The "Elsinore Project" Thread

Mine were first powered by an Integra DTR5.6. The AVR will do it no problems at all. My Elsinores are used for 2 channel and HT and I very rarely feel the need to add a sub to the HT experience. The Blue Man Group DVD will hit a clean 110dB at my chair (about 3 metres back from the speakers) during some of the drum solos, with the Integra at an absolute volume of 80 (which is the "max volume" limit I have set on it).

Hope that helps.

Cheers,
Rob.
 
WS6Fury said:

1) Is there a room size recommendation for these monsters?

2) Is there any special receiver requirement to push the speakers? ( I would be working off a Yamaha 3800)

3) Will this design work for low volume levels in HT use at night time (obviously made for ****ing your neighbors off hehe :cool


1) I use them in my current sound room which is 7M x 4M x 2.6M - I would consider medium size. But very small rooms may not be the best way to show them off.

Are they big enough to be called "monsters"? What you think Rob? Is a 3 cubic foot (for our American friends) a big speaker? Probably, but not unwieldy so, I would think.

2) I use them with my current 20 Watt JLTi EL34 Triode amp as seen on www.customanalogue.com and can generate significant SPLs. Don't worry, they can be physically felt in that room. I love when I have unsuspecting listeners over and let the system rip - Deep Purple "Child in Time" for the In Rock album, then afterward asking how many watts per channel they are listening to. They look dumbfounded when I tell them. :)

3) Some speakers, specially the usual American speakers designed to go with muscle amps, they just don't get lively until you pump a lot of current into them. Lynn Olsson has remarked on this. In my view it is a case of dynamic compression where you need volume to compensate, hence you end up having to turn the volume up. As Rob and other users will easily testify, the Elsinore are highly responsive because the are more voltage sensitive than current sensitive. The original concept behind the Elsinores was EXTREMELY low dynamic compression. I know many will doubt my next statement, but most amplifiers suffer from worse compression than the Elsinores and in fact the Elsinores are limited by the amp in this respect.

Joe R.
 
Hi Joe,

What is the chance of a demonstration at the Melbourne Audio Club? :)
I’m not a committee member, but I think our diy oriented Club would really like to hear your speakers

> the Elsinore are highly responsive because they are more voltage sensitive than current sensitive

Pardon the off topic question, but when designing speakers, is there/ are there driver parameter(s) that you can share, that guide you to voltage sensitive drivers or a voltage sensitive system?

Thanks
 
otto88 said:
Hi Joe,

What is the chance of a demonstration at the Melbourne Audio Club?

> the Elsinore are highly responsive because they are more voltage sensitive than current sensitive

Pardon the off topic question, but when designing speakers, is there/ are there driver parameter(s) that you can share, that guide you to voltage sensitive drivers or a voltage sensitive system?

Thanks

I would love to demo them down there but I haven't been to Melbourne since the 90's and nothing on the horizon to indicate when.

It's the "array" rather than single driver in the Elsinore that makes them voltage sensitive.

Briefly here are the basic maths:

Single driver 8 Ohm 90 dB sensitive nominally = 1 Watt.

Two drivers, Parallel 4 Ohm, 96dB = 2 Watt (half the impedance doubles the current).

Two drivers, series 16 Ohm = 1/2 Watt (halving the current).

Series/Parallel configuration = 1 Watt (same current as a single driver) =96dB.

Peel back to 90dB requied SPL and current is a quarter.

Divide required 1/4W between four driver = 1/16th of a Watt.

The fact that each driver now only has to dissipate 1/16th of a Watt for the same target SPL, the voice coils in the individual drivers now run very cool indeed leading to a massive reduction in dynamic compression... and of course distortion.

Acoustic efficiency gained = 6dB, that's four-fold. That tells the whole story.

There you have the original brief for the Elsinores going back to around 2002.

Joe R.
 
Hi Joe

> I haven't been to Melbourne since the 90's

Shame, but yes Melbourne is a fair way to schlep the speakers.
Are their any owners in Melbourne who might let a potential punter have a listen?? :)

While I understood the big reduction in dynamic compression, and knew that 4 drivers Series/ Parallel gain sensitivity and lower the ohms load, I didn’t realise that ~ if I understand you correctly

> Peel back to 90dB required SPL and current is a quarter.

Current sensitivity is 4 * a single driver

> Divide required 1/4W between four driver = 1/16th of a Watt.

Power (voltage) sensitivity is 16 * a single driver

With excellent drivers as modestly priced as the HDS, I wonder why more manufacturers and diyers don’t do it?
I’ve read most of the material on your website; with 4 drivers, does some lobing occur?

