Synergy horn for 135 dB+

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Synergy horn for >135 dB

Since I first saw (and heard) Synergy horns on YouTube I was intrigued. This coherent sound image was nothing I had ever seen before. And after building a 3-way Synergy with a half passive crossover I was sold. A small 2-way for my living room followed quickly and I wanted another pair that would easily endure a rock band in a small to medium sized space, replacing my 4 Bill Fitzmaurice DR200 cabs; less lugging, better quality, if possible even more balls (more dB’s and a bit better low end), but still a one man transport. That meant > 133 dB, extension < 120 Hz, preferably down to 90-100 Hz, to couple better with subs, and not too big or heavy. It should be pole mountable and (thus) be a one man lift.

With this in mind, more than a year ago I purchased the Beyma CP850ND compression driver, a 2” beast with a 4” voice coil, for a good 2nd hand price, with the intention to design a 2-way Synergy top. Why 2-way? I wanted a passive crossover and to design a 3-way crossover for a Synergy horn that also accounts for the phase curve is a bridge to far for me right now. A 2-way crossover I think I can manage. Than I ran into an offer on 12” B&C 12NDL76 lightweight neodymium woofers. Before I pulled the trigger on them, I quickly threw together some simulations and it looked like it would be a good candidate. So I bought 4 of them. Time passed by and life happened, as many of you undoubtedly know, and the project got put on hold.

On and off I had some time to think about the project and I started reading again for inspiration where I always start, right here on diyaudio :). That is when I read about Art Welter’s SynTripP(C). What a marvellous idea he had with that horn extender. Also, it gave me some confidence that a proper 2-way Synergy could be feasible. I was never 100% satisfied with my small 2-way living room Synergy horns…

Before I continue I’d like to thank Bill Waslo for his SynCalc spreadsheet, which is immensely helpful for designing and constructing a Synergy horn, Art Welter for sharing his SynTripP design, David McBean for offering his HornResp software, Tom Danley for his Synergy concept and everyone else here on diyaudio and on the world wide web for sharing his/her knowledge on this topic!

Let’s start with some simulations:





Note: 80V @ 4 ohms = 1600W, the program power of two 12NDL76s combined.
Note 2: Hornresp 39.40 is, as far as I know, not able to simulate this horn combined with a vented chamber for the woofers. I figured I could get about 6dB from adding vented ports tuned to 100 Hz and even out the bottom response.

Lately I have had some time to turn my thoughts and simulations into sawdust and this happened:

































Not much room left...



Then some measurements:

Tweeter:



Woofer:



Then a passive crossover design:







Which resulted in:



Note: this is @ 2m, my calibrated measurement ‘system’ is limited to 110 dB input. So add 6dB for a 2.83V/1m figure.
Note 2: the dip at 140 Hz is a room dip, as can be seen from the measurement below, which I did right in front of the horn mouth (“0 cm”):



*all measurements without any EQ, just the crossover (when applicable).

The strange phase jump around 760 Hz, that’s where the crossover frequency is, I find a bit odd. Look what happens when I put the mic about 10 cm off axis:



So what I want to do is try out REW’s Impulse Response Calculation. It’s set to “No timing reference”. I’d like to try the “Use loopback as timing reference” but I have to make a special cable for it…

What do you guys think so far?

Cheers!

Edit: Btw, the last three measurements are with four 65-mm bass reflex ports (not in photo's)
 
Last edited:
  • Like
Reactions: 1 user
Thijs666,

I like that you used better plywood this time and they look a much sturdier build than your other's :up:

How did you smooth the compression driver entrance it looks to be nicely done.

Did you reflex port it yet?

Edit: Just saw your edit, makes sense now!
 
Hi Thijs666:
HornResp is now up to version 40.30. You can simulate vented woofers as offset drivers on the horn without issue. You might not be able to do this when simulating it as a multiple entry horn but simulating woofers alone is not a problem

You should take a look and see how closely your CD and woofers are time aligned. The best indication is probably the step response of the full range measurement but you should also look at an overlay of separate driver measurment IRs.

Jack
 

Attachments

  • Attach_1.png
    Attach_1.png
    50.9 KB · Views: 499
Very nice! I was thinking about doing something almost exactly the same. I've got 4 12ndl76s and 2 Faital HF14ATs. My only concern is that the Faital CDs don't have enough low extension to meet the 12s at a reasonable point for a synergy design.

If you can't get a synergy to go high enough, you might follow my thread A 12 inch horn loaded MTM top: 12NDL76 and N314T-8. It will probably work really well with your drivers.
 
Last edited:
I did see that before. Nice project. My plan is to go big or go home. I know I'm going to need mids, and I'm mostly just whining about it because doing a 2 way and ditching an amp sounds nice but impractical to me.

I'm more than likely going to go "whole hog" on this one and do a proper 3 way with 4 4" Celestion mids to keep up with 2 12s.

My other idea that I'm still floating is doing Peter Morris' mid-top design with a 2way synergy center instead of the prescribed BMS coax. Just need some time to agonize over the choices haha.
 
HornResp is now up to version 40.30. You can simulate vented woofers as offset drivers on the horn without issue. You might not be able to do this when simulating it as a multiple entry horn but simulating woofers alone is not a problem

Damn, I must have been living under a rock or something. Totally missed that. Took a shot at it, but I’m not feeling confident I got it right…


(had to guess some T/S-parameters from the Beyma CP850Nd, so that's not accurate!)







