Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Pass Labs This forum is dedicated to Pass Labs discussion.

What watt? Why do we need all that power?
What watt? Why do we need all that power?
Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Thread Tools Search this Thread
Old 8th October 2005, 03:45 PM   #11
GRollins is offline GRollins  United States
diyAudio Member
Join Date: Feb 2001
Location: Columbia, SC
You're assuming that all comparisons are apples to apples.
Compare specifications on low frequency drivers and you'll find considerable variance in sensitivity/W. Worked out in terms of real-world requirements, it turns out that it takes an awful lot of power (and money) to get to 20Hz (or even 30Hz) with anything like flat frequency response.
Broadly speaking, you'll find two classes of drivers. One starts off reasonably flat, but rolls off in the 40-80Hz range. These drivers are generally quite efficient; at least low 90s dB, often higher. If you intend to go down to 20Hz it will take one to two octaves worth of eq to get there. (Watch the cone excursion--it increases by a factor of four with every octave and these drivers do not have much Xmax to play with.) As a result, the power requirement will go up. The other class goes much, much lower, but the frequency response is--to put it mildly--problematical. These drivers are generally at least 4-8dB less efficient, and require eq to remove the grotesque hump. In this case the eq isn't the problem, per se, it's the lower efficiency. Xmax isn't so much of a problem with these drivers. Keyword in sales literature: long-throw. A more subtle hint: lack of real, measured frquency response graphs--instead, they substitute T-S simulations since they don't want you to know about the uneven response. (And most people wouldn't know 20Hz if it bit 'em on the rump. If it vibrates the room, they think it's low, happy in their ignorance that it's actually 80Hz.) A non-subtle hint: the use of a melodramatic name for the driver such as The Blowhard, The Cat-Killer, or The Lease-Breaker.
I'm assuming that the driver in question in this thread falls into this second class.
If you're willing to stop at, say, 50Hz, your best option is the efficient driver with flat response, rather than try to pummel the "long-throw" driver into submission. You'll gain considerably in efficiency and can use a smaller amp. It's when you want something deeper that life gets difficult.


P.S.: I learned more than I wanted to know about this when I went with the Dayton Titanic drivers for my subs. Heed the voice of (painfully acquired) experience. I know whereof I speak.
If I had it all to do again, frankly I'm not sure which route I'd choose. Neither path is easy. And, yes, I demand real 20Hz response, not just simulations on a computer screen. Organs and such just don't sound right rolled off at 40Hz.
  Reply With Quote
Old 8th October 2005, 07:28 PM   #12
Nelson Pass is offline Nelson Pass  United States
The one and only
Nelson Pass's Avatar
Join Date: Mar 2001
What watt? Why do we need all that power?
Thanks for the swell name suggestions!

  Reply With Quote
Old 9th October 2005, 02:16 AM   #13
GRollins is offline GRollins  United States
diyAudio Member
Join Date: Feb 2001
Location: Columbia, SC
Butt-Buster, Neighbor-Knocker, or--if you're seeking a lighter, more romantic feeling (perfect for that overly sensitive significant other)--Heavenly Heartbeat.
Can you tell that I'm in the mood for alliteration?
How many names do you want? I can go at this for hours.
Seismic Shaker (okay, maybe that's a little too close to some of the real names), Richter Rocker...perhaps I should lay off the earthquake related stuff--they just had one helluva quake over in Pakistan and thousands are reported dead. Not funny. Put those names on the shelf for six months before using.
Ceiling Cracker (hey, I've got the cracks to prove it--twelve 12s will test your structural integrity with a Chinese Drum CD or Dafos 45 RPM vinyl).
Okay. Enough. I'm going to go listen.


EDIT: Speaking of names--I've always wondered about the "Pearl." There are many ways to look at that name, one of which (me being a tad cynical) is the old phrase 'Casting pearls before swine.' Or perhaps I shouldn't ask that in an open forum...
  Reply With Quote
Old 9th October 2005, 02:27 AM   #14
Tweeker is offline Tweeker  United States
diyAudio Member
Join Date: Dec 2004
Location: Cincinnati, Ohio
When I think of pearls, I always think of necklaces for some reason.
  Reply With Quote
Old 9th October 2005, 08:51 AM   #15
PRR is offline PRR  United States
diyAudio Member
PRR's Avatar
Join Date: Jun 2003
Location: Maine USA
> You're assuming that all comparisons are apples to apples.

I'm not even in the apple store. Not all possible combinations of parameters are on the market, no. But Natural Law will NOT be violated, and can't even be bent much.

> considerable variance in sensitivity/W.

Sensitivity numbers should be taken with a shaker of salt. Some are taken above 1KHz and reflect directionality more than power conversion. Efficiency is a useful guide, and never a Big Text Spec so if it is quoted at all, it is usually correct. If not quoted, it can be derived from the usual T/S specs.

