This is a set of RTA measurements I did on my Mazda Miata 2000 LS. It came with a factory Bose system which I’m very disappointed with. It BOOOMS.

17Sep2000

Given that the bass has a large boom at 50Hz I checked is if the boom is electronic or acoustic. I played a pink noise CD track and hooked my RTA to the speaker terminals. This time I set the RTA to 1/6^th octave. Here is the result:

The blue line is with the bass driver disconnected and the pink, connected. You can see the system is pretty heavily equalized. The 50Hz bass boom is not there, which means it’s acoustic in nature. The doors are tuned to boom at 50 Hz, like a boomy vented box. It is surprising to see that they boosted some frequencies by as much as 10 dB in some places. The 2 kHz boost is probably to compensate for the rolloff of the bass-mid units, as the paper tweeter unit has a single electrolytic capacitor for a crossover. The 200-400 Hz dip may be to compensate for some cabin boost or something.

The 20 kHz peaking when unloaded is probably the switching amplifier’s output filter peaking when unloaded.

The fact that the pink line is around 2dB down in places means that the wire resistance to the drivers is significant. It’s losing about 2dB of power, which is something like a 30% loss of power in the wires. The Bose system uses 0.5 ohm drivers so that the power amp, with no switching power supply, delivering 12V peak to the drivers, can theoretically deliver >140W into a 0.5 ohm load. The skinny speaker wires means 30% less power is getting to the speakers.

The above measurement was taken with bass and treble setting set to flat, and the volume set so that the LCD indicator just shows 2 notches down from center.

I repeated measurements with 4 combinations of top up and down, and windows up and down. All of them showed a peak at 55 Hz, and a broad one at 1kHz. This 1 kHz peak I believe is mic position dependent. For the new, unshown measurements the mic was sitting very close to the headrest. The 1 khz peak shows in some of the measurements below.

I finally found the source of the 50 Hz boom in the RTAs below (sort of)

I opened up the door panel. I took the door skin off, and stuffed the rear of the Bose basket with foam. This should've killed the resonance, but it didn't. I removed the Bose driver again, and I took my RTA mic and put it in the hole where the driver was. I started pounding my hand around the door, and lo and behold, it was showing a very strong 50Hz resonance. The whole door is acting like a drum resonating at 50 Hz! Sure enough if I listen to the "drum note" the door emits, it's the same note as the annoying "one-note bass" I hear when I play music.

Interestingly, it wasn't as strong if I open the door an inch. Maybe the door resonates more if it's supported at the end when it's closed.

It wasn't the port-like thing that Bose put, the whole door resonates! I put the driver back and played a sine sweep from 60 Hz down, and if I put my hand on the outside of the door, about a foot below the side view mirror, it vibrates very strongly at 50 Hz. That's the source of the Bose boom. I don't know if Bose had anything to do with it. Nevertheless they should have fixed it, like EQ'd it down. I think they thought the boom was desirable.

I don't know why the 2000's are boomier than the '99s. Different door structure, or Bose reinforced it with their door panel port-like thingy?

I think one way to reduce it is to put epoxy or some other glue in places along the side impact beam in the door to make the door sheet metal stiffer.

To solve the problem I'm just gonna put 8" subs on the rear deck. A 3way system will surely sound better than a 2way.

Here are the first measurements I made:

I started with the top down and the windows down, and balance knob to the left, and volume halfway up

Note these RTA readings have an offset, I don’t know what it is. It sounded like it was playing at something close to 90 dB SPL.

Here’s the first graph. The 50Hz boom is very evident. As you can see, turning down the bass still leaves a great big boom. I also measured with the treble turned halfway to max, because that’s where it seems best. If you combine the left part of the curve with the bass down, and right part with the treble turned up halfway, that’s how it’s flattest (if the boom is removed). That is how I listen to it, and my ears were right.

Next I show left vs right channel playing, with bass set to flat and treble halfway up.

Comparing top down vs. up (top up had windows open), a huge lump appears at 1 kHz.

With the top up and windows up, I compared left alone, right alone, and both channels playing:

While using the stereo I noticed that turning down the volume, the bass gets reinforced. I found it a bit too aggressive though.

Compare the blue curve and the pink curve (yellow curve is close enough to pink) As you can see, the midrange went down 10 dB but the bass boom went down only 4 dB. It sounds extra boomy at lower volumes.

Last I wanted to see the effect of the windows when the top was up. I usually drive with the windows up whether the top is up or down. What I forgot to measure is how it sounds when the top is down and the windows are up (my usual driving). It actually seems to sound best like this – plus the wind noise obscures the imperfections of the system.

I couldn't resist opening up my 2000 LS doors to try and fix the

Bose bass boom. Here's what I found. Pictures later.

It appears to be a ported box basically, but the port is more like a

slot in 3D. The driver has a very firm suspension so it doesn't bottom

out.

Driver is an "upside down" driver, where the magnet sticks out on the

front. This makes for a very shallow driver, not that the Miata needs

it at all. Bose may use this driver for many autosound applications

regardless of the shallowness is needed or not.

What you see in the rear is the spider. Is does quite a bit of

excursion, I think around 12mm p-p. Cone area is close to that of an

8". The front of the driver fires forward, and the rear fires into the

door space. The "vent" or "port", if we call it that, fires into the

space between the door panel and the door sheet metal.

The driver has a long plastic basket about 3.5" deep (there's a LOT of

room for a deep driver) This basket is presumably for protection from

the rain.

The tweeter is a paper cone unit, and a single small electrolytic cap is

what passes for a crossover.

Things that are wrong:

The way the front of the bass/mid driver is built, you can't have good

midrange. The way it fires forward into the door speaker grill, 1.5

inches away, is not good for midrange. The way the rear fires into the

"cup", is not good for midrange. The bare metal of the door outer skin

is not good for midrange.

Metal doors are not good for uncoloured bass reproduction. As an

enclosure, it vibrates and rattles.

The whole system reeks of low budget. Granted they had to do some

creative engineering to make it boom on such a low budget.

To reduce the boom, I may try 1 of the following:

Lower the port tuning frequency - but this requires building up this

styrofoam block thingy that forms part of the port in 3D.

Add acoustic resistance behind the driver - stretch a layer of

fiberglass across the rear of the cup or basket.

I will check to see if the boom is also created electronically - that

is, a bump at 50 Hz is EQ'd in.

What I’d like to do is replace all the drivers. Maybe try 8" in the doors, like this one:

and see if I can get decent bass and midrange with a 2way system. The above Audax seems to have

decent xmax, dispersion, and response to 2 kHz. One potential problem is that it has a failry low Qts of ~0.4.

Midrange sounds a lot better if it’s surface mounted instead of deep inside the door like the Bose

speakers. Problem is that the surface of the door grill isn’t parallel with the sheet metal mounting inside. Hmmm.

Any suggestions?