suspect that this is what later leads to the failure of the good group as well - the faulty group finally fails, leaving the good group with a permanent higher voltage, putting more strain on the remaining LEDs, until they either start flickering, or just fail completely. It is not always to see the difference between the two groups unless looking at the LEDs with the light cover removed.



See the information under All of one colour not working for the tests that you can do.



Flickering LEDs - only one group

If all one colour (both groups of that colour) LEDs flicker, then this could be the same reason as for all of only one group of LEDs not working, except that the fault is intermittent. The same comments made in flickering LEDs - all one colour - apply here as well.


This could also be due to a faulty LED in that group - either thee is a problem with that LED or it might have a dry joint. Try GENTLY pressing on the LEDs that aren’t lit. If the group of LEDs light when you are pressing on one of them, then make a note of which LED it was and continue pressing on any other LEDs. One of the problems with a poor connection like a dry joint is that it only takes a very tiny movement of the component or PCB for contact to be made / lost. Therefore, pressing on a LED may light all the LEDs, but it might not be that LED that is at fault as the pressure could be causing the PCB to flex slightly and the contact to another LED or component could be made  / lost. Try flexing the PCB VERY GENTLY to see if the LEDs now light.


If a LED is faulty, you could try the substitute LED test, although this might not work if the fault is caused by a dry joint. If you suspect that a dry joint is the problem and you know where it is, just resolder that joint. Beware of excess heat that could damage components, especially the LEDs - don’t have the soldering iron near the component for more than 2 - 3 seconds.  

See the information under Only one group of LEDs working for the tests that you can do.