The "Elsinore Project" Thread

Proof of concept.

Picture006.jpg
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Joe Rasmussen said:
Hi Guys...

In addition to above it retains the use of series connection of the main drivers. In all honesty I have not seen this used in any commercial speaker and come to think of it, I can't recall any diy designs either. Normally we would connect the drivers in parallel and get a 3-4 Ohm speaker. Yes, the voltage sensitivity would likely match the Elsinore, but the current required would quadruple, so efficiency is not improved, just a heavier load on the amp. I make no excuses for being in a particular camp (a minority one I would say).

OK, over to you guys.

Cheers.

Joe R.

Not to feel alone in series stuff, here is a design of mine, ''The Slim'' as it was published in Positive Feedback Vol.8, #3 Winter, 2000, when PF was still a paper mag. That one hit 93dB at 8 Ohm and was not dipping, so it was endorsed as valve friendly under Doc B.'s (Bottlehead) World of VALVE section.

TheSlimPFvol8-3.jpg
 
Salas said:


Not to feel alone in series stuff, here is a design of mine, ''The Slim'' as it was published in Positive Feedback Vol.8, #3 Winter, 2000, when PF was still a paper mag. That one hit 93dB at 8 Ohm and was not dipping, so it was endorsed as valve friendly under Doc B.'s (Bottlehead) World of VALVE section.

Hi Salas

You have nack of hunting these down. :)

But just two drivers in series (your example is series parallel and I'm a fan of that too, in a manner) is still something I haven't seen, especially 2 x 8 Ohm = 16.

When I get Rob's Hamlet boxes, I will also model parallel. We can the model both and compare what we get. We could end up with two variations, Hamlet 16 and Hamlet 4. Here in Sydney a friend of mine, Brad Serhan, has designed the Orpheus Loudspeaker's Aurora 2 and 3. These are 4 Ohm, but a particularly easy and resistive load.

www.orpheusaudio.com.au/?d=products&p=aurora-2-3

Both the 2's and 3's are nice and easy loads. Indirectly, and I don't think Brad will object, I've had a hand in those designs, but he was in the driver's seat and it's his designs. For commercial designs they are also good value.

What I am getting at, if the Hamlet also lends themselves to easy 4 Ohm, then that will also be seriously looked at.

Joe R.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
I have made medium size mirrored box 2X5.5inch carbon cone Audax HM130C0 (like those in early Von Schweikert) plus a tall purely resistive 8 Ohm ribbon on their side, woofers wired in series 8+8=16. Easy load, soft sound. When I got em in parallel, by adjusting the crossover and normalizing the SPL so to compare fair, maybe more difficult for small amps, but there was a very noticeable gain in transparency and attack, leaving the sensitivity gain aside. Best compromise is series parallel given the woofs are many, in my experience.
 
Salas said:
...Easy load, soft sound. When I got em in parallel, by adjusting the crossover and normalizing the SPL so to compare fair, maybe more difficult for small amps, but there was a very noticeable gain in transparency and attack, leaving the sensitivity gain aside. Best compromise is series parallel given the woofs are many, in my experience.

I hear what you say, but maybe I'm into "soft sound"? Mind you, never been accused of that. As for series parallel, that is obligatory in line sources and I am of fan of that. One day I want to do a two metre high line source, great potential genuine efficiency and equally potentially low distortion.

But will most likely let people opt for both variations in the Hamlet, it all depends how the results stack up.

Joe r.
 
Joe Rasmussen said:


Proof? In these days of photoshopping, can one trust one's eyes?



:rolleyes:

Yeah, while I was practising my photochopping skills, I did these up at the same time ;) .
Each Cabinet weighs 20.7kg according to my bathroom scales (but my wife reckons the scales are inaccurate, strangely though, only when a female hops on them).

Picture007-1.jpg

Picture001.jpg


They are quite large, but alot easier to handle than the Elsinore cabinets.
I haven't routed the handle hole (hole for the port) yet as I still haven't gotten some pipe for it.
 
rob323 said:

They are quite large, but alot easier to handle than the Elsinore cabinets.
I haven't routed the handle hole (hole for the port) yet as I still haven't gotten some pipe for it.

