[b_online]

I recently auditioned a pair of your 1801s in London at a friend's house that just moved from the states.

In fact, I liked them so much that I borrowed them for the rest of the weekend with the hope that they would work well
with my amplifiers.
Unfortunately, my Jadis JA30 amps, while still being able to play them at loud levels wasn't able to make them dance
with proper dynamics.

Since listening to the 1801s , I have been bitten by the DIY bug, and I would like to build my own loudspeakers.
I would like to ask you a favour: could you please give me a heartful recommendation for an easy to drive pair of diy
loudspeakers I could build here in London?
(Something that you consider to be a special gem yourself)

Ideally, I think I would need a sensitivity of around 92 dbs. (Also, if possible, Id prefer to build a pair of
floorstanding speakers)
Price range: around $2000 max

Out of the commercial loudspeakers I have been able to listen so far, I would say these are my favourites (in no
particular order):
Living Voice Avatar (very musical, quite dynamic and very intimate)
Amphion Krypton (an overall alrounder, though sensitivity is on the low side for my amps)
Klipsch LaScalas (although not as intimate or refined as the Avatar, they have awesome dynamics with my amps)

And one of the ones that I least liked so far (out of the big names) was the Audio Physic Virgo speakers...... Not
musically involving at all.


I hope you can give me a hand,,,,,

Many thanks in advance,

B

PD- At the moment I am planing on building a kit I found using a Raven R2 (12 db/oct) plus a Supravox 285 driver in a bass reflex enclosure

[David Ellis]

Your concern is very valid.  Today in hifi the trend is toward smaller/cheaper cabinets, less sensitivity and higher thermal compression/distortion.  According to TC sounds, this is the #2 contributor to motor distortion in cone units.  I am sure many folks would agree that more sensitive drivers have a sense of dynamics not present with less sensitive speakers.

The recipe for a more sensitive speaker is simple.  The bass driver and cabinet are the limiting factors for sensitivity.  It's easy to find a sensitive tweeter, but not so easy/cheap to implement a sensitive woofer.  All things else being equal (i.e. Bass extension and Q) every 3db in sensitivity will cause the cabinet size to double.  Obviously the driver parameters must change, but the limiting factor is the cabinet SIZE.  In this scenario for the 1801 obtain 87.5db/watt, the cabinet size must double.  For it to obtain 90.5 ohms, it must again double.  For it to obtain 93.5 ohms it must again double.  

Before going further, I must explain that there is some "cheating" happening with regard to impedance ratings.  This is unregulated and therfore unpunishable.  Many manufacturer will list a "nominal" impedance of 8 ohms, but their speakers will really dip to 4 ohms.  The word "nominal" is not regulated and not enforced.  If any speaker lists 4 ohms anywere in the impedance rating, the 4 ohm impedance is in the woofer region where most of the juice will flow from your amplifier and where the cabinet size is effected.  

Since speakers are generally rated a 2.83 volts, a 4 ohm speaker will produce 3db more spl than an 8 ohm speaker.  This is simple math.  IF you wish me to explain this, I can.  The rating is gernally db/volt.  If, however, the rating were db/watt, the spl produced would be the same between the 4 and 8 ohm speaker.   As an aside, I believe the onset of 4 ohm speakers is bad for purist hifi.  I can explain further if you like.

Okay, back on the subject...

To increase cabinet sensitivity, starting with the 1801 sensitivity, a speaker would have to grow thus:

22 liters  - at 8 ohms 84.5db/2.83v  // at 4 ohms 87.5db/2.83v
44 liters  - at 8 ohms 87.5db/2.83v  // at 4 ohms 90.5db/2.83v
88 liters  - at 8 ohms 90.5db/2.83v  // at 4 ohms 93.5db/2.83v
176 liters - at 8 ohms 93.5db/2.83v  // at 4 ohms 96.5db/2.83v

It's easy to see how the cabinet will grow dramatically as the sensivitiy increases.  Given the commercial cost of a big cabinet, these speakers are indeed very expensive.  

