There are many things that can cause that. For this problem, we can divide the bike into two distict electrical systems. The high-tension side and the primary side.
To determine if it's a primary side (or battery-side) problem, measure the voltage from the battery's negative terminal to the motor (with bike running). It should be preferably less than .5v. Measure the voltage from the battery to the frame. It should also be preferably less than .5v. And in your case measure the voltage to the carbs. It should also be less than .5v.
Then do the same measurements, but with the meter set to AC volts. The readings should all be near 0v.
If any of those fails, then you have a bad ground connection somewhere.
Now for the high-tension side. With the bike off and the coils disconnected (and sparkplugs disconnected), measure the resistance from the metal core (the mounting bar) of the coils to the high-tension output of the coils. If the wires are easily removable, then remove them to do the reading. Otherwise, measure to the end of the sparkplug wire where it connects to the sparkplug. If there is any continuity, the coil has an internal short. Do the same measurement from the primary contacts to the high-tension side. If there is any continuity, the coil has an internal short. Lastly, check from the primary contacts to the coils core. (Remember, these are all to be done with all of the wires disconnected from the coils and/or sparkplugs.)
The problem with these coil-tests is that the short may only occur while there is several thousand volts on it. That means the resistance test won't show the problem. In order to truly rule out the coils, you have to un-mount the coils (then reconnect the wires)and get the bike to run while the coils are floating and not touching the frame. Then check for your spark-to-screwdriver test.
Oh, and obviously, you need to fix this. Arcs and gasoline don't play well.
Post edited by: loudhvx, at: 2008/01/04 15:33