This can be a little daunting. If I were you I would use a sound level meter and a Stereophile test cd. Probably the first one "Stereophile Test CD". First though, I would try contacting B&W. They may have a way of IDing the polarity that is not readily apparent.

B&W drives usually have different sized posts on the drivers so they can be hooked up only one way because the slide on clips are also different sizes. If this is not the case look for a red dot on or near one of the posts, that would be the positive one. Every B&W driver I've ever seen had one of these two indicators, most had both.

I don't know if you can actually see this on a tweeter, but if you hook the tweeter up to a 9 volt battery and the driver pushes out, then whichever lead is connected to the + is the positive terminal and the opposite is the negative. I know this works with woofers but I don't know if it can be seen on a tweeter.

Please don't hook up a 9 volt battery to your tweeter. A tweeter with a nominal impedance of 8 ohms might have a DC resistance of 6.5 or less. 9 volts across this will produce over 12 watts, which doesn't sound like much until you realize that the true power handling of a tweeter rated for use in a high power system may be as little as 10 watts. The crossover network keeps most of the power away from the tweeter in normal use. Also, the maximum excursion may be less than a millimeter, very difficult to see, and potentially dangerous to exceed.

Were the speakers serviced at an authorized B&W dealer? I'd be very upset with them, and with B&W, for providing such shoddy service. I think a letter or phone call to B&W would be in order.

Do you have access to a microphone and an oscilloscope, or a 'scope program for your computer? With a wave file of an impulse you could try to duplicate Stereophile's impulse response measurement of the Silver Signature:

Note that John Atkinson said the tweeter is connected in inverse polarity to the woofer. It's barely noticible in the picture, but the first movement of the trace is in the negative direction.

Good luck!

Good to know about the power to the tweeter. Like I said, I've only ever done this with woofers. The polarity on your graph looks about as hard to see as the movement of a tweeter.

I also agree that B&W should try to be some help here. Someone there knows the answer.

Quote:
I like your suggestion about the scope program. I mentioned in my original post that I was looking for something like that. Any suggestions?

Sorry, I've never tried any, but Googling "oscilloscope software" produced some promising results, some even free.

You will clearly see a dome on a tweeter move with a 1.5V battery, so go for it. I've done this on about 100 tweeters, actually far more, so it is not problem -whatsoever. That's a 1.5V AA cell or similar.

There is no guarantee what the polarity of the given tweeter is in the given total speaker system, though.

It varies. I believe Martin Colloms reviewed that speaker for Stereophile, in the way back days.

Nope, it was JA. It was Martin who had the silver wired Celestion 700 set.

http://www.stereophile.com/standloudspeakers/272/

The review of the silver signature indicates reverse polarity for the tweeter.

Be sure the review is for the same system.

A note: sometimes the magnet structure or tweeter faceplate has the polarity markings and none on the dome assembly. Look for the polarity markings on those two other tweeter components.