Wednesday, October 17, 2018

New Projects in the works

It's really quite something when it's time to dust off your old blog, which you haven't updated in almost a decade! Now, what should I do with those old posts?

Intense philosophical brain-racking ensues...


Delete

That seems like a waste and it would be a little disrespectful of my younger self, who did a lot of work on speaker related issues. How could I go on, knowing that a future me might savagely delete what I'm doing now?

Do nothing


On the other hand, the posts probably contain old errors or flawed ideas and whatnot — a fearful part of me cares far too much about appearances. What will people think??

→Do something in between

I hereby promise to do some moderate dusting and clean-up, when I have a moment.

New Project

I've been asked to build a spectacular new sound system. I have full creative control over the design, within 'reason'. In other words, there are budget and time constraints, and the customer will know more about what they want when they see it!

As with many projects, the design process is iterative and some of the specifications will probably change as time progresses.

Purpose:
General-purpose home stereo speakers for music and movies.

Vision:
A 3-way system with dedicated amplifiers (6 of them).

Size: 10" woofers will deliver solid bass. This will be a big factor that determines the overall size of the system, and what mid-ranges and tweeters are selected.

Mid-range and tweeters, size and model: TBD, probably 5-6" and 1". A line array of wide-range speakers (perhaps 4) is also on the cards.

Boxes: The mid-range and tweeter will be housed in a unique "leaky box" design, somewhere between a sealed box and open baffle. However, I'm also considering an open baffle design, at least for the mid-range.

Amplifiers will be matched to their speakers, and possibly built-in.


Discussion


Past experience dictates that "2.1" (or 5.1 or 7.1) sub-woofer style systems are more trouble than they're worth in terms of getting the bass right. When each channel has its own dedicated woofer, it eliminates a lot of issues.

What issues? E.g.: far from being omnidirectional, mono bass is often easy to locate. Even with 24dB/oct filter slopes, out-of-range sounds in the kilo-hertz range are often distracting. There is no clear or universal standard for merging stereo signals into a mono signal. When comparing, say, 2-ch vs. 2.1, where the point-1 is summed from 2 channels, there may be significant differences in tone and some loss of information.

Test signal with polarities reversed:
A 2-channel set-up may produce some interesting effects.
A 2.1 set-up with simple stereo-to-mono conversion for the bass could sound relatively bland and thin because the bass cancels itself out.

Mono pink or white noise:
At low frequencies, adding a second speaker tends to double the amplitude because the phases are coherent. However, at higher frequencies, some destructive interference can be expected, especially if the speakers are placed far apart. There will also be a transition region at moderate frequencies with some summing and some cancellation. A design problem occurs when attempting to mimic this behaviour with a .1 channel.

Sealed bass

I've done a lot of experimentation with both sealed and ported designs in the past, and I generally prefer sealed designs. Ported designs certainly have some appeal — they deliver more boom per volt, and they possess a "sweet spot" with low distortion because the cone hardly moves at all and the port does most of the work. Nonetheless, loose, boomy bass can be a problem with ported speakers, even if they are "well tuned". I consider removing unwanted ringing to be one of the hardest aspects of speaker design, and fixing those issues at low frequencies is even harder.

Lossy mids

I'm torn between doing a full dipole design, and boxes with resistive vents. There was a time when I was ideologically opposed to the idea of allowing reverse polarity sound interfere with the sound coming from the front, and I'm still a little unsure. As the waves interfere with each other, some frequencies are cancelled out, while others are accentuated. However, this also occurs with monopole speakers, just at a different set of frequencies.

In addition, dipole speakers, and raw speaker drivers in particular, produce a kind of stereo effect similar to the way that LCD screens play tricks on the eyes when viewed from an angle. The exact interference between the front and back waves varies with the listening angle, and the effect is quite strange to my ears. Then again, attempting to block the entire back-wave with a box also results in strange artefacts.

Is it possible to have the best of both worlds?

Amplifiers 

Low output impedance seems to cause a lot of problems for speakers, reducing the sound quality and making them tiring to listen to. According to one theory, the movement of a speaker cone should precisely follow the voltage generated by the amplifier, and any disturbances should be strongly regulated by the amplifier. However, this can result in unwanted energy storage. Many speakers are already over-damped at bass frequencies, with a Qts around 0.4, and an old rule-of-thumb states that the box size should be chosen so that the overall system is critically damped with a Qtc around 0.7.

But what happens at higher frequencies? Cone break-up (partial oscillations and strong resonances) can be a serious issue that renders a speaker unusable in the affected frequency range. Even if steep filters are used, speakers often have high sensitivity in their break-up zone and the problem is only partially fixed. Therefore, one solution is to tune the output impedance, allowing the speaker's own motor to work as a sophisticated suspension system, rather than a hard reflector.

It will be interesting to hear what happens with the woofer boxes! In the past I've had to deal with unwanted resonances around 300~400 Hz, a region where a 10" woofer should still sound great, and the stiff cone should still move as a piston. However, the air mass inside the box tends to oscillate, and damping materials hardly make any difference.