> I learned more than I wanted to know

Grey, I've seen enough of your thought here to know that, despite a bad experience, you can and probably will understand it better. Sadly, coil+cone speakers were fully elucidated in 1927, all later research has clouded the issue, and nobody today seems to know why a coil+cone speaker can ever be flat over a decade of frequency. Neither the coil nor the cone is at all flat, but that's the key to serendipity.

> two classes of drivers. One starts off reasonably flat, but rolls off in the 40-80Hz range. These drivers are generally quite efficient; ... The other class goes much, much lower, but the frequency response is--... grotesque hump.

You are beating the bush close enough to see the tree.

Take a variety of drivers of the same size and varying efficiency. Tune them all the same: closed-box Q=1 is convenient, it does not matter. Plot them all at the same absolute scale. Ignoring a few poor performers (Qt too high for a Q=1 tuning, or just weak magnet), the bass-corner for all drivers of the same size falls on a 6dB/octave line. This IS the maximum efficiency possible for a driver of that size. If we want a flat broadband response, we pick a midband efficiency equal to the size-limited efficiency at the desired bass response. For a 15" in closed box, we can get 8% from 500Hz down to 140Hz, or 1% from 500Hz to 50Hz, or 0.25% from 500Hz to 25Hz. If the midband efficiency is not similar to the size-limited efficiency at the desired bass corner, we will get a slump or bump. A JBL 2222 (5%) and a Shiva (0.25%) will both have the same efficiency at the bass corner when tuned to, say, 32Hz corner; just that the 2222 rises from there and the Shiva falls. (I may be confusing Shiva with Tempest; whichever is 15" like the 2222.) And at the same bass frequency, both have nearly the same box volume (same if suspension were limp).

Go to 4-inch drivers and you find another 6dB/octave line, just shifted 4 times higher in frequency. There are a few 4" PA midrange drivers with efficiency near 4% but bass corner around 400Hz. For "full" range, a Four can't be designed higher than 1% efficient or it will sound gutless (it isn't gutless; the midrange drowns out the size-limited bass). Recent custom in Fours is to go to 0.3% efficiency for a gentle slope audible below 100Hz.

All our speakers are too small to make bass. It is not a fundamental design problem, we are just buying too-small speakers and making-up with monster amps. Back when a 15 Watt amp was "high power", theaters used 5-foot square radiating areas, horn-loaded to 2 or 4 fifteen-inch speakers. The area gave good coupling to 30Hz (the horn usually cut-off at 50Hz, all you need for excellent talking pictures). In modern PA practice, as in Hilbert's lab, we stack about 16 fifteens tight together, 5 foot square, and get good coupling to 30Hz, similar to a horn. Portability suggests box sizes that stifle response below 50Hz (more than you need for Rock) but such an array can be 20% efficient 50Hz-125Hz and 5% at 30Hz and 250Hz.

Instead of thinking in terms of home-size speakers, think in terms of the 20 foot wave you want to make, and big bass efficiency is easy. Only cost and size stand in the way. And because modest-efficiency speakers can array to high efficiency, and even 5% efficiency is "excessive" in context of 1% mids and tweeters, cost is not the killer issue in today's mass market. Radiating area and back-volume bigger than some apartments, and insane low prices for bulk BJT/FET power, is why we use teeny teaspoon drivers under 15 inches, and whine about huge excursion and all its ills.

Getting anything like the 5%-25% efficiency 30Hz-250Hz of a large array, at higher frequencies, is much tougher. In a direct radiator, 10% at 500Hz is excellent, and comes with problems. Compression drivers give leverage against motor mass, but they can't beat 10% at 2KHz or 1% at 10KHz. And even touching those points requires insanely heavy and costly motor magnets, and fun with midrange horns.

> twelve 12s

That's a good start in size. But a 12 designed to work alone broadband will not have low enough reference efficiency to work well in an array, and won't have a suspension limp enough to stay out of the way at 20Hz even with an infinite box. Dayton Titanic Twelve? It will be flattenable to 50Hz; perfectly reasonable for a solo 12. In infinite box, it slumps 50Hz down to 22Hz, and then falls bad. Probably needs over 12dB EQ to bring 20Hz level with 100Hz-200Hz region (not that you want an array that big working to 200Hz in a residential room!). The fix is to double the mass, sacrifice 6dB efficiency above 50hz for a few dB more around 22Hz, and then set the back-volume for optimum 20Hz output. And you "paid too much": 12 Twelves do not need 18mm and 500Watts to make all the 20Hz your home can stand. Short/small-coil small-magnet drivers would have worked, but there is no single-driver market for the parameters you need for a 20Hz array.
  Reply With Quote
Old 9th October 2005, 12:19 PM   #16
MikeW is offline MikeW  United States
diyAudio Member
MikeW's Avatar
Join Date: Sep 2002
Location: North of Boston
Default Why?