Surely there has to be something between 50mm and 70mm OD? I seem to recall the old 3" pipe and that should translate to about 75mm. That might just about do at a pinch.

This mob makes 65mm in 6M length:

www.pipeking.com.au

NEWCASTLE - NSW
40 SANDRINGHAM AVE, THORNTON, 2322
Phone: (02) 4966 0477
Fax: (02) 4966 0488
Email: newsales@pipeking.com.au

But there must be retailers that will sell 1M.

Joe R.
 
EUVL said:
Actually what is the pipe size (dia & length) for the Hamlet?
Same as Elsinore?

I searched but it wasn't mentioned explicitly elsewhere......
Patrick

Exactly, because it hasn't been determined yet. Diameter cannot stay the same as the Elsinore as that would make the port way too long, so to make the length manageable it has to have a smaller diameter.

So it comes down to what diameter PVC pipe is available. I think 65mm, aiming at a length no larger than 150mm and possibly less than 100mm.

Joe R.
 
Hi Joe,

I just read your write up on your website about the Elsinores. It states that the efficency is 94db and impeadance is 8 ohm.
In your graphs the efficency seems to hover around 90db and is pretty flat. Also I cant see how they can be 8 ohm as the mid bass are 2 8 ohms in series and the bass the same? Graphs I guess tend to indicate 12 ohm.
Am I correct here or have I missed something.
Im reasonalby interested in these but Im not sure they meet my requirements. I want efficency in the mid 90s for and lots of slam, dymanics, speed etc. Think these fit the bill, or are there more suitable projects I should look at? Im sure these are great speakers but its so hard to know without listening.
 
Luke said:
Hi Joe,

I just read your write up on your website about the Elsinores. It states that the efficency is 94db and impeadance is 8 ohm.
In your graphs the efficency seems to hover around 90db and is pretty flat. Also I cant see how they can be 8 ohm...

Im reasonalby interested... I want efficency in the mid 90s for and lots of slam, dymanics, speed etc...

I don't recall claiming 94dB?

I don't have a quick answer, but here goes... [up he goes on his soap box]

This whole topic is a confusing concept and it is not hard to see why. You like high efficiency, yet most speakers are rated @ x dB sensitivity - they are not the same. High sensitivity only makes sense if the speaker is high impedance, only then do you have high efficiency. Sensitivy is SPL @ 1M with 2.83V RMS input and is irrespective to impedance, which could be 1 Ohm (some ribbons are) and hence low efficiency. Efficiency is based on current times volts for a specific SPL and is rather more difficult to quantify.

I am not trying to sound preachy, but what you, and I too, want is high sensitivity while maintaining as high as possible impedance. Also getting the load as resistive and even a little inductive also helps.

Go full range drivers and you will have it most of the time but with other limitations. That and other systems also trade genuine bass extension for higher sensitivity, which means that if you want, and I think you do, some real slam and dynamics you have to compromise a little. But where is the compromise?

Let's look at the individual Elsinore drivers, they are 89dB/1M SPL @ 8 Ohm. Put them in series parallel you get 95dB/1M SPL @ 8 Ohm. Leave there and you would have the goal you seek, right? But not so quick. If you want the bass to be there in a balanced degree you have to compensate for diffraction loss which is a theroretical 6dB. That will get you back to 89dB. (But a full compensation of the single 89dB driver would give you 84dB, no so hot, so it is still way better.)

Do you see the trend? And also the designer has an option to what degree to correct for diffraction loss? Only he can decide but whatever he does he will need to determine it to be between 0dB and the full 6dB compensation. I would think most aim for 4dB approx but there are those quite willing to sacrifice the full 6dB and perhaps even over-cook it and get even more bottom end, but sacrificing sens.

Indeed even with Lowthers @ around 96-97dB sens with full correction is only 90-91dB (rear horns don't boost sens). Bass below 100-200 Hertz have been truncated if you only aim for max sensitivity.

