C/E/X PA Flat to 30 (FT30) PA TH Awesomeness

here is a mod of a model i posted earlier.
as djim sugested ,i put the magnet outside the chamber for better heat dissapetion.
the box is 570lts and is 1+ 2/3 of 4'x8'sheet ,so thats 60 kg (whitout bracing)
semi%20sym%20570%20ltr.jpg
 
Interesting results Epa :up:

Your design is based on (1/3wavelength) Fb of ???? (32Hz?)
While Xoc1-TH18 is based on a Fb of 3.045m = 37.6Hz

Your design has a significant increase in low end extension at cost of a relative high Group Delay at the highest excursion point and at costs of bandpass in the higher part of the frequencies.

The question is will it perform as your prediction...

Epa%252527s%252520TH_vs_Xoc1-TH18_01.jpg



Epa%252527s%252520TH_vs_Xoc1-TH18_02.jpg
 
Last edited:
wel the goal was ,low towards 30 hz with resenable output,and not to heavy.
the group delay is not so good ,needs to be crossed 80~90 hz.
i can't see a reason why it shoulden't work(mayby relative short path towards the mouth?)
what about un-even loading would that be a problem?
it begins and ends as a single fold would,just some bends in the middle sector😀
btw i recalculated hr inputs more precise above.
i think it needs reflectors @ bottem front.
it needs some tweaking to get concistancy

the 18lw 2400 preforms equaly in this design.
 
Last edited:
...what about un-even loading would that be a problem?
it begins and ends as a single fold would,just some bends in the middle sector😀
btw i recalculated hr inputs more precise above...

Hi epa,

I think your design is better than many other multi-folded seen here:
IMO,multi folds in a PA speaker is a SQ disaster when compression levels are reached.

it needs some tweaking to get concistancy

?, See a simulation that reflects your HR input data:

b🙂
 

Attachments

  • epa-18NLW9000-PA-TH.JPG
    epa-18NLW9000-PA-TH.JPG
    639.8 KB · Views: 560
wel the goal was ,low towards 30 hz with resenable output,and not to heavy.
Mission accomplished!

the group delay is not so good ,needs to be crossed 80~90 hz.
Indeed, the 35Hz peak is a little high (+30ms) and the
peak at 150Hz, although very small band so that might be much better in reality (less high).

i can't see a reason why it shoulden't work(mayby relative short path towards the mouth?)
Indeed, if the driver is going to 'see' the path like a more traditional TH concept the sound will start to roll of 10Hz higher.

what about un-even loading would that be a problem?
For cheap drivers it can be a problem but in an age of double spider techniques, high cone stability and exotic materials I don't think that will be a problem.

it begins and ends as a single fold would,just some bends in the middle sector😀 btw i recalculated hr inputs more precise above.
That's the reason I had to try another driver and sure the outcomes look very nice :up:

i think it needs reflectors @ bottem front. it needs some tweaking to get concistancy
In the first 1/6 of the total path I should use corner correction in every corner.[/QUOTE]

the 18lw 2400 preforms equaly in this design.
It sure does!
 
B&C's split-winding technology

After contacting B&C about their split-winding technology they send me a picture. On the left side a voice-coil with Split-Winding technique and on the right side a voice-coil with 'normal' winding. On my question, if or how the gap in the split-winding influences the Hvc figure, B&C answered;

"The height of the voice coil is the distance from the first to the last turn of wire. Even with the split coil technique, this distance is equal for the inside (full) layer and the outside (split) layer."



B%252526C_Split-wire_vs_normal.jpg
 

Attachments

Last edited:
The PDF explains a little, but in short: higher excursion/lower THD.


Edit after your edit 😉
Take for instance the 18SW115:
Xmax calculated [(Hvc-Hg):2] = 10mm
Xmax 10%THD = 14mm
Xvar 50%BL or 50% Cms = 16mm

If you would take out the 'split' it would be guestimated:
Xmax calculated = 9mm
Xmax 10%THD = 12,8mm
Xvar 50%BL or 50%Cms = 14,5mm
 
Last edited:
Awesome! Thanks for that Djim.

I had wondered whether this was a split coil linear motor implementation and it appears that it is. Split coil is public domain with no enforced patent. It is sort of odd that they would utilize it here when the xmax is easily attainable with regular over or underhung. Split coil penalizes peak bl but provides a broad flat plateau. I had a hunch this was the case as soon as I saw the Klippel data from their 18sw115. I have never seen an arrangement such as theirs with an inside outside wind and a split coil on only the outer wind. Intriguing.
 
Over-hung -> higher thermal power compression figures or massive increase of VC weight
Under-hung -> lower Xmax or lower sensitivity and non-linearity’s rise more rapid as soon the coil starts leaving the gap.
Split-Wire -> low pwr comp figures, just a fraction less sensitive in favour of high Xmax (10%THD) and high Xvar (50%BL or 50%Cms).