My Second Tannoy Autograph enclosures

[maxlorenz]

Well, here I must seek help from my friends...Scott, Qi...
As I declared I want to study the possibility of doing a "square" or "trapezoidal" Autograph to better fit modern living rooms

First of all, I want to imitate as close as possible the exponential (?) growth of the rear horn, as this is the main part that will suffer modification. You may use your software to help me

Here is a drawing (not perfect-not to scale-with almost every dimension needed, in cm and '') of the "unfolded" Full-sized Autograph rear-horn. There might be some few cms of error here and there. The upper view was difficult because of the folding (knees) where the width is difficult to calculate, as it expands or contracts.

Maybe "we" can draw some frequency response curves.

For the 2:3 replica, each lenght must be 2:3 of the drawing's lenghts.

Side view:

Upper view:

I hope you understand them

After "we" do the 3D schematic and calculate volumes and paterns, we probably could derive a comparable "trapezoidal Autograph" with longer, bigger folded rear horn

Better than Westminster, wright?!

Cheers,
M

 

[Scottmoose]

No reason why not. GM will probably be interested in helping out too -Greg, you out there? Hope you're connection's recovering?

I'll have to get my thinking cap on. IMO the Tannoy horns, though they sound magnificent, just aren't big enough... A DIY Westmister Royal HE beater should be possible.

BTW, one view I'm increasingly finding myself agreeing with, is that in horns with parallel walls, flare-rate is less important than some other factors such as length and mouth-size. If you can stick tightly to a good hyperbolic expansion (which I believe to be the best compromise) that's a good thing, but I don't think it's as significant in this particular case (parallel walls) as in others.

Scott

 

[AndrewT]

Hi,
I wish I had heard the Tannoy horns, so this is guesswork.

I also feel that the Autograph and Westminster have a mouth that is truncated too early.

The flare rate in the Westminster should allow bass performance down to about 25 to 30Hz.
The Mouth areas are respectively 0.6sqm and 0.75sqm. These start to ripple at about 43Hz for corner loading the Autograph and 60Hz for floor and wall loading the Westminster.

And yet Tannoy claim the Westminster is only 6db down at 18Hz and those that have heard it say phenominal/SLAM!

But could that 18Hz response be at the top of one of the ripple peaks.

I have a sneaky suspicion that both speakers could get more out of the flare rate if the horn had been longer, resulting in a bigger mouth. Twice as big would be an interesting target.

 

[Scottmoose]

It's relying on QW action down there I suspect. Fair enough I suppose, but I'd still like a bigger mouth / longer line.

 

[Diogenes]

I'm definitely interested, though I haven't the technical knowledge in horns to help out. I've wanted to eventually build a Westminster clone for my HPD385A drivers (nearly the same drivers as in the Westminster Royal) when I had the room and money, but an improved model would be even better. I'll be following this thread with interest.

To help out, the manual for the Westminster Royal can be viewed here. Page 18 shows the measured anechoic frequency response, which supports the talk of a too small horn in previous posts. Page 15 shows some of how the fold works.

Hope this helps.

- JP

 

[Scottmoose]

Indeed it does. Confirming, as we all suspected, that these are only horns in the nominal sense of the word, but are closer to QW action. Not that there's anything wrong with that, they're clearly beautifully engineered. But they equally certainly rely heavily on room-gain to boost extension as the mouth size just isn't big enough, nor are they as long as one might think. A rectangular GRF in effect. DIY-ing it should be possible enough. An 'improved' version is also probably possible, but will need careful thought.

 

[maxlorenz]

Hi
Thanks to all for your kind help and interest.

DIY-ing it should be possible enough. An 'improved' version is also probably possible, but will need careful thought.

That is the idea! We can do better. In my case, no hurry and eXcellent WAF, so BIG enclosure is no problem. The throat and first part of the rear horn is better maintained to simplify the work and only mod, enlarge...etc, the foldings. Safe on the 2:3 model wich can use a taller horn first section: longer horn and higher center ("tweeter") point.

Diogenes, thanks for the link!
The enclosure for the fullsized Autograph costed me US$400 (including woodworkers fees)
BTW, I will left out the 10 cm base in both models, fullsized and 2:3; as I understand it will better loaded to floor (?).

