Notes and pictures on the diagnosis and repair of a specific outdoor digital display sign manufactured by AMS Adaptive Micro Systems.

These two panels were purchased in the early 2000's. A couple years ago both failed and since the manufacturer had discontinued the model they no longer supported this model and they recommended purchase of a current model (for thousands of dollars).

The owner found a person who could fix the signs but one of the panels failed again a few months after it was repaired. A couple years have passed and the owner (a close friend) asked me to see if I could have a look.

He took them down from the pole. These are heavy and in the process, the panel that worked slipped off the upper brackets and flipped down on the lower brackets. The only damage that was apparent was bent lower brackets but when the unit was powered up (see above) one of the LED modules was no longer operational. "GREA ----- ERVICE"

The repair process described on this page is very specific to the this model sign and these particular symptoms and must not be construed as general repair procedure for this type of sign or circuit. Please read and understand all safety warnings on the sign and described in the accompanying manual(s). Do not attempt any repair on this sign unless you have a strong understanding of electronics and electronic circuits and skilled in electronic fabrication and repair techniques.

Some views of the disassembly of the sign panel.

Removing the plastic 'lens' cover exposes the LED modules inside. This model has six modules - two of which have been raised and rotated out of the way. I had to disconnect the data cable between these modules (DB9 connectors) in order to get enough slack to move them out of the way so I could see the power supply and control boards underneath.

I didn't need to touch the I/O board or this toroidal transformer. This is the first time I've seen this transformer so I had to look it up on line to learn it's function and properties. These guys are used where there is a need for high current and I got distracted on web sites relating to Tesla Coils.

Back on task - it's fun to look at these LED panel module control boards. It's a simple switch circuit replicated hundreds of times and compressed onto a single board. ( I would imagine that by now these components have been reduced to a single chip.)

Finally I get to the power supply board where the trouble probably lies (with the panel that was completely dead. Actually it wasn't completely dead; when powered up the panel fans would start and run for about 5 seconds before shutting down. None of the display LED's would light (various power up messages) and the unit would remain quiet and dark.

On with the repair.

After disconnecting the power input lines from the toroidal transformer, the power supply board can be tilted up to inspect the underside. Here is evidence of the problem and a previous attempt at repair (the right idea but not finished for some reason.)

The design of this board had the supply current coming from those three 12ga. wires to the transformer feeding the power circuit through this thin copper trace. What happens is the trace under the screws begins to degrade over time and contact with the screw becomes less efficient. Since this corrosion causes higher resistance to current flow the area heats up which degrades the copper even more and pretty soon you don't have enough current flowing into the supply.

The solution is to bypass the copper foil and provide a heavy conductor between the transformer connections and the input (diode) to the power circuit.

Here is a slightly blurry picture showing the screws removed (which feed through the board to the terminal strip connections to the transformer). The traces under the screws had oxidized away leaving only a fraction of the copper to conduct the current. I cleaned the area of burned coating and with steel wool I burnished the copper in preparation for soldering.

Form a jumper from 12 ga. copper wire. I used the mallet to flatten the jumper so it pressed evenly against the board (there is not a lot of clearance under this board when it's mounted on it's stand-offs).

Here the jumper has been installed on the Power supply board. After cleaning all surfaces with steel wool and rosin, I soldered jumper to trace and to the lead for the input diode. I also laid solder around each screw head onto the underlying wire and remaining trace.

This power supply is now ready to re-mount in the unit.

All that remains is to replace the cover and this unit is ready to hang back on the pole.

Before doing this I checked all the fans, cleaned the vents and cleared out all the spider nests (with a couple nasty looking spiders) from the cowls that protect the fans.

The second unit - that worked until it was jarred when removed from the pole - got the same treatment to its power supply.

One of the LED modules had jarred loose from it's power connection - this was re-attached.

The cowling that protects one of the fans had been bunged a bit which jammed the fan. A little bending and a new screw holding the fan fixed this.

-- GREAT SERVICE --

One thing still has us stumped relating to these signs. If you have the answer, we would all be very grateful for an explanation - and a fix: Each sign has memory enough to hold 35 different messages (named A-Z, 1-9). The manual says you can program to display any of these messages based on time of day. We have followed the programming instructions and the sign keeps displaying the default even though the set time has come and gone. This is important because Madison, Wi. has an ordinance that prohibits a sign from changing it's message more often that once every 5 minutes. Of course this makes all the fancy wipes and reveals worthless. And the sign will cycle through messages continuously (this part works) but that too is worthless in Madison.

If you have a solution for this problem please email me. Thanks.