Synergy attempt without compression driver

onni

Member
2013-03-29 9:24 pm
I've been intrigued by the Synergy concept for some years but have some resistance against using compression drivers. The very high sensitivity makes them noisy with my amplifiers. But primarily I've gotten really nice results with the ScanSpeak R2604 in a waveguide which makes me want to try it.

A first sketch:
530212d1455107911-synergy-attempt-without-compression-driver-r2604-synergy.png

Inside. The woofers are Faital Pro 8FE200.

530213d1455107911-synergy-attempt-without-compression-driver-sketch2.png

Outside view.

The size is about 50 x 35 x 30 cm (W x H x D, 20" x 13" x 12"), i.e. a quite compact synergy.

These are the horisontal polars of the R2604 in an early waveguide (210x130x85 mm):
[IMGDEAD]https://i.imgur.com/NPdV543.jpg[/IMGDEAD]

KpluznBl.jpg

Measured waveguide (the brown one).

530214d1455108141-synergy-attempt-without-compression-driver-r2604-wg-raw.jpg

On axis, with phase.

530215d1455108141-synergy-attempt-without-compression-driver-r2604-wg-lt-peq.jpg

On axis with PEQ and LT.

530219d1455108855-synergy-attempt-without-compression-driver-r2604-wg-lt-peq-impulse.jpg

On axis with PEQ and LT.

530216d1455108141-synergy-attempt-without-compression-driver-r2604-wg-dist-85db.jpg

Distortion at 85 dB with 1.2 kHz bessel (for Harsch XO).

530217d1455108141-synergy-attempt-without-compression-driver-r2604-wg-dist-94db.jpg

Distortion at 94 dB with 1.2 kHz bessel (for Harsch XO).The R2604 should be better loaded in the bottom end with a 50 cm wide waveguide (synergy horn) and that should reduce the distortion to enable a 1.2 kHz crossover and 95 dB maximum level.

The plan is to 3D-print the first part of the horn (throat) and then use wood for the last part.

I'm quite new to Hornresp, but have tried a little:
530224d1455109880-synergy-attempt-without-compression-driver-hornresp-input.png


530223d1455109775-synergy-attempt-without-compression-driver-hornresp-power.png


Throat velocity (separate model) is 9 m/s at 2.83 V and diaphragm displacement is 1 mm. I've used 2 x 1" holes for each woofer in 19 mm MDF using minimal volume filler (measured the Vtc for the woofers at 350 ml).

The plan is also to have a leaky supercardioid setup for the woofers to stretch the directivity of the speaker down to below 200 Hz.

I'll post 3D-models later to show what I plan to print :)

/Anton
 

Attachments

  • r2604 synergy.png
    r2604 synergy.png
    6.3 KB · Views: 1,968
  • sketch2.png
    sketch2.png
    4.8 KB · Views: 1,924
  • R2604 WG dist 94dB.jpg
    R2604 WG dist 94dB.jpg
    73.8 KB · Views: 1,867
  • R2604 WG dist 85dB.jpg
    R2604 WG dist 85dB.jpg
    72.6 KB · Views: 1,860
  • R2604 WG LT+PEQ.jpg
    R2604 WG LT+PEQ.jpg
    53.1 KB · Views: 1,868
  • R2604 WG raw.jpg
    R2604 WG raw.jpg
    52.1 KB · Views: 1,918
  • R2604 WG LT+PEQ impulse.jpg
    R2604 WG LT+PEQ impulse.jpg
    48.1 KB · Views: 2,030
  • hornresp power.png
    hornresp power.png
    17.3 KB · Views: 1,862
  • hornresp input.png
    hornresp input.png
    20.6 KB · Views: 1,865
Last edited:

onni

Member
2013-03-29 9:24 pm
CSD

I forgot the CSD plot for the R2604 in the printed waveuide:
530232d1455111379-synergy-attempt-without-compression-driver-r2604-wg-csd-94db.jpg


All measurements look very good to me, except the rising distortion below 1 kHz which should be (at least partly) fixed by a larger waveguide.

