8" Drivers for a small U-Frame

[Tenson]

Hi,

I have been thinking of building some standmount speakers that use a U-Frame for the low frequency component swapping to a separate top mounted section at 800Hz.

Not having any experience of U-Frame design I need some help picking suitable drivers, please!

I was thinking a design about 30-35cm deep (resonance around 280Hz undamped), 22cm wide and 45cm tall with two 8" woofers stacked on top of each other (will be TWW with the T on top).

Hopefully having two woofers like this will give enough excursion and power handling to get down to around 40Hz with EQ, do you think? Max SPL at 40Hz around 95dB?

So, what woofers are suitable? Budget is no more than £200 for all 4 woofers. Which rules out the Seas W22's I originally thought of.

Maybe the Peerless HDS line? Perhaps even the SLS 8". Is that any good up to 800Hz, I can see the FR but what about quality wise?

Any tips on U-Frame design also welcomed. I have read the stuff from John K. I really plan to just do it empirically, adding stuffing bit by bit to get the best rear cancellation.

[moray james]

is probably the best suited to your purpose and they are reasonablein price. Spider noise has to be one of your top considerations. If you can accept the slight increase in size the SLS 10 is only a little bit more expensive and will push more air. I am sure that you will be pleased with the 8 inch units. Best regards Moray James.

[Tenson]

Ooh, maybe the Seas L22? IIRC Zaph liked these for bass performance right? and they seem to have good excursion as well as a smooth response for a nice 800Hz Xover.

I guess I need a driver with as high Q as possible and low FS. But if I am going to EQ I really just need to pay attention to the linear excursion as long as I don't need stupid amounts of EQ as I want to do this passive.

Anybody got a good way to model these? Would modelling it as an open baffle the width of the U-Frame (sides and front) and then adding 6dB more SPL since it will be more directive at LF, be a good way?

[ShinOBIWAN]

Have used the L22-RNX4P myself.

Not bad at all but the better choice is the L22-RNX/P for running to 800hz due to the much lower inductance.

I'm not sure if they'd do 95dB at 40hz though, I used 4 of them sealed and found them a little lacking in SPL, although I was trying to use them as down to 20hz with an LT

[holdent]

Anybody got a good way to model these?

I asked this question in an e-mail to Martin King and he suggested that you can treat a U-frame design as a very short open-ended TL. If you've purchased his MathCad worksheets you would use the TL Open End one for this.

Alternatively, you could use the free U & H Frame worksheet provided by at the FRD Consortium site at http://www.pvconsultants.com/audio/frdgroup.htm. This worksheet was also written by Martin King but he has indicated that he's not entirely happy with it. It also requires MathCad.

Linkwitz discusses U-frame woofers here http://www.linkwitzlab.com/models.htm#C and they are discussed on the NaO site at http://www.musicanddesign.com/u_frame.html (look at the very bottom for a practical example).

[maxro]

Originally posted by holdent
...Alternatively, you could use the free U & H Frame worksheet provided by at the FRD Consortium site at http://www.pvconsultants.com/audio/frdgroup.htm. This worksheet was also written by Martin King but he has indicated that he's not entirely happy with it....

Neither am I. It made my Mathcad freeze up. None of MJK's other sheets seem to do this.

I've been playing around with Svante's Basta, trying to fool it into calculating baffle step for a U frame. Basta won't calculate any 1/4 wave action, only baffle step and open baffle wave cancellation.

What I've done is have Basta calculate two FR graphs and manually take an average. One as a "large closed box" with the frontal dimensions of a proposed U baffle. The second as an "open baffle" with a width equal to the front plus twice the depth of each side. For example a 1' wide x 1' deep U frame would sim as a 1' wide box and a 5' wide open bafle. I have no idea if this actually works, but it seems to make sense in my head.

Max

ps. Tenson, I would go for the Peerless SLS.

[Tenson]

Hi guys,

Why the choice of the SLS? It obviously does have more Xmax but will it still sound any good up to 800Hz? I was leaning more to the L22

[holdent]

The Seas L22RN4X/P seems to be a better woofer than the Peerless SLS213 but is also more expensive. I quickly modeled both in a 25 x 25 x 40 cm U-frame (using Martin's TL_Open_End worksheet) and got poor results -- the response for both drivers drops off from about 100 Hz about 12-14 dB per octave. At 40 Hz the Seas is only putting about 75 dB vs about 92 dB at 100 Hz.

For virtually the same price as the 8" Seas you can get the 12" Peerless SLS woofer. You'll get down to 70 Hz w/o EQ with a 38 x 38 x 50 cm U-Frame and it is overall much more sensitive. The spec sheet for this driver at http://www.madisound.com/pdf/peerless/830669.pdf suggests that at 800 Hz this driver is only down roughly 2dB or less 60 degrees off axis. I think this driver will fit your application better than the other two 8" drivers.

[Tenson]

Originally posted by holdent
For virtually the same price as the 8" Seas you can get the 12" Peerless SLS woofer. ... I think this driver will fit your application better than the other two 8" drivers.

Unfortunately, due to being 12" it wont. I don't want anything bigger than 8", though having two vertical is not a problem.

In the past I have found Martins worksheets model the output rather poorly for TLs - it is always down much further than it is in real life, especially with room gain. What they are good for is getting the acoustic impedance right.

I feel that with room gain, the extra sensitivity of the u-frame over a dipole and a bit of EQ, they could manage 40-50Hz. There are two woofers per box after all.

I might give it a go first of all with a pair of those cheap Dayton 7" that I have.

[tinitus]

Some tough demands

Maybe the MONACOR car-fi 8" autowoofer SP 202C / 102800

37hz / 91db / Qts 0.5 / 0.3mH

problem is 4 ohm but they cost around 20 EUR so maybe four pr side ?