[hwc]

I assembled my new Marchand Bassis kit this afternoon -- scavenging an old analog Fosgate Dolby Surround decoder for the case and regulated +/- 15V power supply. I had ordered the completed circuit board from Marchand and everything in the kit was very nicely made. Pots and switches were meticulously soldered and shrink tubed on twisted jumper cables with molex connectors, so it was just a matter of figuring out a bracket arrangement to mount the board, install four pots, two switches, and a couple of RCA jacks. Got it all wired up and plugged into the system tonight.


This is a very cool piece of gear. Basically a fully adjustable 4-pole Linkwitz Transform Function bass EQ for sealed box subwoofers. You use two of the pots to enter the "Q" and "Fs" resonance (more or less the -3dB point) of the actual driver box combination. Then, you use the other two pots to dial in your target "Q" and target "Fs" (expressed as boost). The EQ changes the subwoofer output from your actual curve to the target curve -- whatever you want. The switches are for "bypass" and an 18 dB per octave subsonic filter at 20 Hz.


What makes it doubly handy is a spreadsheet I found that lets you enter all four parameters and graph the before and after response curves.


My subwoofer by itself had a calculated "Q" of .76 and an Fs of 43 Hz. I replaced that with a Q of .5 and an Fs of about 20 Hz. The response curve is now about -6 db at 20 Hz and about -3 dB at 25 Hz. This requires fairly substantial boost at 20 Hz, but it seems to be within the excursion limits of the rather robust driver, even on the tank scene of Private Ryan. On paper, the boost should be hammering the driver, but the saving grace is that DVDs don't record digital full scale signals at frequencies that low.


I tried a target Q of .7, but at first blush, the lower Q seemed tigher. The theory says that a Q of .5 or lower has zero ringing and nearly perfect transient response.


I haven't had time to do any precise measurements, but I did plop the Radio Shack meter down and looked at a couple of before and after low-frequency sweeps. The EQ works as advertised.


The combination of a sealed box sub and the Linkwitz Transform Function seems like a great way to design a subwoofer. Several companies seem to be taking this route: starting with a smallish high-Q driver/box combo and equalizing for extended low-Q response. But, being able to have full control over it to make you own choice about the bass extension/power handling tradeoff is a good thing.

 

[EMT]

Thanks for your comments on the Marchand unit. May I ask what sub you are using it with?


Also, have you used any of their electronic crossovers?

 

[hwc]

I'm using it with your typical $150 10-inch subwoofer driver: a JL Audio 2.25-inch voice coil, Kapton former, 15mm one-way linear excursion, etc. It's a dual 4 ohm voice coil driver, wired in series driven by a bridged mono 160 watt Audio Research amplifier.


It's in a 1.4 cubic foot sealed enclosure. I've got the filter set to begin boosting bass.


Here's a EQ curve for the particular settings I was playing around with last night.:


Hz Boost (db)

1 18.6

2 18.5

3 18.3

4 18.0

5 17.8

6 17.5

7 17.1

8 16.5

9 15.7

11 14.7

13 13.5

16 12.1

19 10.4

22 8.5

27 6.5

32 4.5

38 2.7

50 0.7

60 0.1

71 -0.2

84 -0.3

100 -0.2


The boost below 20 Hz isn't much of an issue with digital sources, since there is no rumble to worry about. Any information down that low on DVDs is recorded at fairly low levels, down 30 or 40 dB from digital full scale. As you can see, it works like a standard bass control, except that it is adjustable to be limited to extremely low frequencies, whereas a normal bass control starts boosting at 200 Hz and might be 8 dB up at 50 Hz.


With this type of EQ, it really doesn't matter what driver/sealed box combo you have. As long as you know the Thiel-Small parameters and the box size, you can set the EQ to replace the actual performance characteristics with new target characteristics. The limiting factor is excursion, because you are boosting the signal below say 40 Hz, so you definitely want a driver capable of long linear excursion.


Even though the box is relatively small, I could replace the woofer with a 12 inch driver. Even though the box would be undersized (high-Q tuning), it doesn't matter, because the EQ can easily turn it into a low-Q system.


The only system this EQ wouldn't work with would be a vented enclosure. You could use half of it to increase the damping and turn a high-Q vented system into a low Q system by removing the peak. But, you can't boost bass below the tuning frequency on a vented system or you kill the driver. So for vented systems, a peaking "inverted notch" filter with a lo-cut filter would be more appropriate instead of this kind shelving filter.


I've not used their crossovers, but I would expect them to be similarly high quality. They are all available with Burr-Brown op amps as an upgrade -- I didn't figure the ultra lo-noise op amps would be relevant to this particular application.

 

[BruceD]

I use a Marchand active crossover (XM9-L) with the upgraded op-amps and they are truly high-quality devices.

 

[hwc]

Hey guys:


I did the mod on the Radio Shack meter tonight: take it apart and replace about nine capacitors to get rid of the C-weighting curve. The meter now measures flat down somewhere below 10 Hz so you can run frequency sweeps and get a true reading directly from the meter.


I didn't have much time, but I ran a couple of sweeps with the Marchand EQ. Theoretically I had the EQ set to provide Q = .5 with a -6dB point of 20 Hz. My little 10 inch sub actually did better than that, being - 3 or -4 dB dB at 20 Hz. Probably getting some room gain down there -- according to the longest dimension of my room (the point below which room gain kicks in), my lowest standing wave should be right at 20 Hz


Amazingly enough, I have very little problem with room modes. I have minor rises at multiples of 20 Hz -- 40 hz, 60 hz, and 80 Hz -- but they are relatively broad peaks with a max of about 2 dB when close-miking the sub. I am certainly not going to worry about +/- 2 dB variations in the output of the sub!


I can probably back off the EQ a bit, as -3 dB at 20 Hz is more than I expected to get. Even at those settings, the woofer plays the pod race in Star Wars Episode 1 at my normal listening level, and that soundtrack has some serious low bass energy. The woofer is probably right at the edge as a few of the low bass sounds look like they are using up most of the 1 inch peak to peak excursion, but no audible signs of distress -- no clanging, no buzzing, no death rattles -- which is good.


The limiting factor with this type of EQ is cone excursion at low frequencies. It has a 20 hz subsonic filter, but with the shallow broad rolloff down to 10 Hz of a low-Q tuning, I'd prefer not to use it. Looks like I'll be OK even with fairly massive amounts of EQ. I think the thing that would bite me would be a DVD recorded with 10 Hz energy at -10 dBFs. Hopefully, no recording engineer is stupid enough to do that.


Star Wars is bad enough with LFE levels only -20 to -25 dB below max from 5 to 10 Hz, -10 to -15 dB below max at 20 Hz, and practically nearly full scale signal levels from 30 to 50 Hz. That is a brutal test of this EQ because I'm looking at 16 dB of boost at 10 Hz. I think this is why they mastered the DVD at lower levels than the average DVD, requiring higher playback levels to get the same perceived loudness. I think they knew that had speaker-killler bass on the LFE channel.


Oh...the sound of the subwoofer with the EQ? Amazing. I don't think you need 20 Hz bass to have a good sounding subwoofer, but it certainly does add a lot "weight" to the presentation!