Playing the modular through my home stereo

[Magnus Danielson]

From: JWBarlow at aol.com
Subject: Re: Playing the modular through my home stereo
Date: Wed, 1 Mar 2000 22:45:24 EST

> 
> In a message dated 3/1/2000 3:47:11 PM, pfperry at melbpc.org.au writes:
> 
> >Only time I blew the speakers out, I was using a r*land
> >D*gital soundcanvas card & running a midi file shifted
> >down 4 octaves.
> >This gave very low freq square waves.. I couldn't hear 
> >anything except this 'clicking' sound.. thought the gain was
> >too low.. turned out what I was hearing was the cones bottoming
> >out on squarewaves at 5 Hz or so.
> 
> What I've wondered about for a long time is the effect of instantaneous 
> transitions on speakers (even in the audio range). It seems hard enough to 
> get electrons to move like saw and square waves, let alone speaker cones. 
> Anyone know how traumatic this is for speakers?

There are several things that can happend, but here are some faults:

Xmax overstress.
        Xmax is the limit in cone displacement the speaker can widthstand.
	When you pass beyond Xmax you will overstress the cone suspension,
	coils crashing into the magnet at high velocity etc. If you are unlucky
	the mechanical pull from the suspension together with the pull from
	the coil and the accumulated heat on the coil can make a week coil-cone
	interconnect break - with the probable result that the coil flies into
	the magnet.

Coil overheating.
        The coil can quickly accumulate heat, since it is sitting in free air
	basically it must radiate this heat to its surrounding, this is easilly
	done when the whole speaker assembly is cold since then with the heat
	flux be good due to the temperature difference. However, as the magnet
	assembly, which is the primary target of this heat, gets warm the heat
	flux reduces due to the smaller temperature difference, this can
	happends after several hours of concert for instance, naturally will
	the temperature of the coil rise. Now, as the temperature rises the
	coil changes properties and its resistance has rised for instance.
	Also, now the coil is MUCH more vunreble to excess energies since there
	is basically nowhere to radiate the excess energy. Any energy hitting
	the element at this point we should be able to radiate. Add to this
	fact that for high frequencies will the air volume in front of the
	element be very reactive, so you will not emit much energy there, and
	for low frequency you will also have a very reactive load so no energy
	is being emitted there either... thus, the impedance mismatch will
	cause the energy to be eaten by... the coil!

So, this really suggests that you:

1) Have control over the frequency range where you insert energy into your
   element (notice, we are talking element-by-element here, not speaker).

1a) You want the cross over filters to be well adapted to your element, so
    that you only try to use the part of the frequency range where the element
    migth provide a good energy transfer for you. This is good practice both
    from a element survival and total responce view.

1b) You want cross over filters that are stable, does not change with applied
    energy and that can be pretty sharp. This suggests active cross-over
    filters prior to the amps.

1c) You want to ensure that your amps NEVER can clip, this means that the amps
    shall be able to deliver MORE power than the elements are marked for before
    they clip. Yes, your amp should be able to basically overheat the element
    with PURE energy (say sine) before it clips, because when it clips you
    get MUCH more energy and not only more energy, also with overtones into
    areas where the element simply does not know where to waste it. Clipping
    amps is a major contributor to wasted elements. Propper sizing of the amps
    is a good thing.

2) You would like to control and limit the amount of energy you insert into
   an element. This becomes a per-element (-type, thus, per register) limiter.

2a) You want the limiter setting be such that you never can overheat it (1c).
    This is really a moving target since you have to consider the energy you
    allready have pumped in.

2b) You want the limiter setting to be such that you can ensure you do not
    exceed Xmax. Engineering this can be a bit difficult, but manegable.

3) For all these measures, you really, really need a good view on your whole
   gain/level/power chain from the cross-over/limiter to the element. You
   really need to know how your limiter level relates to your poor element's
   treshold of giving up. While at it you can do neat trimming of your dynamic
   range capabilties.

Now, doing it this way really work and really pays of. Take it from one who
have been engineering catastroph-limiters and cross-overs for larger PAs.

Interestingly enougth, if you go out and by a pair of Genelecs, they provide
this all solved into one box (well, maybe except for the 1035s where the
electronics goes on the side). The only thing you need to care about is not
to clip the inputs.

Oh, but the cone, how about the cone?
If you have an old paper cone, then you could break it, but modern cone will
just "break up", that is get momentarilly bent into one of many patterns. Those
bendings relate to energy inserted and are just the resonance flexings. When
you hit a resonance in the element you will not emitt much energy, the same is
naturally true for the speaker assembly.

Further, a speaker which has a good impedance matching to the air (like in a
good horn speaker) then will much more of the energy get out into the air and
not be a nuance to speaker survival. Also, you will get more bang for the
buck ;)

So, in the end, no "strange" energies (like DC, RF, other regiser material)
into the elements is a good starter.

Cheers,
Magnus