RGBA Wash - One channel dimmer than the others

Device:

Pixel Range PixelLine Micro Wash RGBA        

Issue:

Green LEDs considerably dimmer at 100% compared to the other colours.

Green is at full and yet completely lost

This lighting fixture is one of a six units that our theatre received for free in varying states of functionality. After a bit of love and care most of them only required some re-seating of their internal connections and some new fuse holders to get back into working condition.

This lighting fixture is one of a six units that our theatre received for free in varying states of functionality. After a bit of love and care most of them only required some re-seating of their internal connections and some new fuse holders to get back into working condition.All but one were able to go into service. When comparing their total light output against each other it became quite obvious that one of the fixture's green LEDs were not getting as bright. This suggested a more technical problem and as such and this one sat on a shelf alone for 6 months. 

PixelLine Micro W

But then lockdown happened and once our venues were able to open up again LED units were suddenly in greater favor than their filament counterparts due to the lower running costs and so a new found sense of urgency arose to get this fixed.

5 minutes of unscrewing was all it took at to get into this unit. Everything is neatly packed inside and so unfortunately the first thing to do was chop off the cable ties in order to allow better access to each circuit board. 

Inside the unit

Now that we can see things better its time to do some problem solving and work through the signal path. By the fact that the LEDs come on at all means we know that they are receiving power. It's also clear that DMX is being transmitted as we have control of the green LEDs so the next thing to check is that they are receiving the full range of DMX values from 0 (OFF) to 255 (FULL). The easiest way to do this is to cross-plug another colour with green and see how the problem changes. If it changes to the other colour then it must not be decoding the correct DMX value as we know they are able to reach full when wired normally.

If the issue remains with the green LEDs then we know they are receiving DMX correctly and something else must wrong. 

From the image below you can see the outputs from the DMX decoder are on the left with coloured red, green, blue and yellow wires. We simply swap the green cable with another, power up the light and send the green channel to full. For this example I swapped with blue.

DMX arrives from the right, passes to the micro-controller (green label), and then outputs on the left

Sending green to full took our blue LEDs to full brightness and sending blue to full brought our green LEDs to their usual dim level. This test confirms that DMX is being received and getting decoded as intended. The issue must lay somewhere upstream of this stage.

Looking deeper at the unit I discovered another circuit board sitting behind the LEDs and after gently removing the diffusion cover, paying special care in not spilling the individual lenses all over the floor, new components revealed themselves for investigation.

It seems each colour had its set of these components in the corners of the board consisting of what look to be tiny resistors and something larger. Fortunately the label is quite clearly read and a little googling reveals the black component to be a LM1117 Voltage Regulator. These types of component are used to provide a stable power supply voltage independent of load impedance, input-voltage variations, temperature, and time. The accompanying resistors are likely used to set the value of the output voltage.

The LM1117 Voltage Regulator

This is good progress as we can now use the pinout on the datasheet to test that they are all correctly regulating the voltage passing through. The right pin takes the input voltage and outputs it through the center pin based on the resistance through the left pin.

Note - The next step requires turning on the LEDs which, without their lenses, get very bright and I do not recommend looking directly into them. At this stage I covered each row with a strip of LX tape.

Powering up the unit once again I sent all 4 channels to full and broke out my multi meter to measure the voltage of the output pin to ground.

Red, blue and amber were all outputting 7V but the green regulator was putting out 12V. Almost double! This must be where our problem lies. On closer inspection after this test I noticed that the soldering on this component looked a lot "blobbier" than the others, suggesting that it had potentially been changed in past.Clearly it must have been replaced with a faulty regulator, right?

...well I was able to find a replacement part on eBay and was almost ready to checkout when I decided to see if anything happened if I simply re-flowed the solder. Out came the soldering iron and 30 seconds later I had the solder joints looking shiny once agian......and what do you know....I send the greens to full and for the first time they actually went to FULL brightness. With the multimeter at hand I checked the output voltage once again and this time it was reading 7V like all the rest.

After reassembly - the greens are back!

What a simple fix that turned out to be! A couple of pound saved and another light back into service.

All LEDs equally balanced

Problem solved!