Hi AndrewT

The flare rate in the Westminster should allow bass performance down to about 25 to 30Hz. The Mouth areas are respectively 0.6sqm and 0.75sqm. These start to ripple at about 43Hz for corner loading the Autograph and 60Hz for floor and wall loading the Westminster.

Yes, it has slam. Maybe it doesn't reach 20Hz, I don't know. With the diagram provided and with my tranducers specs, could a response curve be made? (Beyma KX15 coaxial)
Can be the corner loading be ploted???

Anyway this is "Tannoy bass": ample; expanding, fast and articulated. The high sensitivity transducer + the high sens (efficience) enclosure allows for easy high SPL listening, without distortion, but it is a little scarry sometimes when fortissimo approaches
Corner loading really makes magic here as the whole response is felt as round and coherent. It will be hard to compensate for it with the present project.

The little 2:3 replica is very satisfying in the bass. Actually one does not feel that something is lacking because all the bass notes it reproduces are felt like I described above, with natural bass armonics and natural attacks and decays (piano is great), wich is one of the reasons why I am sold to this enclosures (I sent my Tannoy Definition500 for sale )
I simply can't go back to boring, square, bass reflex boxes

I think no one made my recommended experiment: make a 1:4 replica with 5mm MDF and a tinny "full-range". It only takes one week but it allows for perceiving what this enclosure is capable of. I'm not joking
Hey! maybe a 1:4 model will be OK for experimenting this new one! Thanks...Besides that, a complete speaker pair with Audio Nirvana super10 should cost under US$400...and Mamma should be less worried/mad for the big box

Bye,
M

PS: does this project deserve it's own thread?

 

[Scottmoose]

Initial thoughts: Keep the chamber and throat the same volume. Boost height back to the full Westmister Royal HE for a good 15in co-ax. Be worth seeking out a pair of genuine Tannoys for this I reckon. Slow down the flare rate and increase the length by adding two more back-front folds. Make sure the final section has a rapid final flare though which will help smooth peaking at Fc.

 

[AndrewT]

Hi,

Make sure the final section has a rapid final flare though which will help smooth peaking at Fc.

could you explain this in more detail? new flare rate, when to start, how to assess the Fc(whatever that is), etc.

 

[Scottmoose]

Fc = lower frequency cut-off. Little trick I learned from GM. In certain designs, the application of a rapid final flare / expansion to the horn causes a sudden decrease in pressure. Ground-effect cars, like, say, the Lotus 79 Formula One racer from the late 1970s do the same thing with their shaped undersides. This in turn will reduce or eliminate any potential peak at cut-off. I'm just idly thinking at this point in time, nothing concrete in mind yet. Roughly though, an additional front-back-front fold spliced into each side in front of the existing two. It'd need the angles all changing, as the expansion in these new paths would be very slow, but it could be worth doing for the extra length it should afford. Still doesn't do much about the relatively small mouth size, but the Fostex cabinets deliberately under-size their mouths and go for a longer-then regular length in compensation. Works well enough for them. The 208 box can get down into the low 20s, and that's a lot smaller than whet we're dealing with here.

 

[maxlorenz]

Hi Scott,
Thanks for the tips.
For the 15" tranducers, as I am stuborn I will try those "pro" 15" P-Audio BM-15CX38, instead of passing through the various steps of a used Tannoy purchase
Activelly biamped as always.

http://www.paacoustic.com/Product_Detail.asp?CatalogID=&ProductID=BM-15CX38

http://www.paacoustic.com/pdf/BM-SERIES(COAXIAL)/BM18-15CX38.pdf

What do you think?

What (free) software must I get to play with my different ideas???
(I am a nullity regarding software; only intuition is my guide )

Regards,
M

 

[AndrewT]

Hi,
if the extra front to back fold were incorporated then it would have to stop about 9 or so inches from the front to allow the wave to travel around the corner. Then it would expand to about 13 or 14inches wide for the extra horn length instead of the 10inches at present. This would make the speaker an extra 28inches wider. Is that managable? and be rear exit!
That's just one extra fold. If you add on a second extra fold the front to back becomes silly because the gap for the wave approaches the depth of the cabinet.

Now if you were to make the cabinet much deeper, then redesign all the folding for minimum wasted internal space and sell it back to Tannoy.

