Visual and electrical simplicity

[RobHolt]

Commenting on a thread recently about loudspeakers using very simple crossovers it struck me that there might be some confusion between something that looks visually simple being electrically simple. This isn't necessarily the case.

In the example above we were looking at a two way loudspeaker system using just a single cap on the tweeter. No padding resistor and most importantly no inductor.
Of course in that example, were you to express the circuit in purely electrical terms you would in fact see the various LCR elements - they just exist within the driver rather than being applied externally. Suddenly doesn't look so simple.

ISTM that the same applies with certain amplifiers.
Many of the hair shirt - 'look no filters' - designs do in fact contain them, they just aren't where you might expect to see them.
Take for example a classic example of the passive potentiometer feeding the 'wide open' power amplifier. You will rarely see mentioned that the passive pot and cable loom in fact forms a low pass filter, meaning that, as a whole, our wide-band amplifier 'system' suddenly isn't.

Interesting when you delve a little deeper than the seemingly obvious.

 

[felix]

Very good point.

A related issue in using the 'natural roll-off of a driver' - it also requires exceptionally graceful physical behaviour of the cone assembly, critical internal dmaping in the cone and so forth. Any out-of-band break-up also has to be very well controlled - so paper cones only, forget most plastics and don't even think about trying it with most metal ones.

Anyway, a few more examples to provoke discussion:

Naim's use of the speaker cable as the output inductor. IMO it's a very elegant solution - if you need to use longer cables, with more stray capacitance, which (any amp) needs more inducatance to stabilise - the cable itself provides it. Elegant, especially when compared with potentially hobbling the amps at some arbitrary large output uH value for universal guaranteed stability. The trade -off is specifying the right cable - which they do, and it's actually a very basic spec (you don't have to use NACA5 - but it's very good value for so much copper)

Interconnects on active pre-amps! It's such a simple thing, but staggeringly few preamps appear to be properly designed for capacitative loads (and yes, just100pF can very significant on small -signal, high-bandwidth amplifiers) It's very easy to fix, as little as one resistor per channel. But this is usually ignored, to make good numbers on reviews ('low output impedance') or just because of cost - or it was just not considered. Then users report swapping between even similar 'sensible' interconnects (shielded, or sheilded twisted pairs etc) all sound different 'when they shouldn't because 'this <pats box> can drive anything'. Well yes, but only becasue it aint entirely happy doing so. One single 50 or 100ohm resistor added between the board and the socket output could make the damn thing entirely cable-agnostic (even for bug-loco cable ideas).

One going the other way: a simplification which appear more complex than it really is:
Magic power blocks , 'hydras' - the fancy-earthing myth.
It's just plain good engineering - if you return the mains earth for all the connected boxes to exactly the same point, there can be no voltage difference between these wires. Therefore no currents can flow around the earth/sigbnal 0v loop (these arise from perfectly normal things like the difference in stray capacitance in different sizes of transformers). No 50Hz + harmonic noise current = lower noise floor. It is simply keeping the mains plugs physically and electrically close that make the difference - not the fact it was dipped in LOX and blessed by virgins, contains a fancy filter or isolation transformer and so forth.

 

[RobHolt]

The Naim cable example is a good one.
I've seen it said often that the lack of output inductor is an important aspect of the design on sound quality rounds. Of curse as you say, it does have an output inductor but its external.

Maplin used to sell a good and very effective Hydra substitute for little money.
Looked like a large, flat 3 pin mains plug with multiple inlets.

 

[YNMOAN]

Commenting on a thread recently about loudspeakers using very simple crossovers it struck me that there might be some confusion between something that looks visually simple being electrically simple. This isn't necessarily the case.

It seems to me that you are referring to a post I made. I'm well aware that the drive units posses their own resistive and inductive properties. However, this is true of all drive units. The fact remains that it is a less elegant solution to tailor the drive unit performance by adding additional, resistors, capacitors and inductance. Obviously, you can't just take some randomly chosen drive units, plonk them in a randomly chosen cabinet volume and expect them to integrate with a minimal X-over (as at least one 'designer' seems to think).

My experience is that, irrespective of what the computer modelling may say, speakers with complex X-overs (all things being equal) sound less dynamic than those with simpler X-overs.