OB Qts: how high is to high?

[Chaz H]

To serve as the evil twin to the "how low can you go?" thread, I want to get opinions on when woofer Q is TOO high for OB.

In particular my question pertains to the results from MJK's OB/H-frame/U-frame comparison. In the fostex+alpha15 paper MJK recommends a woofer with Qts between 1 and 1.2. But due to added air mass the U- and H-frames increase the total Q for the eminence from 1.26 up to around 1.47 and 1.65 respectively. Do I have it right that a woofer with its Q increased this way will behave the same as a woofer with such high Q to begin with? Does that mean I can take a woofer with a Qts of 1.5 and an Fs of 35 hz and expect the same bass response on an open baffle as the alpha-15 will have in an h-frame? And in that case, where should it stop? When does Qts become a dealbreaker?

 

[hitsware]

Any Q >~ 1 will cause some ripple, but is advantageous.
However >~ 2 the ripple starts to become 'one note bassish'.

 

[TerryO]

The Qts figure by itself is not really that important, unless coupled to a Qms to a high enough number and it (Qms) remains linear.
The Carver "Amazing" OB speaker started with specifically designed 12 inch woofers (built by Sony to Carver specs) with a Qts of about 3.12 or so and, IIRC, a Qms of about 10. This allowed the woofers to remain flat without EQ'ing down to about 20 Hz.

Later versions used a Tonegen woofer with a lighter cone and a Qts of about 2.85 and were targeted at 28-30 Hz, again without EQ.

This was the result of original research conducted at Carver, as there really wasn't any information available and the driver build houses had never attempted anything like this before.

Best Regards,
TerryO

 

[mbskeam]

hi,
where did you get this info?
specifically that Sony built the 1st gen woofer.
thanks

 

[Chaz H]

Thanks Terry for that very interesting response. I have heard of a driver with such a high Qts so that must indeed be an oddity! I'd like to know: what do you mean by the Qms remaining linear?

Here's another question: I've seen it posited as a general rule that to determine the rolloff of a driver on an infinite baffle, you divide the Fs by the Qts. Doesn't that mean that a driver with Fs 35 and Qts of ~1.5 will extend to nearly 20hz as well? I'm looking at these GRS drivers for a possible cheap open baffle project and it seems like extension is no problem if efficiency is not your priority...

GRS 12PF-8 12" Paper Cone Foam Surround Woofer

 

[ra7]

A simple forum search would've revealed this:
http://www.diyaudio.com/forums/full-range/186571-qts-openbaffle-how-low-can-you-go.html

 

[Andrew Eckhardt]

Most people aren't massively sensitive to transient response in the low bass region. Either by conditioning or general acceptability, the response of ducted ports is still popular. If you plot out the phase and group delay of a Q=1 OB speaker with Fs of 30 Hz next to a maximally flat single ported speaker with tuning frequency at 30Hz, the OB speaker is going to beat it in both areas. But if you get to the point where ported sound annoys you, you might eventually get to the point where a Q=1 woofer also starts to sound like bass notes are all getting drug toward the resonant frequency and anything that drives the speaker at resonance seems to be a little loud and lacking a clear beginning or end. As the resonant frequency increases its easily more offensive. I have some Beta12A2's (Qts raised by current feedback method) running at Q=.9 right now and that's about the limit for me. If their free-air Fs weres 30 rather than 47 it might not be such a big deal. This is just a test situation for the time being. Those drivers wont be running below 70Hz for much longer.

 

[TerryO]

hi,
where did you get this info?
specifically that Sony built the 1st gen woofer.
thanks

I got that from the designer Jim Croft, who lives here in Seattle.

Here's his statement:

I created the Carver Amazing woofer system and Large Area Ribbon (co-authored with David Graebener) for Bob Carver in the mid 1980's as a consultant to Carver Corporation and was VP of Research and Development at Carver in the 1990s. I would be glad to answer any questions anyone has about these devices.