In fact… I’m pretty sure I didn’t get it correct :s.

I would be happier with something like this: (I hope I summed the vented OD horn correctly)





That would mean I would have to enlarge the vent holes from 39.10 cm2 to 120 cm2. So far for my simulation with WinISD :smash:.

You should take a look and see how closely your CD and woofers are time aligned. The best indication is probably the step response of the full range measurement but you should also look at an overlay of separate driver measurment IRs.

Jack
I tried letting my DCX2496 calculate the time alignment and it suggested to delay the CD by about 4 ms. I don’t believe that would be correct. Below is the impulse response of the full system (from the second from last graph of my first post). Am I correct in reading a 20 microsecond delay on the woofer section?

 
Very nice! I was thinking about doing something almost exactly the same. I've got 4 12ndl76s and 2 Faital HF14ATs. My only concern is that the Faital CDs don't have enough low extension to meet the 12s at a reasonable point for a synergy design.

Thanks :cool:

That would be a concern indeed. I don't see a way to get those 12's playing any higher than you can see in the graphs above. They are literally up against the mounting plate of the CD. So you need a pretty beefy CD, hence I picked up the CP850 and I fear even this one isn't capable of keeping up. Distortion measurements will have to point that out.
 
120 sq cm is a pretty large hole but then they are large drivers. You might get a better answer if you simulate a shorter port length - equivalent to a thinner horn wall or frustrumizing the port. A shorter port should allow a smaller hole for the same velocity.

I should be the last one to give advice on time alignment since I struggled with it in my own design - but I might have learned something along the way. Using IIR DSP or passive filters, I believe you want the drivers aligned at the start of the IR. Using linear phase filters, you want the IR peaks aligned but if I recall correctly DCX does only IIRR filters. The IR peaks are easily determined. IR start can be fuzzy, especially if there is any pre-ringing, but you can judge start and then fine tune by trying to optimize the step - look for a sharp almost vertical initial rise followed by a ramp downwards to the right. What you have isn't bad; would look better zoomed in. Try small changes in delay and move in the direction that makes the initial rise of the step steeper.
 
120 sq cm is a pretty large hole but then they are large drivers. You might get a better answer if you simulate a shorter port length - equivalent to a thinner horn wall or frustrumizing the port. A shorter port should allow a smaller hole for the same velocity.

The measurements above are done with 4 vent holes of 65mm diameter.

3.25^2*pi = 33.18 cm2 ... *4=132.73 cm2; hmm, doh, I think I screwed up at least one hornresp sim by using only 33.10 for AP. But the sim for 33.18 measures up better to my measurements than the sim with 120 cm2:confused::confused:

On the other hand, looking at Art's SyntTripPs, they also have quite generous vent holes...

I should be the last one to give advice on time alignment since I struggled with it in my own design - but I might have learned something along the way. Using IIR DSP or passive filters, I believe you want the drivers aligned at the start of the IR. Using linear phase filters, you want the IR peaks aligned but if I recall correctly DCX does only IIRR filters. The IR peaks are easily determined. IR start can be fuzzy, especially if there is any pre-ringing, but you can judge start and then fine tune by trying to optimize the step - look for a sharp almost vertical initial rise followed by a ramp downwards to the right. What you have isn't bad; would look better zoomed in. Try small changes in delay and move in the direction that makes the initial rise of the step steeper.
Zoomed in? REW doesn't let me zoom in any further on the time scale :(. For the record: I tried to estimate the delay between CD and woofers using my DCX2496, but that's the only thing I did with the DCX. All measurements are done with the passive crossover. So no DSP involved! That's what I'm aiming for.
 
Zoomed in? REW doesn't let me zoom in any further on the time scale :(. For the record: I tried to estimate the delay between CD and woofers using my DCX2496, but that's the only thing I did with the DCX. All measurements are done with the passive crossover. So no DSP involved! That's what I'm aiming for.

Sure, REW will let you zoom in on an IR or IR overlay . Click on the "limits" button with the crossed double headed arrows and type in the time limits.
 
I admire anyone who can do a passive linear phase synergy xo even if I can't help thinking they have a masochistic streak :)

The art of passive XO synergy is choosing a port location that doesn't require any DSP or all pass filter delay for time alignment. So if this is just a proto build you should find the DSP delay you need for time alignment and then see if its within range of a port location change. I don't know if that can be done with a 12" woofer with the voice call so much further back from the horn wall than would be the case for the small mids usually used.
 
I admire anyone who can do a passive linear phase synergy xo even if I can't help thinking they have a masochistic streak :)

The art of passive XO synergy is choosing a port location that doesn't require any DSP or all pass filter delay for time alignment. So if this is just a proto build you should find the DSP delay you need for time alignment and then see if its within range of a port location change. I don't know if that can be done with a 12" woofer with the voice call so much further back from the horn wall than would be the case for the small mids usually used.

I still have to make that adapter for REW to be able to use a loopback as timing reference. I hope that will clear up the questions I have regarding the strange phase drop that occurs on axis around the xover point, but doesn't happen 10 cm off axis. Then, maybe I'll be able to determine the delay.

Also, the CD is behind a 2nd order filter, while the woofer has just a 1st order filter to attenuate the frequencies above the xover point (as you can see in the graphs, the woofer has quite some output up to 5 kHz. It's only about 10 dB down from the 200-Hz figure). The CD is also reverse connected. I could still go for a 3rd or 4th order filter on the CD...
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.