I think it would be worthwhile to read my 'cut & paste' from a thread in TGP. I left a link to the original thread below. Read the whole thing. It's interesting to see how SS and tube amps differ in this regard.
Originally Posted by TENAX
the speaker resistance the amp sees at the load (speaker end) should always be more than the amps output ohm rating..so for example, 8 or 16 if you have a 4, 8 , or 16 ohm output options on your amp..you could then use for example:
4 or 8 ohm amp outputs with 8 or 16 ohm speaker resistance...or...
4 ohm output with a 4, 8 or 16 ohm speaker load resistance.
i have a valve jr that can do either 4, 8 or 16 ohm output. i have a 2 x 12 with 2 16 ohm speakers in parallel so it's an 8 ohm speaker load. so i would use either the 4 or 8 ohm outputs on the amp..but not the 16..
speaker load always to be matched or higher than the output on the amp or it could cause too much current to flow in your amp and fry your transformer.
easiest way i find to look at it is why is it said to always make sure you have a speaker hooked up..or a "dummy load" when you turn on an amplifier?" because otherwise, there would be no resistance in the circuit so say a transformer designed for 4 ohm resistance is presented with zero load with no speaker connected..too much current flowing..flames too follow quite possibly
Here's the response....
Sorry, this is absolutely wrong (again - every time this question comes up someone posts this)... especially the bits I highlighted.
It is is true for solid-state amps, as is the reason - causing too much current to flow - but it is NOT true with a tube amp, as the mechanism for damage is entirely different. Excess current is not the problem, as it's limited by the loss of efficiency of the OT as the mismatch becomes more extreme. Flyback voltage causing insulation breakdown is the risk, and occurs when the load is too high impedance, not too low.
You're also confusing 'no load' with 'no resistance' in the last paragraph.
'No load' means infinitely high impedance - an open circuit - and is dangerous for tube amps (although almost never for solid-state ones). 'Zero impedance' - a short circuit - is in fact usually safe for a tube amp, at least for a short time (although it's very hard on the tubes), but will fry a solid-state amp.
Generally, tube amps are safe with between half and double the correct matching impedance. Lower is more risk to the tubes, higher is more risk to the transformer and other parts, and it is NOT a good idea to run a 4-ohm tube amp into a 16-ohm load. If you must mismatch, it's safer to go lower rather than higher, with most tube amps.
Here's the link to the thread at TGP:
http://www.thegearpage.net/board/showthread.php?t=274034