/Anton
 

Attachments

  • R2604 WG CSD 94dB.jpg
    R2604 WG CSD 94dB.jpg
    82.5 KB · Views: 1,816
Nice. :)

How are you going to get the woofer bandpass to go past 1k (preferably 2k) in order to have a good overlap with 1k XO for tweeter? The steep cliff at 1k will make it impossible to do, say a Harsch XO at 1k. Or, with a larger WG, you expect the ring radiator dome to go well below 1k?

Will you print the whole thing WG and all? Commercial printer or in pieces then glue together?
 

nc535

Member
2011-04-08 5:12 pm
NC
I struggled mightily to make 8FE200s work in a 2-way synergy. I could get response out to 1.3 Khz (using a volume plug) but no matter what I did, I had a null around 900 hz. The configuration that minimized this null was with no enclosure around the back of the 8Fe200s and with those drivers draped with absorption. That was how I prototyped it but my celebration was premature.

Like you, I was trying for a small, tight box for each 8FE200, just big enough to fit the magnet and with the back wall of the box parallel to the horn wall. My conclusion was that reflections from the back wall were arriving 180 degrees out of phase at the port through the horn wall and cancelling. Others have shown successful use of the 8FE200 in larger enclosures. My advice is make sure you have enough space behind the drivers for enough absorption to absorb the back wave at the resonant distance.
 

onni

Member
2013-03-29 9:24 pm
Nice. :)

How are you going to get the woofer bandpass to go past 1k (preferably 2k) in order to have a good overlap with 1k XO for tweeter? The steep cliff at 1k will make it impossible to do, say a Harsch XO at 1k. Or, with a larger WG, you expect the ring radiator dome to go well below 1k?

Will you print the whole thing WG and all? Commercial printer or in pieces then glue together?
Good point! I'll probably need to print volume fillers as well. One upside of using the R2604 is that the membrane is in the horn throat and not deep inside a driver, so the notch from the reflection should be high in frequency.

I'll print the throat, maybe first 7-8 cm.

/Anton
 

koja

Member
2009-10-04 12:10 pm
Ontario
very interesting work. I also share interest in extrapolating on Synergy approach and was thinking of inserting a tweeter in the back of this horn which Hoerst did with Visaton B200:
Horn-Sat
not quite sure how something like that could be done but the thought crossed my mind when I was messing with playing B200 through cardboard boxes (when breaking it in for use in my OB), and that driver does classical music and deeper voices like you would not believe it. but off a flat baffle and run FR it is just awful.

if there was a way geomerically to combine two drivers in a horn where the back radiaton is adding to the front outputs for increased sensitivity, that could also be interesting for CD use.
 
Last edited:

nc535

Member
2011-04-08 5:12 pm
NC
re' just printing the throat

I've done a rectangular wooden conical horn synergy with a square throat from BWaslo's spreadsheet. I did the round to rectangular conversion with modelling clay like most others. Unsatisfied with this and inspired by XRK's thread I just 3D printed an insert for the horn that does that conversion much more smoothly than I did with clay. Now I've got to scrape off the clay and slide the adapter back into the throat in its place and measure to see what improvement I got.

I think this insert approach is easier to execute than joining a wood section to a plastic printed section, provided one is comfortable working with wood in the first place. The hard part, for me at least, was doing the CAD for the adapter and prior to that, choosing an appropriate profile. I did this in Sketchup and generated the shape heuristically rather than mathematically. It looks a little rougher than it actually is or I would like but is at least a proof of concept. I just exported STL from the Sketchup drawing and sent it and $30. off to "Sculpteo". I felt very lucky when I got the piece back a few days later.

Some pictures below. The adapter is roughly 2" by 3.7" by 1.5" for a 90x45 horn. The third image is the google file itself showing the adapter in the horn. It ends about .5" short of the mid ports with final thickness of perhaps 1/32".
 

Attachments

  • RO Throat Transition.png
    RO Throat Transition.png
    124.1 KB · Views: 337
  • back of adapter.jpg
    back of adapter.jpg
    63.6 KB · Views: 327
  • front of adapter.jpg
    front of adapter.jpg
    70.7 KB · Views: 301

onni

Member
2013-03-29 9:24 pm
I struggled mightily to make 8FE200s work in a 2-way synergy. I could get response out to 1.3 Khz (using a volume plug) but no matter what I did, I had a null around 900 hz. The configuration that minimized this null was with no enclosure around the back of the 8Fe200s and with those drivers draped with absorption. That was how I prototyped it but my celebration was premature.

