how to wire separate power for motor shield

I am trying to configure this NKC motor shield for a Duemilanove Arduino. The attached image shows that the VIN and ground pins for powering the motors are sticking up. The 5v and ground pins are sticking down, so that they plug into the Arduino. I have tried plugging a 9 volt DC transformer into the Arduino to provide it power, then plugging in an independent 6 volt lamp battery into the VIN and ground pins. The yellow PWR light on the motor shield does not come on and the B and A pins that should be powering the motors do not provide any voltage. By the way, the Arduino is running the standard "B" motor test code that is at the end of the NKC build instructions. This code causes the B pins to turn on and off every few seconds, causing the motor to turn and then stop repeatedly. If I solder the pins all pointing down and sticking into the Arduino so that the Arduino powers both the board and the motors, the B motor pins get power and runs the B motor as expected without any problem. For starters, can anyone confirm that the power pins are configured up and down correctly? If that's true, any clues as to why the A and B pins aren't getting any voltage (off and on, depending on the Arduino test code I'm using.)

motorshield.JPG

Looking here: http://mcukits.com/2009/03/12/assembling-the-freeduino-arduino-motor-shield/

It would appear that your power pins are installed upside down.

Well obviously the pins are the other way up to enable connection to a separate power source - that's what the posting is about!

The schematic suggests this should just work - the grounds are commoned on the shield so providing motor power to the sticking-up pins should work, though I'm not sure its a good idea to be able to power the Arduino up and not the motor supply.

I'm wondering if the polarity of the supply is correct, the current is sufficient? A little checking of voltage rails with a multimeter ought to establish such things.

Powering a motor shield via the Vin header isn't a great idea anyway as those headers are not rated for much current, its asking for trouble.

Went back and double-checked the polarity being supplied correctly to the power pins for the motors--the ones sticking up from the shield in the image. Didn't make any difference. I did notice that the B pin, which the test code is currently cycling from low to high, goes from -.04 to +09 volts, as though power from the arduino was somehow leaking into the B pins from some connection. Also, you mentioned that the ground for the 5v and ground for the VIN were connected. I'm not sure that's true. One ground pin sticks up and one ground pin sticks down. Don't think they're connected. I could solder between them to connect them electrically. I should mention that the motor shield does have a 0v pin on the far side from the motor pins that sticks down and connects to GND pin on the Arduino, next to pinhole 13. This might provide a common connection or not. I can't tell.

Good thinking..

Discovered the VIN pin was no longer delivering power to the VCC2 pin of the SN754410 chip. The resoldering of the pins in the upright position might have caused this. Once I soldered an external wire between the VIN pin of the shield to the VCC2 pin, the motors were powered by an separate power source as expected. I did take the trouble to solder a bridge between the 5v ground and VIN ground pins as required by the original instructions.

After all that, I don't feel I got much "boost" in amperage from the motor shield. It can only drive up to 1 amp, or so the original instructions claim. I think the main advantage of the shield is that it can drive two toy motors simultaneously, controlled by the arduino.