A little insight from Ayre Acoustics founder.
“This is another reason why I've not experimented with fancy fuses -- the penalty factor for any sort of problem is pretty high...
Real fuse companies rate their fuses by a quantity called "I^2t", pronounced "eye squared tee". This is an amount of energy that is the current (in amperes) squared times time (in seconds). This puts a quantitative value onto a subjective label such a "fast blow" or "slow blow". So for example, we can look at the Littelfuse website and see that their Series 218, 5x20 mm, 1 amp, slow blow has an I^2t value of 6.73 A^2sec. This is what we use in our products.
http://www.littelfuse.com/~/media/Electronics/Datasheets/Fuses/Littelfuse_Fuse_218_Datasheet.pdf
Now if we go to Digikey, who sells a large number of brands and put in the same specifications -- Glass Cartridge, 5x20 mm, 1 amp, slow-blow, 250 VAC, we can find I^2t values ranging from 2.4 A^2sec to 28 A^2sec, a variation of greater than ten to one! So a 1 A slow-blow fuse is
not a 1 A slow-blow fuse.
The peak surge energy drawn by any specific piece of equipment will depend on many factors. First of all the largest surge occurs at turn-on. There are two main culprits. The first is the transformer itself. If you put a multimeter across the primary leads, the DC resistance will generally read in the range between a few tenths of an ohm (for a large transformer) and a few ohms (for a small transformer). If this is all there were to the situation, the transformer would be more like a light bulb, glowing red hot until it melted. But we are not sending it DC during actual operation but instead AC. Now the inductance of the transformer comes into play and limits the current. Every transformer has some "leakage inductance". This is inductance in the primary winding that is not coupled to the secondary winding, but instead creates a stray magnetic field around the transformer. This requires energy, and hence some current draw. At turn-on, it takes a fraction of a second to build up this magnetic field, and while it is doing so, it draws more current than normal.
The other culprit are the filter capacitors. These are charged up from zero volts to their full operating voltage. usually within about ten cycles of the 50 or 60 Hz incoming voltage. For the first half cycle, the capacitor is virtually a dead short and the only thing that limits the current are the stray resistances of the internal wiring, the primary and secondary windings of the transformer, the fuse resistance, and the impedance of the rectifiers, et cetera. We have a Tektronix current probe that connects to an oscilloscope and we were measuring the peak current at turn-on for the new VX-5 power amplifier, and even though it idles at about 2 or 2.5 amps, the peak turn-on surge was well above 50 amps!
In this case the only thing to do is use a higher value for the KX-R. You can ask Synergistics what the I^2t value is for their fuses and get one that is at least 7 A^2 sec. I'm sure it won't hurt to go to 10 A^2 sec. However I would be surprised if they have that data. If not, just go to the next larger size. The I^2t values climb
very quickly. For example the same Littelfuse in a 1.25 A rating has roughly
double the I^2t as the 1 A fuse, and the 1.6 A fuse has roughly
double again the I^2t rating. But at $30 a pop, it is pretty expensive experimentation...”
Good luck,
Charlie Hansen
Ayre Acoustics, Inc.
