Dipole EQ Problem

[m@]

Hi everyone,

I have a problem I'd appreciate some help with - and I promise not to throw out any more half-baked theories.

I'm running an open baffle bass setup with two 15-inch 97 dB pro woofers with fs just under 40. The baffle size gives an fequal of 90hz.

I'm using a sound card to figure out what EQ is needed before I build and analog EQ. I'm running the sound card out to a single LM3886 with about 150 VA / 2*4700 uF on the chip. I'm running the drivers in parallel at 4 ohms total.

The total EQ I have set is over 30 dB at 25 hz, which is right about where it should be, or even high, given everything I've been able to read about dipoles. The sound card is not clipping - nor is the amp - I have a scope on the speaker terminal.

My question is, when I run an RTA set at about 1 meter with pink noise I still show the low end rolling off fast - it's about 6 dB down at 40.

I know the fs on the woofers isn't that low, and the xmax isn't exceptional, but I haven't been able to exceed xmax on them, so I'm not sure that's an issue. These woofers have a good build quality and should be able to take just about any type of abuse. (That's an understatement actually, they're built like tanks.)

Two questions:

Am I missing something about using an RTA in this setup? I would assume 1 meter away would show the dipole roll off properly.

Also, I have the overhead in the drivers to get as much EQ as I need, it would seem, yet the EQ I'm throwing at it seems to have little effect. I'm not running the amp into clipping, but is it simply running out of juice? I'm sure the amp is undersized for the application, but would an undersized amp exhibit a roll off towards the low end like that - given the amount of EQ I'm trying to push through it?

Any thoughts would be helpful =)

[nunayafb]

we'll need more info about the sub and eq, is it parametric, 6dB/oct upward slope to counter dipole roll-off, etc...

My question for you is did you factor in the speakers natural rolloff or just dipole rolloff?

[opp]

This is a dilemma - normally only nearfield measurements can be trusted in the low frequency region due to room modes and nearfield doesn’t work with dipols. I can see two alternatives:

1) outdoor measurements
2) acoustically adding nearfield measurements from front with nearfield measurements from back of the baffle. The back added the back-to-front delay.

With the aid of Speaker Workshop, I have used the latter method I my dipole project with reasonable success.

[MBK]

As pointed out above, one would need info about the woofers - not just Fs, but Q as well! With Q=0.5 the woofers alone (w/o dipole rolloff) would already be down 6 dB at Fs. With a typical low Q pro woofer, say Q=0.3 or so, you might be down 6 dB well beforew Fs.

Additionally, room effects can also work to the negative. I have a severe dip at 40 Hz in my room for instance. Definitely, outdoor measurements would help here. That's my problem too - because outdoors it is very easy to run into S/N issues at larger measurement distances. In my area even the outdoors has a strong ca. 50 Hz hum from background traffic noise.

[Rudolf]

@ m@,

a possible strategy: Run a RTA in the extreme nearfield - distance max. 1/2 cone diameter. That way room influence can be almost omitted. Through your planned passband the response should rise from high frequencies to low frequencies with at least 6 dB/oct. I would not recommend to use the driver below fs, but if you insist so, you will need some added EQ from 40 Hz down. When you move into the room this gain will be "devoured" by the dipole roll-off.

If you can´t achieve that 6 dB rise in the nearfield, the driver, the amp or probably the soundcard will be at fault. Are you sure your soundcard is linear down to 25 Hz??

Rudolf

[Nuuk]

I'm using a sound card to figure out what EQ is needed before I build and analog EQ. I'm running the sound card out to a single LM3886 with about 150 VA / 2*4700 uF on the chip. I'm running the drivers in parallel at 4 ohms total.

The total EQ I have set is over 30 dB at 25 hz,

You are expecting that amp to properly control that driver down to 25 hz? Good luck!

[MBK]

You are expecting that amp to properly control that driver down to 25 hz? Good luck!

Actually, if you calculate power vs. excursion demands using one of Linkwitz's worksheets, you can see that amp power as a rule of thumb won't be the limiting factor below 80 Hz or so - it will be excursion - and depending on the driver it only may take 20-50W to drive it to full Xmax at lower frequencies.

In terms of control, around resonance no amp will have real control over the driver: impedancwe will be at maximum, current drawn at minimum, power consumption will be at its minimum, and you'll depend on the mechanical parts for linearity.

In any case it appears that clipping was not the problem, no?

[Nuuk]

Well, if that's the case (and I have no reason to doubt you or SL ) perhaps I can get away with a bridged 3886 with a 300VA traffo to drive a pair of Hawthorne Audio Augies!

[MBK]

Well, I don't know these speakers... but, once you throw in a crossover things get a bit more complicated, usually because of an impedance minimum though, and not at Fs but higher.

FWIW the spreadsheet is here . It works for closed boxes and infinite baffle / dipoles (set box volume >>Vas). It gives you the limits in terms of Xmax, Imax, Vmax, and Pmax. Drwback is, you need the complete driver parameters, which some manufacturers don't give in enough detail.

From my playing around with it, below 80 Hz most drivers just don't have the necessary Xmax for decent SPL, and extra wattagecan't fix that - you have to add Sd or Xmax. The wattage limitation to maximum SPL usually comes in between 80 and 200 Hz. Above 200 Hz it's usually easy to get SPL in both terms of Xmax and Power, for the typical drivers used there (say, 6.5 to 8").

[Nuuk]

Well, I don't know these speakers... but, once you throw in a crossover things get a bit more complicated, usually because of an impedance minimum though, and not at Fs but higher.

I'll be using an active crossover!

FYI, Hemp Acoustics are working on a 15 inch woofer with 14 mm xmax that is intended for OB use.