Cheers
 
otto88 said:

With excellent drivers as modestly priced as the HDS, I wonder why more manufacturers and diyers don’t do it?
I’ve read most of the material on your website; with 4 drivers, does some lobing occur?

Cheers

There seems to be a reticence re connecting drivers in series. The bottom two drivers are in series and gradually rolled off from a couple of hundred Hertz up - less than single order, so they form a 16 Ohm combination. The same applies to the two upper drivers which form a point source with the tweeter, so you have two different filters seeing 16 Ohm because of series connection. For some reason - and it's a blind spot - this just doesn't seem to happen.

IF I was to do a Two-Way 3 driver system a la D'Appollito (using same drivers as Elsinores) I would seriously consider doing the opposite what everybody else does; they parallel them and you get a 4 Ohm speaker system, I'd consider doing it as a 16 Ohm combination. A 90dB speaker would stay 90dB but only draw 1/2W nominally. I haven't done yet but I might try it.

Joe R.
 
Joe Rasmussen said:


There seems to be a reticence re connecting drivers in series. The bottom two drivers are in series and gradually rolled off from a couple of hundred Hertz up - less than single order, so they form a 16 Ohm combination. The same applies to the two upper drivers which form a point source with the tweeter, so you have two different filters seeing 16 Ohm because of series connection. For some reason - and it's a blind spot - this just doesn't seem to happen.

IF I was to do a Two-Way 3 driver system a la D'Appollito (using same drivers as Elsinores) I would seriously consider doing the opposite what everybody else does; they parallel them and you get a 4 Ohm speaker system, I'd consider doing it as a 16 Ohm combination. A 90dB speaker would stay 90dB but only draw 1/2W nominally. I haven't done yet but I might try it.

Joe R.

Hi Joe,

It really doesn't make any difference form a power perspective. Two drivers yield a 3dB increase in efficiency any way you connect them; in series or parallel. You will require 1/2 watt to have the same SPL as a single driver drawing 1 watt either way. The underlying consideration should be on what load the amp operates better into. If the amp is capable of high voltage swings but low current use series. If capable of lower voltage swings but high current use parallel.

But there is another reason to shy away form series connections. In a series connection the back EMF for each driver modulates the other driver. How much of a problem this is would depend on the degree of differences between the drivers. It may not be too significant, but parallel connection avoids the interaction for the most part.
 
Member
Joined 2004
Paid Member
Re: Alternate pvc pipe sizes

Ed LaFontaine said:
Francios G,

I obtained 3 1/2" pvc conduit (3.5" i.d.) from my local electrical supplier. It seems they stock sizes that aren't available from plumbing suppliers.


Ed,

Thanks for the pointer - much appreciated. I will check the local electrical suppliers, followed by US-wide ones.

Francois
 
john k... said:

It really doesn't make any difference form a power perspective. Two drivers yield a 3dB increase in efficiency any way you connect them; in series or parallel. You will require 1/2 watt to have the same SPL as a single driver drawing 1 watt either way...


Hi John

Good to see you join in.

Agreed, as you say it comes down to the amplifier. The real benefit kicks in when using a four driver array covering the same frequencies and series/parallel - then you get a real 6dB @ 8 Ohm and THAT is significant. I freely admit it, I have not done a 16 Ohm system in the way I described, but being a zero feedback tube amp user I wouldn't mind trying a D'Appolito three driver 2-way design and just try series/16 Ohm - and if I didn't get a desirable result, revert to 4 Ohm. Anybody out there with sixteen Ohm taps on their tube amps? :)


But there is another reason to shy away form series connections. In a series connection the back EMF for each driver modulates the other driver. How much of a problem this is would depend on the degree of differences between the drivers. It may not be too significant, but parallel connection avoids the interaction for the most part.


I used to think along those ways too. BTW, the one who made me change my mind was a conversation, and a kind of disagreement, with a Mr R. H. Small. We had a wide ranging talk on series connection, even adding series resistance to achieve an erosion of Qe (I really balked at that one) to achieve a desired alignment (would you like me to expand on that as this is surely controversial, especially considering the source). But he had no problem with any series connection with the proviso that we are talking similar/same drivers. I have done a number of designs using series connection, even a 3-Way with a 10" Bass driver, two 5" mids with a tweeter between them forming a point source above 350 Hertz. The two mids were connected in series. The combined SPL was the same as a single driver and 16 Ohm. They were mounted on an open baffle above the bass bin. That design has been heard by more people I could count and was very highly appraised. The midrange had some really nice attributes and it wasn't just the open baffle that did it.

