Xbee Adapter Board #32401 - Connection Problem Resolved
Greg P
Posts: 58
Attached is a slightly more intuitive schematic of the Xbee adapter board #32401. It lacks buffer chip 244A pin # assignments but otherwise should be complete.
PROBLEM:
When a Xbee was plugged into the board it would not attach to the wireless network (node discovery failed). This was true of multiple Xbees plugged into the board. If I retracted the Xbee pins from the board socket by about 1/8" proper activity was restored.
OBSERVATIONS:
The path from the DIP ground pin to the XBEE ground pin was several 100 M-ohms when operation failed, but a good connection when operating correctly. All direct ohmmeter tests of the 2mm receptacle strip on the adapter board proved good. The customer would have no clue from just these ohmmeter tests that a problem exists ! Various wires inserted into the 2mm receptacle strip always exhibited excellent conductivity and showed no sign of physical obstruction.
SOLUTION:
With the Xbee plugged into the board I applied my solder gun to solder joints of the 2mm-spacing connector on the TOP SURFACE of the Xbee for about 3 to 4 seconds per joint. My first solder gun heating attempt was applied to the ground pin. This restored continuous wireless operation regardless of Xbee insertion depth immediately. I noticed other LED indicators (Tx & Rx) were also "flaky" in their behavior, so I repeated the solder gun heating operation for all Xbee pins. Immediately, the LEDs showed stable behavior.
BEST GUESS:
The underside (nearest the PCB surface) of the 2mm connector strips that the customer must solder to the adapter board is open. My best guess is that flux from my solder penetrated from the solder side of the board (through the plated-through holes) and accumulated near the base of the 2mm connector strip on the component side. The tip of the Xbee pins would press against this flux when fully inserted and this in turn would provide a lateral force for pushing the metal spring clips away from the pin, creating the open circuit condition.
The brief heating of the Xbee pin may have been sufficient to evaporate most of this flux.
I would not recommend this solution during low humidity conditions as there is the slight risk of static discharge and damage to the Xbee. Best to insert a short lead of solid wire into the socket and heat the wire with the soldering gun.
This is just my best guess at what happened. I'm providing this information for the benefit of others which might encounter the same problem. Theoretically, a 2mm connector strip with a CLOSED configuration at its base (PCB contact side) would not allow flux to flow up and into the metal spring clip area. If this part is available, it would be nice to substitute the closed connector for the existing one.
PROBLEM:
When a Xbee was plugged into the board it would not attach to the wireless network (node discovery failed). This was true of multiple Xbees plugged into the board. If I retracted the Xbee pins from the board socket by about 1/8" proper activity was restored.
OBSERVATIONS:
The path from the DIP ground pin to the XBEE ground pin was several 100 M-ohms when operation failed, but a good connection when operating correctly. All direct ohmmeter tests of the 2mm receptacle strip on the adapter board proved good. The customer would have no clue from just these ohmmeter tests that a problem exists ! Various wires inserted into the 2mm receptacle strip always exhibited excellent conductivity and showed no sign of physical obstruction.
SOLUTION:
With the Xbee plugged into the board I applied my solder gun to solder joints of the 2mm-spacing connector on the TOP SURFACE of the Xbee for about 3 to 4 seconds per joint. My first solder gun heating attempt was applied to the ground pin. This restored continuous wireless operation regardless of Xbee insertion depth immediately. I noticed other LED indicators (Tx & Rx) were also "flaky" in their behavior, so I repeated the solder gun heating operation for all Xbee pins. Immediately, the LEDs showed stable behavior.
BEST GUESS:
The underside (nearest the PCB surface) of the 2mm connector strips that the customer must solder to the adapter board is open. My best guess is that flux from my solder penetrated from the solder side of the board (through the plated-through holes) and accumulated near the base of the 2mm connector strip on the component side. The tip of the Xbee pins would press against this flux when fully inserted and this in turn would provide a lateral force for pushing the metal spring clips away from the pin, creating the open circuit condition.
The brief heating of the Xbee pin may have been sufficient to evaporate most of this flux.
I would not recommend this solution during low humidity conditions as there is the slight risk of static discharge and damage to the Xbee. Best to insert a short lead of solid wire into the socket and heat the wire with the soldering gun.
This is just my best guess at what happened. I'm providing this information for the benefit of others which might encounter the same problem. Theoretically, a 2mm connector strip with a CLOSED configuration at its base (PCB contact side) would not allow flux to flow up and into the metal spring clip area. If this part is available, it would be nice to substitute the closed connector for the existing one.
943 x 756 - 57K
Comments
I haven't use Parallax's yet, but the one's I've used (from SparkFun) can very easily fill with solder. I've ruined several female header this way, even when I'm aware of the problem and trying to be extra careful, I've still managed to ruined several female 2mm headers.
http://www.sparkfun.com/products/8272
Also, look at this photo of an extreme example:
http://picasaweb.google.com
/marc.bayerkohler/Misc#5256440319164643730
One reader offered the following remark:
"You should solder headers at around 250-325F. Also, on these types of headers, just tiny piece of solder wick at the base before soldering it, so any solder dripping down would be absorbed."
I presume the cooler tip would limit the distance at which the solder would remain liquid and reduce penetration of solder through the plated-through hole into the spring assembly. The solder wick ideal is interesting ... but seems like a difficult task if doing a dozen boards pin by pin !