Herein I will offer some personal subjective opinion.  The Tannoy Churchill in that big whoppin' cabinet sounds very good.  That big paper cone will flex, but there is relatively little thermal compression in that speaker.  Given a SET amp, I believe the most appropriate speaker is a VERY big speaker with VERY high sensitivity.

There are many sources of distortion in modern loudspeaker drive units.  There is motor linearitiy distortion, spider linearity distortion, cone flex, magnetic variation distortion (uneven gaps), cone flex, cone resonance, frame vibrations.... and thermal compression.

So... my recommendation is potentially twofold.

1.  Buy a pair of big old Klipsch speakers, and update them with modern crossover components.

2.  Danny Richie did this project a few years ago and the parts are still available  http://www.gr-research.com/discontinued/alpha.htm .  The only question I would have about this project is why Danny Richie dumped it.  I suggest you call Danny via telephone and as him this question.  Danny is a good guy and will be very honest with you.

At present, there are NO DIY speakers at 90db+/8ohms with high quality stiff cones and low distortion motors (i.e. bigger 1801) for tradional single amplifiers.  Eventually I may change this.  

There are other issues regarding amplifiers too, but... I must go to work.

Dave

[JoshK]

I love that Dave took the time to answer you in great detail even though this isn't related to his speakers and his sales.  Dave has shown that he is truely a great guy.

I think one of the more intriguing HE DIY designs out there is the BassZilla.

But alas, as Dave says, there is no getting around the big box for bass if HE is a requirement.

[bhobba]

Hi B Online.

I have read Dave's excellent reply and do not necessarily agree.  The PHL 1220/1230 drivers are very sensitive, among the best in the world, and will go down to 44hz in a ported enclosure.  Zalitron offer a kit with them in that are audiophile grade (see under the audiophile section).  
http://www.zalytron.com
You could save a little by using the Arum Cantus or Fountek tweeter instead of the expensive Raven (may require some slight crossover mod).  Some people whose opinion I respect claim that are as good or at the most only slightly inferior.  

Thanks
Bill

[David Ellis]

The PHL 1220/1230 drivers are very sensitive, among the best in the world, and will go down to 44hz in a ported enclosure.

Please clarify.  Given modeling software using the standard T/S parameters (not always accurate) what size cabinet, alignment (port size/length), and F3 can you obtain.

Specifically, are you inferring that somehow PHL manages to cheat physics with regard to sensitivity, cabinet volume, and bass depth?

Dave

[Mudjock]

I plugged the 1220 T/S parameters into Unibox and it recommended a really small box (around 6 liters) with an F3 of 91 Hz.  PHL recommends a 12-15 liter box tuned to 45 Hz, which seems to give an EBS alignment that is about 6 dB down at the knee frequency of 50 Hz.  F10 is in the low 40's.  The bass could be extended a bit more by adding full BSC to the crossover, since the response is already down a couple dB at 200 Hz, but that would be eating into the sensitivity a bit.  So, although the PHL may be a very interesting and potentially very good driver, the laws of Physics still appear to hold.

[bhobba]

Hi All

In no way was I claiming the laws of physics had been violated.  I have not run the parameters of the PHL driver through simulation software and have no reason to doubt what Mudjock said.  I was basing what I said on Zalytron which claims a response down to 44hz.  Of course what that means is open to interpretation and is in itself rather misleading due to room modes.

That being said I am a bit confused.  Having reread both B Online and Daves post I can see nothing in there about bass response.  My concern had to do with Daves statement - 'It's easy to see how the cabinet will grow dramatically as the sensitivity increases.'.  The PHL's are very sensitive drivers and are available in 16 ohms with a sensitivity of 90db at 2.83v.  In parallel they would be 8 ohms and give 96db at 2.83v.  This is plenty sensitive to my way of thinking.  My focus was not on bass respnce - just sensitivity.  If one wants deep bass to say 30 hertz from a 6.5 inch driver then one is going to be looking for a long time IMHO.

As an aside I personally am not a big fan of ported enclosures and believe in the use of closed boxes, TL's, open baffle etc with a sub-woofer - and an actively compensated closed box one at that - but each to their own.

