Quoted from jimbeam42:

Right number of balls installed?

Quoted from pin15:

Mine was doing this after mist multi ball. Everything was working in test mode and game play. In researching the topic on pinside- they said could be a flaky opto- I replaced the mist opto on the right side- fixed the problem. Again it worked and tested perfect in game play etc. Just did this couple nights ago- was bugging for a long while. Good luck with it- great game.

Quoted from phishrace:

Before you tear anything apart, run the magnet (Mist) test a few times and watch for errors. While watching for opto errors (on the display), also watch the trough switches on the display for anything funny there. The optos will give an error if there are problems, but I'm not sure the trough switches will.

Quoted from pin15:

http://pinside.com/pinball/forum/topic/bsd-game-shuts-down-sometimes-after-multi-ball
Here is an earlier post that may help.
Good luck- I know your pain!

Quoted from ryan1234:

This was happening to me and was REALLY frustrating. It only happened occasionally and I was sure my Dracula was haunted. I found that it was caused from weak solder joints on the opto board(under the playfield) . The long opto worked correctly in switch test... but if I put a little pressure on the playfield it would trigger. I reflowed the connector joints and the problem has totally gone away. It is worth a try and cant hurt. Good luck

Quoted from ryan1234:

Yes-
The long range opto board.
It only takes a few minutes to remove and reflow. definitely an easy place to start.
My personal guess is your problem has to do with the long range opto beam. It could be the sender, receiver, wiring, or board. When "mist" mulitball is running it looks at the long beam to see if the ball was knocked off. When the ball IS knocked off the magnet the switch will be "closed" with nothing blocking it (the receiver sees the transmitter beam).
The following scenarios are seen as the same to the CPU and the program will assume that the ball is blocking it.
1. The long opto works and is blocked by a ball.
2. The long opto- sender, receiver, opto board, or wiring, are faulty.
The program assumes the last ball is in the mist gates and ends your play.
Mine would run the "mist test" perfectly every time- but during a game the vibration from other balls and flippers would cause the problems. I could slightly press or shake the playfield and get the opto the open(as if something were blocking the beam.) I used the "switch test" (not mist test) with no ball in the path of the long opto and ran my hand back and forth to test the switch. try shaking the playfield. With nothing in the way the opto should always stay closed.

Quoted from millerjl:

It sounds like you're already barking up the right tree, but just in case you need to keep digging never underestimate the power of putting the game in switch test, closing the coin door and hitting the flippers for awhile. If you really want to work on your coordination try it with a ball as well (an assistant to watch the switch test display on the DMD can help here quite a bit as well). Obviously pay special attention to the trough switches and the long range opto on the matrix.

Quoted from phishrace:

You haven't mentioned the shooter lane switch. You'll want to check that switch closely too.
Since they all seem to be working in test, I would suggest checking the trough and shooter lane micro switches for borderline switch activation. Meaning when a ball rolls over it, it barely triggers. If one is close, adjust the actuator so it triggers earlier. Pounding on the playfield or flipping the flippers may not trigger the iffy switch, but balls being juggled around in the trough might. A switch being too sensitive (triggering too early) may also be part of it, but I would suspect a borderline switch first. See when they activate (slowly move a ball over each switch) and get them all activating at about the same point.
I would also check the entire trough for anything out of place. Seems like I remember BSD having a gate there to prevent balls from going backwards. Make sure all the parts are there and none are bent. IPDB has lots of good pics to compare to.

Quoted from phishrace:

If you've wiggled the wires to those switches while in switch test and not found anything, I'd probably go ahead and replace the switches. Especially if they're factory. Diodes too.

Quoted from kevinleedrum:

None of the sling switches or flipper EOS switches double-register as trough switches in switch test? No problems from the outhole switch or the other trough ball switches?