Can the AKSA55 drive speakers that dip to 3.7ohms, but average just over 4oHms?
Regards,
Rookster
Rookster, low impedance is the GOOD part of the deal. Most half-competently designed amps will come to some sort of grips with a simple impedance drop to 3.7 ohms.
The problem is far more complex than that. The trouble is that impedance variations in ANY speaker, no matter what the price and the source, will not only drop to say 3.7 ohms, it will also show quite significant phase shifts as well, and worst of all, typically at the point where the impedance drops.
Ideally, an electrical signal will have both of its constituent parts, the voltage and the current, in ideal phase, ideally following each other's changes. A phase shift occurs when voltage leads and current follows, or vice versa, i.e. when they are not ideally synchronized.
If voltage leads, we say we have a positive phase shift; if current leads, we say we have a negative phase shift.
And this is where the amplifier's troubles REALLY begin. We measure and express phase shift in degrees. Not to bore you with the maths, let me just say that a phase shift of 45 degrees is very common in speakers, and 60 degrees is not unheard of, though not very common either. When you work it out, -45 degrees is a factor of 1.41, and a -60 degrees is a factor of 2.
So, if by some misbegotten chance that 3.7 ohm drop is coupled with a common phase shift of -45 deg., the actual equivalent impedance the amp sees as its working load is (3.7 : 1.41) 2.62 ohms. And if it should be -60 deg., the equivalent impedance will be halved, just 1.85 ohms. This spells trouble for the amp; it will have much difficulty riding over that one, no matter what it is.
To ease your mind, the now legendary Apogee speakers had an impedance drop of 2.4 ohms with a -60 deg. phase shift, making the amp see 1.2 ohms as its load - practically a short circuit. No wonder they required behemoth amps to make them come on song.
Now, Hugh's approach is to let his amps work into low loads and he doesn't fear that for two reasons:
1. His output stage and power supplies can ride it out, and
2. He knows this is not a steady state signal, but in effect short term pulses.
If this was steady state, Hugh's amps would overheat at best, as would the vast majority of amps on this planet, or their overheat protection circuits would trigger and siconnect them. As things stand, they will ride out such dips and go on working.
I ran into this problem myself some years ago when I developed a power amp which was then, and still is today, my favorite project. It is capable of riding out peaks of as much as 130 amps (the largest Krell can do "only" about 85 amps), assuming the power supply could deliver it, which of course it couldn't. But in the end, the problem boiled down not to the power supply, which you can beef up any time you want to, and I was already using 800 VA toroids for a nominally 100W/8 ohm amp (did I mention I was a power supply freak? No? Well, I mention it now!). The problem was in the heat produced - I ended up needing inordinately sized and priced heat sinks, or go the forced air cooling (fancy term for fans). I hate fans, so I had to make do with the knowledge that it could theoretically deliver this.
The last say I had on the subject was in writing the design brief for my speakers. They were developed by a friend of mine, then in the speaker manufacturing business but sadly no more today (see the test at my site,
http://www.zero-distortion.com , B&M Acoustics 1041). Anyway, that speaker has no less than 6 ohms, no more than 14 ohms, and the worst case phase shift is -25, +20 degrees. Coupled with an efficiency of 92 dB/1W/1m, a teeny-weeny tubie job of 8 watts can drive them easily, let alone 50+ watt per side jobs.
The speaker took 3 months to come to life, but developing the crossover for it to deliver such ease of drive took 6 months all of its own.
Cheers,
DVV
