I can't believe, that in this day and age, that any amplifier is built that doesn't have proper protection in place to protect itself against faulty wiring. This has existing for ages now, and has been incorporated by all the top tier manufacturers to ensure that when you buy their $,$$$.00 amplifier that they don't get repairs to deal with because you crossed up some wiring on day one.
What an absolutely insane thing to tell a customer. If you got a dead short, I would ask first "You mean your amplifier doesn't handle a short in the wiring like every other manufacturer does?"
It's not a minor detail, it's of phenomenal importance.
Unfortunately, even if your speakers are fine, there are about 100 ways a pair of exposed wire ends and speaker terminals, and amplifier terminals can short out. Minimal contact at any point would do it with any piece of metal. But, I haven't seen a new amplifier in years which aren't protected from this.
Not to pull you off site, but this was an interesting hit on Google for checking for Bryston dealing with shorts in the connection, and a response from a reputable member saying that there is short protection with the amps, and others chiming in with other manufacturers. Seriously, this isn't acceptable IMO.
http://www.audiocircle.com/index.php?topic=90327.0
There are many amplifiers that do not employ robust protection circuitry, as it is felt by some to impair sonic performance (relays must be placed between the power transistors and the speaker connection binding posts). Broadly speaking the more consumer-oriented the product, the greater likelihood it has some form of output protection. Many (probably most) DIY amplifiers do not employ any, but of course a DIYer is unlikely to short the outputs.
The more exotic, or "High End", or unconventional the circuit, the less likely that protection circuitry will be employed. Stuff sold at Best Buy will probably have protection circuitry as it reduces (obviously) warranty claims and the products are marketed to unsophisticated uses whom are not expected to have significant audio expertise.
Most amplifier protection circuits are designed to monitor output current, or temperature of the output devices, and to shut down or place in standby the amp if the limits are exceeded (such as by driving too low impedance speakers)*. Output protection relays age and contacts pit or degrade, so over time they may not offer the same protection as new.
However the biggest issue is that the consequence of output shorting is very fast acting, and in many cases the protection circuit cannot act fast enough to save the output devices.** Protection circuits can save you from amplifier damage under a wide variety of conditions but you cannot expect any output protection circuit to save you from a direct short.
The simple solution is to never adjust, connect, disconnect or otherwise fxxk with the speaker wiring while the amp is powered up. Shut it down, do whatever you want, check your work, then power it up.
If you want protection from short circuits of the output devices you can always look to the QUAD 33, QUAD 405, and similar "Current Dumping" amplifiers. The 33 (45 wpc RMS 8ohms) is "unconditionally stable" and will drive a direct short indefinitely without damage. The 405 is rated to survive a direct short for five minutes at full power (100w pc RMS 8ohms). Both amps were designed to safely drive QUAD's electrostatic loudspeakers.
When I was asked to spec and install systems in restaurants and bars for background music, I always used the 33 for the installation. With many rotating staff members and occasional ham-fisted renovators typically involved, you cannot insure a short will not be encountered. If a short was introduced the music would just stop, and a sticker on the system insured that they would call me. I never had an amplifier failure with that system, some still in operation 35 years later. Horses for courses.
But there are consequences to such a circuit topology, and although not considered "bad" amplifiers, they cannot compete sonically with the very best. They also require precise parts matching not normally found on less expensive power products. No longer manufactured by AMCL (UK) under new ownership after the death of Peter Walker.
* In order to obtain UL / CSA / TUV etc electrical safety certificates, amplifiers are required to operate without interruption into a specified load at specified power and distortion. Should the amplifier shut down during the test (protection circuit is activated) the manufacturer cannot rate the amplifier into the load and power that caused the protection mode to trigger. Thus you get amps rated for 8 and 6 ohm loads, but not 4, presumably because the amp shut down at full rated power into the 4 ohm load during certification.
** A relay appropriate for use to protect the output of an amplifier will have a reaction time of perhaps 0.5 ms. Mechanical relays could be ten times slower. A 20 KHz signal has a periodic time (swings through the crest of the sine wave twice, once for each + and - half-wave) of 0.05 mS. Thus the amplifier would produce full power into 20 KHz 20 times before the relay could trip. Any frequency above 1 KHz would allow at least one swing of the half-wave (ie reaching full power on the + power rail) before the relay could trip.
Use of relays may require snubbers (capacitors, resistors, diodes, etc) to become part of the output circuit. Thus there are those that prefer not to design with them employed at that location.