You don't need all that power. I built some ten watt amps and am very happy with them. Start small and work your way up if needed. Bi-amp if you need more bass.
Remember this site is for fanatics. Not too many people have a water cooled amp or put 4000 watts into el pipe-o.
  Reply With Quote
Old 10th October 2005, 03:08 AM   #17
GRollins is offline GRollins  United States
diyAudio Member
Join Date: Feb 2001
Location: Columbia, SC
I don't know whether the problem here is too much theory or not enough. You're leaving out cone breakup (not that it's not subject to analysis, just that such information isn't readily available). That's part--probably most--of the hump in these drivers; the cones aren't very stiff. I wouldn't go so far as to call them rubbery, but they're, well, fairly floppy. Tap them with your fingernail and instead of a "tip-tip" sound, you'll get something more like "thup-thup." As near as I can tell the cones on these things have a purely mechanical resonance all their own. Granted, it's old hat to use cone breakup to extend the high frequency response, but this trick of pumping up the 80-100Hz range is another matter entirely. Note that I didn't say that there's any new science--just a purely cold-blooded marketing decision to tweak the response by peaking the driver in that range.
The flatter drivers give semi-predictable passband response in a normal T-S enclosure. The "long throw" drivers don't. Choose a maximally flat enclosure and what you actually end up with is a response that looks suspiciously like a Q of 2, 3, or even 4. Tolerable pass band response, by which I mean, say, +-3 or 4dB, followed by this great galumphing peak right in the middle of the same frequency range that boom-boom car systems, PAs, and disco-dance systems have taught folks to recognize as bass. So they go home and hear the same sort of thing and think it's wonderful low end.
Accident? I think not.
You say flattenable to 50Hz? The keyword here is 'flattenable.' I choose to interpret that word as "eq." The actual, measured corner frequency is half an octave higher. Or worse. I've seen some graphs online that indicate that some drivers in this class peak at more like 100 to 120Hz. These things aren't flat to 50 followed by a rolloff, they're so-so flat to 150, followed by a large hump centered (in my case) at 80, followed by a rolloff. 12dB of eq to get 20Hz even with 100Hz? Unh-uh...more like double that.
Again, I am not positing 'new' science, nor am I breaking any laws of physics. It's just a question of manipulating the cone material and thickness (read: stiffness), to produce a peak right where it would do the most good in the average consumer's mind. Cone-related response problems do not show up in any of the usual numerical specs. The only way they give themselves away is an actual plot of a real driver in a real test chamber...the very thing that the "long throw" manufacturers don't show you. Golly, what a surprise!
You came at this from the other side when you mentioned cone mass. Stiffness takes mass. A stiffer cone would be more massive from the git-go, and would satisfy both of us.
A decade of response? Perish the thought. I'd be happy to get from 15-20Hz to 40 or maybe 50Hz, where my Tympani woofer panels could reasonably be expected to take over. Unfortunately, the 80Hz peak is so pronounced that it fouls up the crossover slope something awful. The best you can hope for (short of applying a notch filter set for the peak) is to set the crossover slope for something around 40 Hz and hope/accept that the declining slope and the rising response will combine to something like flat response until the resonance has had its way with the driver and the whole shebang begins rolling off somewhere around 90Hz. That's almost an octave higher than I want to run these things, although others may have other ideas as to what they want from a 12" driver. Another strategy I've played with is to start with a more modest 16Hz LP filter that more-or-less flattens the response until the resonance gets done, at which point the LP slope takes over--but again, you're faced with a higher crossover point. This sounds good on paper but for some reason is unsatisfactory in practice, even if you make allowances for the higher than desired crossover point.
Paid too much? Maybe. Maybe not. When I got these drivers, they were, what, $130-140 ea.--I forget. The flatter drivers were going for double that or more. I seem to recall looking at a $280 driver and thinking that was pretty inexpensive by comparison with its peers. So which route is cheaper? As I said above, even knowing what I know now, I'm not sure what I would do if I started from scratch. The ones I've got now aren't chrome plated or anything--they're actually pretty pedestrian looking compared to the drivers out there now. They're just cast baskets with a big magnet (no Allen bolts or anything, just glue) and this silly talc-filled poly-something cone. Rubber surround, by the way. Foam surrounds are an absolute non-starter in my book. I absolutely hate self-destructing driver suspensions. Every year or two, they claim to have solved the problem of foam suspensions rotting, yet the stupid things persist in crumbling to dust within ten to fifteen years. (Grey, you're ranting about foam surrounds again...stop it...)(Oh...okay...)
Arguing theory is wonderful. Great stuff. Many pints of beer can be consumed during such discussions. But if you want to go that route, remember to add the cone breakup (dictated by marketing soft-science, another interesting variable) to your model. Please, quit arguing with me until you've investigated these drivers. You're assuming perfect or near-perfect pistonic behavior, and that's undermining your entire argument. You're where I was before I bought into the whole long-throw thing. I knew I was going to need to eq to get as low as I wanted and 15s were rare then (the few that were available were unacceptable due to foam surrounds). 18s? Whazzat? Given the need for eq, I knew I was going to need a large Xmax and so I went with Titanics. I kinda knew what I was getting into, and I kinda didn't. That peak threw me. On reflection, it's easy to see where they're coming from, but it makes life difficult for someone who wants high fidelity rather than boom-boom. Bummer.
And, yes, I think we can both agree that sometimes the published numbers are sometimes...ah...optimistic, shall we say...