In the Elsinores the idea was carried on as it is 16 Ohm nominally in the midrange and 8 Ohm below that. But the two lower drivers are also in series but they fill in the diffraction loss and balanced against the two upper full range (up to the Tweeter) so there are TWO series elements - but they combine to 8 Ohm in the lower midrange. If we are to think in terms of driver differences causing problems in the bass, this simply is not the case. The bottom end is VERY capable. :D

Bottom line, the results have been good and I have problems doing it again and likely will.

Cheers.

Joe R.
 
otto88 said:
I thought all four formed a line source; do you know how (only) the two upper drivers form a point source?

Cheers

As the above discussion with John indicates, the effect only occurs when the drivers cover the same band of frequencies. But I am backing a little away from claiming the line source effect as it only occurs down to a certain frequency. But ultimately at LF they all become pulsating sphere or gradually so with descending LF. But this explain why the top three drivers can become a point source as the other two bottom 6.5' drivers are falling away at the mid frequencies. Hope that helps, but this explanation is on the website.

Cheers.

Joe R.
 
Here is my 2 cents worth

I think there are two factors at work here :

1. Dynamic power compression is really a measure of driver sensitivity.

ie 92db is close to 1% efficiency for 1 watt.

Therefore a large portion of the electrical input power must be dissipated in heat, the harder you push the lesser sensitive driver the hotter the voice coils get..basic physics.

The heating effect in woofer voice coils is not short term but more so medium to long term because of the size of the voice coil and the low frequency program crest factor.

Unfortuately steel pole pieces are slow to absorb heat so once the voice coil gets hot and the the pole heats up the voice coil will say that way for a while. Venting of poles and spiders helps get rid of the heat but heat still happens in voice coils.

Only way around this is large 4 or 6 inch voice coils and high sensitivity like 3% or 98.5 db sensitivity

The benefit of Joe's design is you share that heat across 4 voice coils even though the drivers have relatively low efficiency.

Therefore the onset of power compression will not ocurr until there is 4 x the arbitrary power input for the driver to start entering power compression.

Each driver is about 0.3 % reference sensitivity.

http://www.tymphany.com/hds-platform
http://www.tymphany.com/830875

2. Line array effects power response.

By arranging the woofers in a vertical line array this focuses the lower frequencies in a narrower vertical pattern as it does with the HR array above 1000 hertz with just two woofer.

Ths focus adds punch and impact as there is simply more direct that reverberant power reaching the listener th with one woofer.

Ian
 
Joe Rasmussen said:
Hi John

Good to see you join in.

Agreed, as you say it comes down to the amplifier. The real benefit kicks in when using a four driver array covering the same frequencies and series/parallel - then you get a real 6dB @ 8 Ohm and THAT is significant. I freely admit it, I have not done a 16 Ohm system in the way I described, but being a zero feedback tube amp user I wouldn't mind trying a D'Appolito three driver 2-way design and just try series/16 Ohm - and if I didn't get a desirable result, revert to 4 Ohm. Anybody out there with sixteen Ohm taps on their tube amps? :)

I had some EAR 509 amps for which the output could be configured for 16 ohms. But I soild them years ago.

We had a wide ranging talk on series connection, even adding series resistance to achieve an erosion of Qe (I really balked at that one) to achieve a desired alignment (would you like me to expand on that as this is surely controversial, especially considering the source).

Adding series resistance is pretty common knowledge these days in the DIY circles I travel in. As is using the second coil of a dual VC woofer to control the alignment.

But he had no problem with any series connection with the proviso that we are talking similar/same drivers.
Cheers.

Joe R. [/B]

Same... that is the caveat. Same has to mean identical parameters, identical loading... Tolerances on T/S parameters are pretty wide. And that is only looking at the linear side of the picture. When we consider distortion we have the motor nonlinearities of one driver affecting the other. I have a design that uses dual woofers and I can clearly hear the difference between parallel and series connections. Now, this is a dipole like application and excursions are being pushed to the limits so these nonlinearities are probably more significant in this case. As I said, I don't know how significant these effects ultimately are in general as I have not taken the time to measure them, but they are worth thinking about.

I understand you reason for the series/parallel connection in your system and I would agree that it is why correct path to follow in a 4 driver application. Not too many of us have amps that are happy with 2 ohms loads. :bawling: Now if I could just find a way to modualte the output of my arc welder.
 
> ultimately at LF they all become pulsating sphere or gradually so with descending LF. But this explain why the top three drivers can become a point source as the other two bottom 6.5' drivers are falling away at the mid frequencies. Hope that helps

Thanks Joe, it does.


To modify the earlier question, with two drivers, if a wanting to use tube amp which has say 4, 8 and 16 ohms taps, and if SPL is sufficient either way:
would a tube amp more likely operate better with the two drivers wired in series or parallel?

Thanks