Bose MusicMonitors


This is going to be partly a review, and partly some technical stuff. I like these little speakers, and decided to crack them open to see what makes them work so well.


A few years ago, I visited a Bose shop at a local shopping centre, mostly just to pass a bit of time and see what they were up to. After listening to a few items on display and coming away largely unimpressed (the home-theatre systems are particularly poor - rumbly and tinny at the same time), I spotted a pair of very small speakers sat by themselves. They weren't connected to anything in particular, so I asked permission and plugged in my phone to play a few tunes.

... Wow.

I don't impress easily, but these little speakers managed it. Not only do they put out a reasonable impression of the entire frequency range, they also get astonishingly loud for the size of them. Naturally, I had to have them.

Since then, I've used them in lots of different situations, and have found them to be very useful. Whether it's as a pair of small monitors for on-location recording, playing music in the kitchen while cooking, or (as Bose intended) at a desk, they've consistently sounded much better than they have any right to. Remember, there's only a 2" driver per side.

I won't go into a flowery audiophile-esque description here, but suffice to say, these little speakers are very impressive, punching way above their (physical) weight.

Tech Stuff

I ran some acoustic measurements and found these speakers are essentially flat from a few hundred Hz up to around 16kHz, and run flat to 70Hz. The 70Hz-few-hundred-Hz range is very dependent on nearby acoustics - being small, they're essentially omni down there. Measured free-field, there was a consistent dip in the 120-200Hz range, even close-mic'd. With some back-wall reinforcement (sat on a desk, say), that dip often filled in.

When confronted with something this good, I like to open them up and find out what sort of voodoo is going on inside. So, here we go:

The sides are cast aluminium, and the top/back/bottom/baffle appears to be a one-piece extrude. Neat.

Below, there's a close-up of the driver.

The driver itself doesn't appear to be anything special. It's a 2" nominal unit with a neodymium magnet. We can see some venting in the former to help move heat away from the top-plate and voice coil, and into the cabinet. The cabinet itself, being aluminium, shouldn't have much trouble conducting heat to the outside world. It's possible that the drivers are high-tech units with shorting rings etc, but from the outside, they look very conventional, with black untreated paper cones and rubber half-roll surrounds.

The passive radiators are quite neatly arranged. There are two of them, mounted to a rectangular section, which fires the LF output out the sides of the cabinet. The PRs are mounted with both of them facing the same direction. However, they're always moving in mechanically opposite directions - towards/away from the slot, due to the air pressure inside the cabinet.

The net effect here is two-fold:

1 - Some mechanical non-linearities in the passive radiators are cancelled out.

2 - Vibration reduction.

Here ends the summary of the acoustic design. In short, it doesn't appear to be anything particularly exciting, but it's obvious that sensible design choices have been made.

On to the electronics.

All of the electronics are housed in the right speaker, with the left speaker (shown above) being fairly empty. There appears to be a couple of bits of signal processing that occur:

- Some EQ

- Limiting, or at least soft-clipping

- Mixing low-frequency signals to mono - there's a cross-feed with an approx. 2nd order lowpass around 120Hz.

Graph below. Here, I connected the amplifier outputs to a DI box, and ran sweeps while increasing the drive level.

We can see there's a general boost towards the extremes of the frequency range. The steep cut around 16kHz is curious, though.

The 120-200Hz dip mentioned earlier appears to be designed-in, which is interesting. It's likely the result of desktop/wall positioning, but I do wonder if there's some psychoacoustic effect, too - by cutting some upper bass, that might draw more attention to the 60-100Hz range.

At high levels, such as the red curve where something is clearly running out of gas, distortion remains relatively low, around a few percent at worst. That suggests a good limiter is operating here: instead of allowing outright clipping, the level is being reduced to maintain a clean waveform.

In use, this is audible - when you crank it up (as much as you can, given the diminutive size of these speakers), the sound becomes compressed, but not particularly distorted. Any audible distortion seems to come from the driver itself.

The cross-feed mentioned above is a neat trick. What that does is make it so that, no matter what the mixing engineer has done, any low-frequency signals always come out of both speakers. That way, they're always making use of the relatively little cone area they have.

Electronically, then, there are a few tricks afoot which mean these speakers put up a performance that belies their size. Combined with the generally sensible acoustic design, the result is a decent-sounding little speaker.

I plan on investigating the 16kHz cutoff - it might be the case that the driver itself is capable of 20kHz, so why would Bose electronically roll off the upper treble?

Answers on a postcard.