One question that was asked was whether the Amazings had a high or low Qms. The Qms was quite high, we attempted to achieve a Qms of nearly 10, which is required if one is going to end up with a Qt of much greater than 2 and still maintain reasonable efficiency.
The trick is to be able to maintain 'linear' Qms. One of the big problems we found early in the development of the Hi-Q open baffle was that many of the Hi-Q drivers have Qms values that change depending on level. This non-linear Qms causes a non-linear Qts, which causes a non-linear frequency response with level...i.e. compression/distortion. This Qms modulation is one of the main reasons that some have found higher Qts drivers to sound bad. It’s is not the higher Qts, but the Qms -> Qts nonlinearity that causes the poor sound quality. Proper Qts match to baffle cut-off frequency will perform quite perfectly, even for higher Qt values if the resulting frequency response is flat and the Qt value is maintained at all listening levels.
The Amazing suspension components, spider and surround, took nearly 9 months to work through the design problems to chase out the non-linearities. No one had optimized high-Q woofers previous to the Amazings so the driver vendors were not able to help. We had to drive the component development ourselves. It was hard to get spiders with low enough damping (they actually call them dampers in Japan) to achieve high-Q and then it was even more difficult to develop spiders and surrounds that maintained constant damping.
Okay, that is probably more than what anyone wanted to know on that subject, but I felt it was important to start getting past this myth that “high-Q drivers are bad”. It is a systems approach. High-Q is not inherently bad, but ‘mismatched’-Q is. A high-Q driver is appropriate when matched to a ‘low-Q baffle’, just as a Low-Q driver is appropriate to match to a ‘high-Q’ Helmholtz enclosure.
The original Amazings had even higher Qts and lower Fs than the later Series. As someone in the forum suggested, lower Fs requires higher-Q, but this becomes problematic. For a fixed baffle cutoff frequency, with the standard 6 dB/octave high pass characteristic from cut-off down to the resonant frequency, the ideal high-Q gain would match that first order slope over the bandwidth from cutoff to fs. Unfortunately, the 12 dB of gain we needed in the original Amazings to achieve reference level at the Fs of 20 Hz required a Qts of about 4. This created a narrow band peak that doesn’t match the smooth, 6 dB roll-off slope of the open baffle. Because of this we had to use some shaping networks to smooth things out from 25 Hz to 100 Hz. We ended up with a fairly smooth response with a peak at 20 Hz and Fc at 17 or 18 Hz.
The later Amazings had a new lower mass paper cone woofer to replace the heavier honeycomb devices in the originals. We were able to get the moving mass down to less than 15 grams, which is quite low for a 12” woofer, but required for getting decent efficiency.
The Fs was raised to 30 Hz and the Qts was lowered to a value between 2.5 and 3. This worked much better in matching the Fs, Qts, and baffle Fc to get a smoother response and much higher efficiency.

I hope that I've managed to retain a modicum of respectability in your eyes.

TerryO

 

[Chaz H]

That's pretty amazing. So the perfect OB woofer requires a high Qts matched with an appropriately high Qms, but that Qms has to remain consistent at different power levels. Is that correct?

I can't help but ask, as a fellow Seattle-ite: does Jim Croft have a pair of Amazings which he'll let people listen to?

 

[TerryO]

That's pretty amazing. So the perfect OB woofer requires a high Qts matched with an appropriately high Qms, but that Qms has to remain consistent at different power levels. Is that correct?

That's also my understanding.

I can't help but ask, as a fellow Seattle-ite: does Jim Croft have a pair of Amazings which he'll let people listen to?

I really don't know if he has a pair, but there are a few in the general area. You may want to check the Audio Karma forum as I know that several local owners hang out there.

Best Regards,
TerryO

 

[mbskeam]

TerryO,.....
thanks,.the reason I ask is that the surrounds on my plats have rotted out.
and if you have seen theses you know there are no replacements out there....
and the company that is now getting the woofers reproduced, is mum on there supplier,there are Many of us out here that need replacement foams, not the whole speaker, and at 50bucks a pop this adds up.
and they sound okay but are not the same as the OEM woofers.....

who wants to hear which model.....

take a cruise thru these pics and you will understand.....
carver speakers pictures by mbskeam - Photobucket

 

[mikejennens]

Thanks Terry for that very interesting response. I have heard of a driver with such a high Qts so that must indeed be an oddity! I'd like to know: what do you mean by the Qms remaining linear?

Here's another question: I've seen it posited as a general rule that to determine the rolloff of a driver on an infinite baffle, you divide the Fs by the Qts. Doesn't that mean that a driver with Fs 35 and Qts of ~1.5 will extend to nearly 20hz as well? I'm looking at these GRS drivers for a possible cheap open baffle project and it seems like extension is no problem if efficiency is not your priority...

GRS 12PF-8 12" Paper Cone Foam Surround Woofer

Hello,
Did you ever do this build? I'm thinking of putting a pair of these drivers in an H-frame to augment the low end od a build I'm doing. I'm also looking at the GRS PR15-8 in an H-frame. I have a Dayton APA 150 I plan on using to drive them. Any suggestions or comments?
Thanks,
Mike

 

[freddi]

that's nearly unbelievable to get a 12" mmd or mms down to 15 grams. Here's some outdoor plots of a cheap Revere 12 - think it was called a DC1220 - I can see it looked clean on the sine but had great difficult playing a well recorded zarb drum - in contrast a 12" coax with 80oz ferrite magnet slug (and old model APT50 compression driver) in a slit vented little Karlson 12 did great on the drum. The slit vents really mess up sine near fb - thankfully some music is comprised of transients and not pure sine.


Revere DC1220 - fs probably ~50 - qts ~2 IIRC - graphed outdoors 2'x4' baffle with ~1ft deep wings

An externally hosted image should be here but it was not working when we last tested it.

 

[cradeldorf]

I tried the alpha's in an h-frame, worst sounding things I ever heard. But don't let me stop you from giving it a try.

 

[freddi]

here's an indoor measurement of a Goldwood GW1858 in my H-frame - (never measured system Q and its taken down) subjectively its a bit "ripe" for some classical music but tuneful on hip-hop - at least it seems to poop out gracefully rather than go chaotic. I think relatively low moving mass is on its side. I had a P-Audio neo 15" coax on a small open baffle -- qts for the woofer ~0.38 and liked it quite a bit so high qts may not really be needed.

An externally hosted image should be here but it was not working when we last tested it.