My recommendation to B Online for what it is worth is, if he still wants a sensitive speaker, consider a closed box PHL and fill it out with a subwoofer. But I also want to add that I do no necessarily understand the reason for sensitive speakers - high quality amplification is available at a reasonable price.  About the only reason I can think of is you want to use low powered valve designs. If sensitivity was no longer an issue then B Online might like to consider a closed box 1801 which I believe Dave also sells.  Couple that with say a Warfdale closed box subwoofer then you may have something that you would find 'dance with proper dynamics.' even down to deep bass.  BTW the above is not a recommendation for the Warfdale - merely a subwoofer I have heard and think is OK but a bit pricy for what you get.  

Thanks
Bill

[JoshK]

The Seas Excels have flatter response than the corresponding sized PHL's (which means lower linear distortion) but I do hear a lot of good praise for the PHLs and I am intrigued at least to hear them for myself.  If one is willing to suppliment them with bass drivers then a moderate sized box can be accomplished, but still the physics apply so the box will be larger for a given cutoff than the Seas.  

Personally I think subs + monitors is the way to go anyhow for most reasonable setups, but who said any of us were reasonable?  

[bhobba]

Hi JoshK

I agree.  Despite the fact PHL drivers are among the best in the world they have paper cones and regardless of how well implemented IMHO can not give the same detail as Magnesium cones - as I think a perusal of Dr Linkwitz's measurements on his site will quickly show.  The only reason to use them IMHO is for low powered setups such as the some of Audio Note valve stuff.  Although it must be said that due to their graceful clipping valves can sound a lot bigger than they actually are.

In fact B-Online may find that once the Ellis 1801 is relieved from outputting deep bass with that part handled by a subwoofer the 1801's may sing to his/her satisfaction even with his current amp.  Worth a shot IMHO.  

Yea - as you can see I am a big fan of subwoofers and not just because of the deep bass they can deliver.  Once the mid bass is relieved from trying to reproduce deep bass it will produce less distortion and not 'shake' the speaker as much giving improvements in other areas. In fact I am very seriously considering basing my next system on the DEQX because I believe with modern low distortion drivers like the Seas magnesium the room is by far the biggest problem.

Thanks
Bill

[David Ellis]

Bill,

Kindly, there are several areas to address.  I sincerely hope this response is eductional.  I also hope it isn't to wordy.  A telephone conversation might be helpful.  If so, I have no problems discussing any of this via telephone - or any of the above information.  

I notice the Mention of Mudjock.  HE understands physics.  I suspect his knowledge of physics probably surpasses everyone who posts at audicircle.    My knowledge of physics is focused purely on speakers, and is not very deep in other areas.  Getting past the specifics of loudspeaker application for me is difficult.  However, in the realm of speakers my understanding is fairly solid.  His modeling appears correct.

I have heard at least 1 very stiff paper cone that was extremely rigid and had a nice cone resonance to prove it.  I think this paper cone had the resonance of a metal cone.  Also, not all metal cones are rigid.  The newer SEAS curvalinear cone on the L18 driver has less output at 2khz than the old straight-sided cone.  Someone from SEAS commented this was due to cone decoupling.  Another word for this is flex.  Even the metal curvalinear cone will flex before 2khz.  The flex happens when the inside and outside of the cone are not moving like a piston.  The voice coil is pushing harder than the cone can endure.  The edge of the cone doesn't maintain pace with the cone.  So, while it seems that all metal cones are pistons - this isn't true.  It's unfortunate.

I have heard some very good paper cones.  Indeed PHL implements a very good paper cone.   However, I don't believe they are stiff.

I will make a statement of belief here:  All stiff cones will have a cone resonance.  I am about 98% sure this is true, but don't really have the physics backround to prove this.  I watched a few professors prove formulas, and learned many of them.  However, this problem is beyond me.

I was basing what I said on Zalytron which claims a response down to 44hz

.... Can you model T/S parameters?  

