2 things would help the 6290 fan situation.
First thing is that the amp was designed to be stacked and racked and left on for a bazillion hours SO airflow through it's chassis was designed not to interfere with other rack components ( at least too much ).
If you removed the top and bottom covers, and cut out ALL the metal over *and under* the heatsink fins, and gave this a rack space above and below it would run a lot cooler and the need for high speed might never present itself.
If it did it would be the LEAST noticeable volume in that room...
Second is the value of the low speed resistor. The fan is a standard 120V bearing type and it needs to see 120V at start up then drop to the low speed otherwise it won't engage ( esp if worn ).
The stock value can be increased to make the fan spin slower OR not unless the "magic" temp is obtained ( rewiring required ) for home use.
I have 9 of them here from an install and later this month will be doing my analysis on this issue and will post a DIY thread.
There is also the issue of bias voltage but that is a little more advanced BUT easily checked / adjusted. This amp series tends to drift up with age and that WILL cause heat buildup and early fan engaging.
sub