Like you, I was trying for a small, tight box for each 8FE200, just big enough to fit the magnet and with the back wall of the box parallel to the horn wall. My conclusion was that reflections from the back wall were arriving 180 degrees out of phase at the port through the horn wall and cancelling. Others have shown successful use of the 8FE200 in larger enclosures. My advice is make sure you have enough space behind the drivers for enough absorption to absorb the back wave at the resonant distance.
Alright, I'm going for a leaky cabinet and maybe there is a point in putting the vents directly behind the drivers.


re' just printing the throat

I've done a rectangular wooden conical horn synergy with a square throat from BWaslo's spreadsheet. I did the round to rectangular conversion with modelling clay like most others. Unsatisfied with this and inspired by XRK's thread I just 3D printed an insert for the horn that does that conversion much more smoothly than I did with clay. Now I've got to scrape off the clay and slide the adapter back into the throat in its place and measure to see what improvement I got.

I think this insert approach is easier to execute than joining a wood section to a plastic printed section, provided one is comfortable working with wood in the first place. The hard part, for me at least, was doing the CAD for the adapter and prior to that, choosing an appropriate profile. I did this in Sketchup and generated the shape heuristically rather than mathematically. It looks a little rougher than it actually is or I would like but is at least a proof of concept. I just exported STL from the Sketchup drawing and sent it and $30. off to "Sculpteo". I felt very lucky when I got the piece back a few days later.

Some pictures below. The adapter is roughly 2" by 3.7" by 1.5" for a 90x45 horn. The third image is the google file itself showing the adapter in the horn. It ends about .5" short of the mid ports with final thickness of perhaps 1/32".
Sure looks easier! My approach will however be very sensitive towards shape of the throat as the wave is in no way flat when entering the horn compared to a (good) compression driver. The R2604 is however quite light, which should make the connecting easier.

This is from a simulation I've done of a WG that should perform better than the one I've measured:
505892d1443127315-3d-printing-se1.85-20k.png


Simulated result:
505880d1443124896-3d-printing-se1.85-hor.png

Constant directivity from 2 kHz to 10 kHz. The low end should improve with a larger WG.

/Anton
 

onni

Member
2013-03-29 9:24 pm
New hornresp attempt

I tried to get the response for the 8FE200 up to 2 kHz and arrived here:
530295d1455132421-synergy-attempt-without-compression-driver-hornresp-input2.png


530297d1455132421-synergy-attempt-without-compression-driver-hornresp-power2.png


530302d1455132550-synergy-attempt-without-compression-driver-particle-velocity2.png


530296d1455132421-synergy-attempt-without-compression-driver-group-delay2.png


I've assumed that the throat adapter only needs to be 5 cm long. I also reduced the holes to 21 mm diameter (from 25.4 mm). Looks like this:
530303d1455133810-synergy-attempt-without-compression-driver-21mm-holes-58mm.png

Probably even better to make them oblong from diffraction point of view, but that reduces the possible length of the throat adaptor.

Another possible look:
530305d1455134466-synergy-attempt-without-compression-driver-sketch3.png


/Anton
 

Attachments

  • hornresp input2.png
    hornresp input2.png
    19.6 KB · Views: 1,818
  • group delay2.png
    group delay2.png
    16.5 KB · Views: 1,713
  • hornresp power2.png
    hornresp power2.png
    16.8 KB · Views: 1,714
  • particle velocity2.png
    particle velocity2.png
    12.7 KB · Views: 1,710
  • 21mm holes at 58mm.png
    21mm holes at 58mm.png
    3 KB · Views: 1,709
  • sketch3.png
    sketch3.png
    5.8 KB · Views: 1,715

nc535

Member
2011-04-08 5:12 pm
NC
that looks very doable. You wouldn't get the same results using a compression driver, because, as you said, you'd have to add perhaps another 3 cm to the reflection distance. The L12 on my 8FE200 proto was 2.25" which suggest that your 5.8 cm is practical.

