The **** Synergy: learning the art.

Hi guys,
Thought I’d start a thread to capture my foray into DIY synergy horns and use some of the expertise on these boards to make my attempt a little less… well... shitty.

I’ve been enamoured with the tech of Danley and read a lot of the threads on here with DIY homebrew options. I don’t have quite the technical background as some of the other builders, but hopefully with some help I can learn something and hopefully end up with a reasonable result.

I am living in Laos and getting components is really tricky so this first attempt is using existing components that I already have lying around- so by no means is ideal.

Here is what I’ve got to work with

Nad 3020B amp (just using the pre-amp section)
MiniDSP 4x10HD + UMIK
Marantz SR-7500 HT amp with 7.1 input section and discrete amplification (90W p/c)
2x PRV Audio D290Py-B polyimide 1" CD
2 sets of similar floorstanders that I have harvested for mid/woofer drivers. All drivers are 6”. One pair of floorstanders will provide all the drivers for one channel of the new system. The plan for the synergies is to play to about 80hz where they can pass off to my DIY tapped horns (recycling 10 and 12” drivers) from some old home theatre Subwoofers I had lying around.

I’ve made a 50x50 horn out of 12mm mdf as a prototype using the BWaslo spreadsheet. The CD is installed, mids mounted (sealed in plastic tubs with small amount of polyfill, with each driver firing through 2, 25mm ports. Mid ports are placed 72mm centre to centre away form each other (1180Hz horn cross sectional position). Mid ports are about 82mm axially from the screen of the CD. There is a 12mm thick, 35mm radius cone volume plug installed under the mid drivers that I used to push up their response after measuring them falling a bit short of the XO position with the CD (albeit this is when they were not yet sealed so that may have been why)

The placement of the mid ports was informed by some measurements of the CD roll-off, and the same is the case for the Woofers. I only have 2 small woofers for each Channel so I have them installed diagonally opposite, on a 12mm thick stand off mounting ring, firing through a 44mm port . The other holes you can see have 44mm right angle pvc connectors hot glued to the other side. I have some extra PVC that I can use to make ports longer to tune the bass section but have not got that far yet (flying blind on the theory behind the bass section here). The whole box is roughly sealed in a 50cmx50cmx50cm cube that screws onto the back of the waveguide.

I have done a (really) rough time alignment using the physical location of drivers and horn length but would like to do a better job of it as I know that can really make a speaker like this ‘snap’ into place.

If I post some raw- un eq’d responses would there be some willing people to help me whip this prototype into shape? I have been familiarising myself with the UMIK and REW but am by no means proficient in it yet. I’ve been doing tests indoors (because out house is surrounded by a big 10 foot concrete wall that would make outdoor measurements non-ideal) using Sine sweeps but as yet have not figured out gated measurements (that can only be done with pink noise right?) and don’t know much beyond pursuing a flat FR (I know phase, and impulse response need to be properly considered to achieve the synergy effect). Where is a good place for me to start?Advice on what level of smoothing to use on charts for them to be useful etc...


Thanks in advance for your knowledge and time!

Mark
 

Attachments

  • FullSizeRender(2).jpg
    FullSizeRender(2).jpg
    406.5 KB · Views: 1,168
  • photo(4).JPG
    photo(4).JPG
    182.3 KB · Views: 1,153
Hi Scott,

Forums are not letting me attach the .MDAT files. I'll find another way to get them to you. Measurement is with some attenuation of levels using the mini DSP. Measurements are raw- no Eq yet applied.

CDs are -13db
Mids are -6
Woofers are +2

Measurement is from about 90cm (indoors) directly in front of the CD exit.

Here are some response and phase charts for those playing along at home.

Mark
 

Attachments

  • combined raw FR.jpg
    combined raw FR.jpg
    69.9 KB · Views: 1,002
  • CD 50x50 .jpg
    CD 50x50 .jpg
    65.9 KB · Views: 970
  • Mids50x50.jpg
    Mids50x50.jpg
    73.8 KB · Views: 962
  • Woof50x50.jpg
    Woof50x50.jpg
    65.7 KB · Views: 165
A swept sine measurement can still be gated (and is a better way to measure than is pink noise,as log-swept sine can reject harmonic distortions and is much more robust against timing and clocking problems). Check your software manual, there should be a way to do it.

One of the features of the Synergy approach is that a linear-phase response is possible using passive components. (How much that affects audibility is up for debate -- to my ear, it seems to be "not much if at all"). If you want linear phase, though, I'd suggest a better less torturous way (and ultimately less expensive, probably) would be to design for flat response and then use an FIR DSP eq to flatten the phase. It's almost trivially easy with RePhase (and something like an OpenDRC) if you have a good magnitude and phase measurement of your unequalized speaker.
 