 

[Scottmoose]

I said front-back-front. You need to add another path to it. Did you mention something about them being wide?

Wouldn't automatically need that kind of depth to the gap -depends on the flare-rate. And potentially adfding more folds could help -LF isn't that bothered about corners, but it would help attenuate any stray HF leakage. And the large filter chamber should help clobber any potential reflections back through the cone.

A thought. There's a large area of dead-space in the first front-to rear path fo the existing cabinet. We could slow the flare-rate dramatically, and in that space get a triple front-rear-front rear fold, stacked vertically. Shift the place which defines the final expansion from back-front to the bottom from the top, and presto: a good few more inches of line length. No idea if it'd work, but I can't see any genuine reason why it shouldn't. I might get chance to run some form of sim after Wednesday -I'm a bit snookered until then. The front horn might need changing as well.

The PA drivers seem OK. Not as good as Tannoy units though.

 

[maxlorenz]

Well, here we go...a first drawing of a monster modded Autograph, based on what I translated from what you said and what the Good Lord led me understand

Here I designed a prolongation of the central rear part of horn. Three times the lenght, so far ...and twice the depth of the box. Maybe good for the 2:3 replica only. Of course the angles/flare rate must be wrong.

Next...

 

[maxlorenz]

Here a view of what could be the base of the former version. I added colour fro the bored out there

Only one side if the lateral foldings is shown. The central rear horn is in colour.

Next...

 

[maxlorenz]

Now, the complete view. Again, flare rate could (is) wrong, but you see that the total height escapes the paper...
That is what I understood from Scott's guidelines
Please, bring your imagination...

Now, rear horn's lenght is more than double than the original. Mouth area can be easilly increased.
Well, I could do a tinny replica with the central core fixed and play around with different lateral folding flare rates

I choose that the (split) mouth be close to the front horn to maintain the original configuration. That makes narrow "corridors".

How do you like it?
(Maybe it's too much? )

Cheers,
M

 

[maxlorenz]

Now, my first thought was a simpler mod, from now on version2:

This one maintain box depth but increases height of vertical horn throat by any lenght that you prefer. Square base actually doubles lenght of lateral foldings and mouth area can be increased at will (almost)

You can compare the lenghts of the lateral foldings in modded and unmodded versions (only one side is modded).

Do you like V2? or it is only a little gain?

Regards,
M

 

[maxlorenz]

Hi all,
I will build a 1:3 replica of the "square" Autograph, V2.
Just to see if the bass integrates well with the rest of the spectrum.
Here are some picks:

This is the base of the speaker. At the end it resulted very similar to the former quick drawing.

To plan the rear folded horn I designed an unfolded horn on paper and then I folded it to accomodate the dimensions of the enclosure I hope it is fine this way...

Next I will post a diagram of the lateral view. I made the height of the vertical horn 20% taller. Based on my calculations this 1:3 speaker will have 1m long rear horn with 1260cm2 mouth area.
If we built a full-sized modded Autograph the rear horn would be 3m long and mouth area would be 11340cm2 but tweeter would be a little high.
Maybe I will make a 2:3 replica if I feel that my V1 (folded vertical horn) is too complicated.

Bye,
M

 

[AndrewT]

Hi Max,
the one third replica with 1260sqcm mouth and corner loading will cut off at 95Hz or so. Does the flare rate match that cut off frequency?

If it is only floor loaded then the cut off rises to about 190Hz.
this would make a good horn for a mid range driver, except you would probably want it off the floor to minimise reflections and keep it as near as possible to the treble unit.

 

[maxlorenz]

Hi AndrewT,
Well, this will only be a test speaker to evaluate the performance.
An enclosure for the Audio nirvana super10 (2:3) will be:
Horn lenght: 2m
Mouth area : 6160cm2.
Could you calculate the performance based on ANsuper10 specs?


I made a mistake: the mouth area for 1:3 speaker is 1540cm2; for full sized Autograph it is 13860cm2

This will be the side diagram:

This is the exact folded horn shape you saw above. Throat is 9.3cm (28cm in full sized Autograph) wide, then horn expands gradually to 22cm width.

Once I have settled the right enclosure dimensions it is easy to augment the size

Thanx,
M