I noticed in doing these horn response sims that if, for example, you changed both your "Ap1" and "Lp" parameters to half the values in your posted model, you would get the same results - both for the frequency response curve and the offset throat particle velocity. If you can thin out the horn wall at the port location (e.g. frustrumize), then you can use a smaller hole and thus better polars If one takes this to the extreme, you end up with an extremely thin horn wall and an extremely small hole. What is the catch?
 
If you can thin out the horn wall at the port location (e.g. frustrumize), then you can use a smaller hole and thus better polars If one takes this to the extreme, you end up with an extremely thin horn wall and an extremely small hole. What is the catch?
An extremely thin horn wall would imply a larger VTC (volume of throat chamber). The VTC needs to be filled to reduce it's volume to raise the mid driver's upper cut off frequency high enough for the tweeter to not run out of excursion in it's lower range.
 
Last edited:

onni

Member
2013-03-29 9:24 pm
that looks very doable. You wouldn't get the same results using a compression driver, because, as you said, you'd have to add perhaps another 3 cm to the reflection distance. The L12 on my 8FE200 proto was 2.25" which suggest that your 5.8 cm is practical.

I noticed in doing these horn response sims that if, for example, you changed both your "Ap1" and "Lp" parameters to half the values in your posted model, you would get the same results - both for the frequency response curve and the offset throat particle velocity. If you can thin out the horn wall at the port location (e.g. frustrumize), then you can use a smaller hole and thus better polars If one takes this to the extreme, you end up with an extremely thin horn wall and an extremely small hole. What is the catch?
Interesting find indeed! I had to do a control myself cause it seems weird, a smaller hole should result in higher particle velocity.

/Anton
 

onni

Member
2013-03-29 9:24 pm
One thing I wanted to ask given your comment about the shape of the wave entering the horn from the ring radiator tweeter - are you able to model that so as to accurately predict the polar response of the horn driven by that tweeter?
Well, I'm trying as apparent in the 3D-printing thread (starting at post 174 here).

Here is comparison of measured and simulated horisontal polar response for the brown WG:
505819d1443097834-3d-printing-measured-vs-simulated-brown-wg.png

I discuss sources for discrepancies in the thread.

Hopefully I can make a model of the synergy later.

/Anton
 

onni

Member
2013-03-29 9:24 pm
Sketchup model

Played around a little in Sketchup. Outer dimensions are 50 x 28.2 x 30 cm and the waveguide mouth is thereby 40 x 25 cm.

530413d1455191145-synergy-attempt-without-compression-driver-50x28.2x30-rounded-wood.jpg


530414d1455191145-synergy-attempt-without-compression-driver-50x28.2x30-rounded-white.jpg


I'll make a model of my media console to see how it looks later.

/Anton
 

Attachments

  • 50x28.2x30 rounded wood.jpg
    50x28.2x30 rounded wood.jpg
    104.1 KB · Views: 951
  • 50x28.2x30 rounded white.jpg
    50x28.2x30 rounded white.jpg
    98.4 KB · Views: 953
Nice looking design Onni. It's got a certain 1970's aesthetic and I mean that in a good sense. Very clean lines. A lot more WAF friendly than my rectangular box with a black baffle. I hope you can stuff all that into that little box. Those 8FE200's have a much larger bezel diameter than the smaller DC200's I used. I know I sure had a tough time fitting mine into a 46cm tall x 28cm wide x 33cm deep. This was using the Faital Pro LTH142 waveguide.

http://www.diyaudio.com/forums/multi-way/285030-bookshelf-multi-way-point-source-horn.html

529809d1454908866-bookshelf-multi-way-point-source-horn-baffle-black-2s.png
 

onni

Member
2013-03-29 9:24 pm
More sketchup

Here's how it would look on my media console (290 x 40 x 50 cm, W x H x D):
530519d1455222208-synergy-attempt-without-compression-driver-50x28.2x30-rounded-white-console.jpg


Showed this to my SO and she was surprisingly positive! I think I could make these in a way that she would let them in to our living room :)

I probably need to come up with some kind of feet to angle the mouth upward (towards ear-height while seated).

/Anton
 

Attachments

  • 50x28.2x30 rounded white on console.jpg
    50x28.2x30 rounded white on console.jpg
    163.5 KB · Views: 1,343