If one wants deep bass to say 30 hertz from a 6.5 inch driver then one is going to be looking for a long time IMHO

... Bass depth ability from any driver is based upon it's Qts and Fs .  The cabinet size required is dependent on VAS.  While these numbers are somewhat related to size, the correlation is extremely far from 1:1.  At a glance, the PHL driver's stock parameters Fs of 45hz and Qts of .27 will push an F3 of 85hz.  It might have some energy at 45hz but this is a marketing stretch.  In this vein, the 1801 has some energy at 27hz too.

Hmmmm.... are you interested in actually learning about T/S parameters and how they apply to bass systems?  If so, I could dig-around for my older copy of the Loudspeaker Design Cookbook and send it to you.

I like sealed woofer systems too.

I am a fan of dual subwoofer setups.  I am not a fan of single subwoofer setups.  The latter always seem to eschew the soundstage for my ears.

I sincerely hope I responded to all of you concerns.

It's late and I need to sleep.  I do plan to eventually address the amplifier issue specifically....

ZZZZZzzzzz

Dave

[bhobba]

Hi Dave

Thanks for the reply.

After reviewing what was said I believe there was some misunderstanding about what was meant.  I was purely addressing sensitivity - not bass extension.  Getting loud bass from  a small driver in a small box without using a lot of power is simply not possible - the laws of physics do not allow it.  And yes I have some background in physics being formally trained and self taught in mathematical physics and am a regular poster to physics newsgroups - although my interest is in relativity, QM and QFT rather than acoustics.

I have a copy of Vance Dickinsons classic (somewhere at least - in may have gotten lost in a recent move I did), have read it, and believe I understand what it says.  I also have software that models speakers.  While not the area in physics that I know well I am familiar with the principles of loudspeaker design and what those principles tell me is this.  Overall speaker efficiency is not affected by box size except as far as bass extension is concerned.  Over about 200hz or so there is no difference in the output of a driver compared to cabinet size. So, in considering the overall response up to say 2k there is little impact on overall sensitivity regardless of cabinet size.

I may have expressed myself badly but I have no problem with what Mudjock said - I in fact concur with it.

My comments about the seas driver has to do with the measurements made by Dr Linkwitz.  Although he did not test a PHL driver I would be very surprised if it had distortion lower than the seas - or even equal to it for that matter.  Magnesium is inherently more rigid than even treated paper so logic indicates it should be lower.  This is not just my opinion - others such as Dr Linkwitz have stated such as well.  The problem with the Seas driver IMHO is resonance problems and its sensitivity - which his only a problem if you have a low power amp.  

I also agree with your comment about dual subwoofer systems - I believe they are superior to single subwoofer systems.  However I believe a single closed subwoofer system is superior to the booming bass of ported systems - they always sound bloated to me - with a caveat.  I have a single ported subwoofer - a Richter Evocotour that reaches down really low - 15hz I think.  This does not suffer the same problems largely because at such low frequencies you feel it rather than hear it - my main speakers reach down to about 50 hz easily - at least that is where I crossover.  They are ported as well and the bass IMHO is not as good as a closed system.

I am quite willing to continue this over the phone.  But the bottom line is now I understand you are referring to bass I do not disagree with anything you say so I suspect there are no actual concerns.

Thanks
Bill

[David Ellis]

If you seek a good musical subwoofer for your system, the SCC300 is killer, and very cheap.  It has a good stiff cone indicated by a metal-like resonance at 800hz.  The bass is very dynamic, and the bombs in the movie Pearl Harbor sound very... tuneful.  I believe one of the reasons is a very high sensitivity.  This woofer is also darn cheap at $99.

The only downside is this woofer requires a 3-4 cubic feet sealed to work best.  The high sensitivity (and dynamic ability) and bass depth come wih a cost - a big cabinet.  As you cite above, for a bass drivers, higher sensitivity reqires a bigger cabinet.  Such is very true with the SCC300.

All in all, the SCC300 is the best sounding woofer I have heard.

I also agree that above 200hz cabinet size is not really a factor.  I must admit not understing your point in this regard.  I thought the subject was woofer enclosures/systems.

And subjectively... I think sealed bass with an F3 of 40 is the perfect set of compromises.

Dave