Finding the middleground with a bandpass design

[AllenB]

The last bandpass (4th) enclosure I built sounded bad, embarrassing even. The transient response just didn't work. I think I may have strayed beyond the drivers comfort zone.

I know about the gain/bandwidth tradeoff but is there such thing as the "appropriate" response for a given woofer? Is there a "best" closed volume that I can relate to standard closed box design?

Put another way, for a given bandwidth and box size constraint, is there an "ideal" Thiele/Small set to look for?

[GM]

Originally Posted by AllenB

Put another way, for a given bandwidth and box size constraint, is there an "ideal" Thiele/Small set to look for?

Yes, Prof. Leach's compression horn math is very accurate.

GM

[AllenB]

Thank you, GM. I've run a sample driver through the first page of mlutil to see how this works.

I assume you're suggesting that the resulting front/rear cavity volumes will produce the optimum loading for that driver? and any other response from that driver would be "less than optimum"

I assume I can ignore 'M'?

Why does it give me a 'new upper f3' when I already told it where I wanted it to be?

and, will I need to guess and work out the front chamber tuning frequency?

[GM]

You're welcome!

When combined with the vent's loading, correct.

I can't run mlutil to play with it, but IIRC it only designs a horn based on driver specs hence the new Fh3. In his papers though, ML also rearranged mlutil's math to calc the driver's specs for a given horn alignment.

'M' defines the horn's flare factor, so can't be ignored, especially when accurately flaring the vent, though some say a simple radius round-over suffices.

At this point though, I'm thinking that without a better understanding of horn theory it's probably best to just use a simple BP designer and try different driver specs.

Note that for a 4th order max flat you want both the front and rear chamber to be tuned to the same frequency. From this we see that regardless of the alignment, the BP's mean tuning frequency will track the 'sliding scale' roll off response of its 0.707 Qts sealed one and all the design variables to get the desired eff./gain BW that this implies.

GM

[Patrick Bateman]

Originally Posted by AllenB

The last bandpass (4th) enclosure I built sounded bad, embarrassing even. The transient response just didn't work. I think I may have strayed beyond the drivers comfort zone.

I know about the gain/bandwidth tradeoff but is there such thing as the "appropriate" response for a given woofer? Is there a "best" closed volume that I can relate to standard closed box design?

Put another way, for a given bandwidth and box size constraint, is there an "ideal" Thiele/Small set to look for?

Did you measure the resonant peak of the woofer in the sealed chamber, and then tune the front chamber to that frequency?

The reason that I ask is that i've built a *lot* of bandpass boxes. And they all sounded sub-par until I learned how to properly measure the impedance curve, and tune accordingly.

The bad reputation of bandpass boxes is mostly due to mistuning IMHO..

And whatever you do, don't EVER rely on the predictions from programs like WinISD. In the real world, loudspeaker ports are WAY too complex. You *must* measure the impedance to get good results.

[AllenB]

Originally Posted by Patrick Bateman

[font=verdana]Did you measure the resonant peak of the woofer in the sealed chamber, and then tune the front chamber to that frequency?

This is what GM is saying. I take it that this is the key to the optimum BP, using volumes from leach.

No, my last design was some years ago and was just a way to force a response from a driver. I realise I've been looking at BP through jaded eyes since.

So should I start with the two volumes from leach, mount the driver in the rear chamber, measure fc, assemble the front enclosure and choose a port to match fb to fc?

And whatever you do, don't EVER rely on the predictions from programs like WinISD.

Off topic, I have some concerns about WinISD so I don't use it. I tried AJ, hornresp, spreadsheet and others.

[GM]

WinISD Pro is pretty good and WRT vent length for a given alignment it matches HornResp. Unibox is said to be good also. The most accurate I've used though is MJK's MathCad software. It's a pity he dropped his FLH WS as it does a great job of matching ML's predictions, ditto 4th order BPs.

Play with sims first, then do as PB recommends. BTW, WRT minimum vent area, calculate it the same as you would for a BR of the same net Vb, Fb for 'x' max power at < 5% mach (~17.2 m/sec).

GM

[AllenB]

GM, finding drivers here is not terribly easy. Would you perhaps know of a few drivers that are 8" units with Qts less than 0.25, Fs less than 25-30Hz and Vas less than about 1'^3(28L)?

[GM]

The only one that comes to mind is the popular MCM 55-2421 that Bill F. uses in one of his FLHs, so do a search to get a variety of measured specs, performance info.

GM

[Patrick Bateman]

Originally Posted by GM

The only one that comes to mind is the popular MCM 55-2421 that Bill F. uses in one of his FLHs, so do a search to get a variety of measured specs, performance info.

GM

I really like the 2421; it's inexpensive and it works well in both conventional and tapped horns. For a bit more money, the Tangband W8 is reputed to have better build quality, and more xmax. Having said that, I've never destroyed a 2421, even after feeding it with something like three to four times the rated power.

If neither of those options is available, you might check out Image Dynamics. It's going to need a bigger box than the MCM, and it costs more than twice as much. But it also handles more power, and has more than twice the xmax.



ID8V.3-D4 - Image Dynamics ID 8" Subwoofer

Impedance: 2/8 Ohms
Resonant Frequency 24.7 Hz
DC Resistance 1.95/7.8 Ohms
Electrical "Q" .367
Mechanical "Q" 3.572
Total "Q" .332
Equivalent Volume 1.55 cuft(44 liters)
One way linear 15 mm
Cone Area 231 cm2
Sensitivity SPLo: 85.6 dB
Power Handling 150W RMS/300W MAX
Power: 50-350W RMS
XMax: 15.5mm
Sealed ft3: 0.40
Ported ft3: 0.75
Mounting Depth: 4.75"