  Reply With Quote
Old 10th October 2005, 01:19 PM   #18
hilbert is offline hilbert  Sweden
diyAudio Member
hilbert's Avatar
Join Date: Sep 2005
Send a message via ICQ to hilbert
Seems like I've stumbled upon a hot potatoe here...
End of the line for myself, though, is that I have the long-throw in a small closed box. I could of course sell it to some dB-dragster boom-boom head and buy something else, but (again) in the spirit of Mr. Pass I like to use what I have lying around. So I guess eq is the long and narrow path to take. And that calls for power, I guess, as I will have to compensate quite a lot here and there. Especially if I want it to go down to 20Hz. Without having measured anything yet, I would guess the resonance freq. of the system is a lot higher than 20Hz, and with a 6dB/oct roll off, I have to insert a lot of power. What I will probably do is measure input electrical power to obtain equal loudness down to 20Hz at highest requested loudness level, using all kinds of signals; noise, tones and transients. As mentioned one difficulty is influence of room modes. The anechoic chamber we have is only valid down to about 60Hz. However, we have another room which is of roughly same size and shape of an ordinay livingroom, and it is at least semi-anechoic, so I could probably do measurements there for a worst case rating.
Some time way off in the future I would like to build something from scratch, selecting drivers and designs carefully and spare little cost to obtain flat response down to 20Hz. Oh, and yes, in honour of my nordic heritage I would of course call it Tor (or Thor) - god of thunder. The upper frequency range could then be taken care of by the more subtle Oden's (Odin's) - god of wisdom. How's that for a name Mr. Pass? Although I really like "Butt-Buster" too. It wonder how that would appeal to the gay community.

As far as power goes, maybe I should build a three channel zen-like amp, with two channels being the very fine super symmetric class A's as described in the articles by Mr. Pass, and the third being a less subtle AB, B or whatever other design that could pump a lot of power. It should then be fitted with a switch to either run the two Zen-channels fullrange, or the Zen-channels above say 70Hz and the third channel below. Perhaps also a "pre-out main-in" connector pair to be able to apply that compensating eq.
The name? Well, Asgaard, of course - home of the gods.

On a related note, Mr. Pass, You seem to have a fablesse for MOS-FETs. Do you by that mean that one shouldn't bother with bipolars? I think we might have some old tubes lying around at the department, and I was wondering if they could supply the voltage gain and then utilize darlingtons for current gain? I guess the problem is that they require very different supply voltages...
Could there ever be a tube/hybrid Zen? Maybe the tubes won't allow for the Zen-like simplicity...
  Reply With Quote
Old 10th October 2005, 08:33 PM   #19
Nelson Pass is offline Nelson Pass  United States
The one and only
Nelson Pass's Avatar
Join Date: Mar 2001
What watt? Why do we need all that power?
Originally posted by hilbert
Mr. Pass, You seem to have a fablesse for MOS-FETs.
I like all gain devices, but exercises in minimalism have put me
in the Mosfet (and now power JFET) camp as they are the only
devices that lend themselves to single-stage power amplifiers.
After I wear out the single-stage thing, I'll probably get around
to re-visiting Bipolars and multi-stage amps.

I doubt, however, that I will do any tubes for public consumption.
There are still too many other people flogging that subject.

  Reply With Quote
Old 11th October 2005, 05:04 AM   #20
tmblack is offline tmblack  United Kingdom
diyAudio Member
Join Date: Jan 2005
Location: UK
Any chance of seeing BJT in the front end of Aleph style amps?
I plan on during that.

  Reply With Quote


What watt? Why do we need all that power?Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
10 watt Simple Power Amp danville Solid State 16 29th June 2008 03:15 AM
WTB in Oz: 150-200 watt stereo power amp otto88 Swap Meet 4 20th December 2007 06:06 AM
First Watt F1 with 30W output power Adams_Leo Pass Labs 20 18th November 2005 03:14 AM
150 watt power amp bassplayer Solid State 1 23rd July 2004 11:59 PM

New To Site? Need Help?

All times are GMT. The time now is 06:08 AM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 16.67%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright ©1999-2017 diyAudio