Now let's look at what the Elsinores do. They series the upper two drivers to give nominally 16 Ohm and 89dB. 89dB may not seem so good, but this does not need compensation and has high efficiency. Now look at the bottom two drivers and things gets interesting. we've halved the current while maintaing same sens. They too are in series and 16 Ohm, but at low frequencies are in series parallel with upper two drivers, so nominally 8 Ohm. But the roll-off is so gradual that it does two things, it fills in the diffraction loss at low frequencies and also, because it's so gradual rolling off (lese than 6dB/Octave), also manages to boost the overall sensitivity by several dB up to above 1500 Hertz and then the gradual take-up by the tweeter maintains it above that.

In practice when comparing with other speakers rated at 88-89dB, the Elsinores sounds noticeably louder. This was noticed when changing speakers without disturbing the volume control setting. At one time this was done with a known reviewer in the room and he actually was the one who turned down the volume control. I took the opportunity to ask how much he estimated he reduced the volume by and he said 4 or maybe 5dB. This certainly puts it above 90dB somewhat.

As for my graphs you comment on, well I am conservative when I set it up and you have just given me full justification for doing that. At least I make no false claims. :) But it is known that manufacturers claims are not always correct and also SPL is not the easiest thing to calibrate and measure.

Re the rated/nominal impedance of a loudspeaker, I follow the well estabished convention that the impedance must not drop below 0.707 times rated impedance. So an 8 Ohm should not drop below 5.65 Ohm at any frequency and 4 Ohm not below 2.85 Ohm etc. A 16 Ohm not go below 11.3 Ohm.

Re coming back to full range drivers, I have a pair of Visaton B200 and the manufacturer recommends a contour filter that compensates for diffraction loss. It ends up about the same as the Elsinores even though rated 96dB. Some do similarly with other full-range drivers and some like it, others maybe not. But it is tweakable to suit taste. So full range high sensitivity do not offer as much hope if you want a bottom end to match. So what do they do (and I heard a major Lowther horns on the weekend doing this) add a sub-woofer that can be very hit and miss.

Finally, and this sorta sounds a bit black magic, but the Elsinores have another quality that numbers just don't even hint at. Their ability to provide that certain jump factor and incredibly low dynamic compression, they just don't need a big muscle amp to get going and produce a really powerful sound. Do they sound louder than they measure? I can say that we did a controlled comparison between two CD players and matched their outputs to 0.07dB and yet the player that had superior dynamics gave the impression of being several dB louder. It was actually 0.07dB lower. This was heard by everybody in the room. I am quite sure the same can be applied to speakers etc.

I use the Elsinores currently with a 20W amplifier and have also used them together in a large carpeted community hall where Audiophile Society of NSW meet (the amp won the major blind shoot-out earlier this year) and it went quite loud indeed. It has also been tried with 6 Watt amplifiers and often surprised listeners that a speaker of this size is so capable with such a small amp. But since I like "Child in Time" by Deep Purple at fairly high SPL, then 20 Watts is quite nice in a thunderous way.

Now what about a 2 metre high Line Source? That could give you 101dB before compensation using 12 of the same drivers. It would be a 6 Ohm system. One day...

Joe R.
 
Hi Joe, thanks for your thorough explanation. I understand what you say about specs being misleading, and I see how difficult it can be providing specs of loudspeaker.
I need to read your post again several times to fully digest it all. But I do get the gist of what your saying.
 
Luke said:

I need to read your post again several times to fully digest it all. But I do get the gist of what your saying.

Thanks, I think reading it a couple of times and it's not too hard to get the gist of it. I do hope that you give serious consideration to building the Elsinores. I don't think you will be disappointed as this is a speaker with a big sound and yet surprisingly modest power requirements.

Joe R.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Joe Rasmussen said:


I hear what you say, but maybe I'm into "soft sound"? Mind you, never been accused of that. As for series parallel, that is obligatory in line sources and I am of fan of that. One day I want to do a two metre high line source, great potential genuine efficiency and equally potentially low distortion.

But will most likely let people opt for both variations in the Hamlet, it all depends how the results stack up.

Joe r.


I got softer sound in that experiment, I don't say its the norm.