BTW, something I've found out from a recent synergy/unity-like project -- you don't really need to make the drivers symmetric in the horn. In fact, for home use you can easily get by with a single midrange. Notice that you have something like 20dB more sensitivity on the mid than the tweeter!

Multiple midrange and woofer drives gives ability for high output (damn near insane levels at midrange), but adds additional port diffractions to the tweeter's path. For point source behavior, the requirement is the drivers be close together, not necessarily that they be symmetrically clustered around the tweeter. So using the additional mid drivers costs more, is more work, and hurts the tweeter -- and their output isn't needed. I don't see much reason in a home for more than one good sized midrange. For woofers, there never seems to be enough cone area (and the horn doesn't really help it), so the argument might not hold there.

Losing placement symmetry could affect the crossover, though. If midranges are symmetric around the tweeter, you can basically ignore the phase shift between M and T as long as you get the magnitude right -- the waveguide sets the pattern, not the 'array'. But if you trim down the big farm of drivers, then you do need to pay some attention to phase at crossover if you want the one output coverage lobe to be symmetric in shape and not start to null. On the other hand, using a single midrange does take out the center-center distance requirement between the midranges themselves so it can actually help ease placement of the ports in the horn.

This is a new realization to me, but I'm getting pretty confident that it is a better approach. I made a 'Small Syn' speaker with a single mid and two woofs on just one side of the horn and the output pattern looks at least as good as my 9-driver-per diy synergies. (Build thread to appear here soon after I try a few new things, got some new tweeters in to play with)
 
Last edited:
Bwaslo, you make a good point about typical home needing less dB than the local Fairgrounds or Arena :) However, are you sure that a single cone mid will be much hotter than a compression driver tweeter? I would think the opposite. In any event the laggard in a Synergy system is the woofers. It's so hard to get 120 dB SPL at 20 Hz 1 meter out of a single woofer for some reason :)
 
Soldermizer -- No, the one mid won't be hugely hotter than the tweeter, but four of larger sized ones will be, well usually at least more than 6dB hotter than the tweeter. See the OP's curves at http://www.diyaudio.com/forums/multi-way/292178-shitty-synergy-learning-art.html#post4732939. Dropping to one mid would be only 6dB below that (assuming series-parallel wiring). With a constant directivity (essentially conical) waveguide, the tweeter isn't really all that hot since you have to eq its entire response down to the tweeter's lowest inband point (somewhere around 97dBSPL usually). With a tractrix or (gasp!) an exponential horn the highs get squished into a narrower pattern so at least on-axis the tweeter gets a little more sensitivity then.
 
Last edited:
Thanks for the interest and responses so far!

Interesting observation Bill about not needing to make the number of drivers- particularly the mids, symmetrical. I am a bit of a SPL junkie and will probably keep this lay out for the moment as I use my system for largish outdoor garden parties (and lazy afternoons at home when my partner is not around to tell me to turn them down :D)

After a bit of a chat with Scott (speaker Scott), I've got a better idea as to what is needed measurement wise to be useful to fully tune these cabs. This weekend I'll try to get a new set of measurements outdoors (2m ground plane), with a time alignment benchmark (trick with running an old silk dome tweeter in parallel with the drivers in question).

I'm also going to spend a bit more time sorting out the CD transition to horn, and making some frustrums for the mids to help push up the passband.

@Pano- thanks for the tip re: zipping the .MDAT file to post it to the forums.

I'll update the thread when I have some more usable info for you guys to look at.

Cheers!
 
Another possible benefit of symmetrical mid/woofers, especially for 2 way synergy horns, is that the woofers are almost push-push across the horn, so force cancellation reduces vibration, especially if their mounting plates are braced across each other outside the horn..
Otherwise one woofer pounding on one side of the horn wall could cause some vibration issues?

Sent from my GT-I9505 using Tapatalk
 
Bwaslo: thank you for clarifying. I will keep this in mind for my first "real" Unity project: I just won a pair of Yorkville U15 "for repair". Even if all the drivers are trashed (a not unreasonable assumption coming for old Pro gear...?) I will probably buy only the parts needed for home use, such as one mid per horn. I will plan to actively Eq it so one driver or three, who cares :) Further details will be in my own private Ijit quarantine thread :)
 

mayhem13

Member
2008-09-22 4:37 am