For anyone following the various DIY Synergy or Unity builds, or planning to build one, I've worked out a revised crossover on mine that gets it pretty close to linear phase (comparable to the Quad esl63 phase plots I've seen). With some help from Tom Danley, who also worked out an active crossover to get linear phase for the same speaker (I haven't been able to try that yet, but plan to in the near future). Thanks Tom for the crossover, and for the whole Synergy idea. Doesn't get much better than this.
Details at --
DIY Synergy/Unity spreadsheet
Details at --
DIY Synergy/Unity spreadsheet
Cool! I was just thinking about this yesterday, how I'm going to do the crossovers once I get my mids in and get a linear phase response. Or at least close to it.
Thanks guys. There is also a Youtube video of squarewave responses at coSyne DIY Synergy horn reproducing square waves.. sort of - YouTube
Concerning my approach: modeled in PCD, adjusted common delay to all drivers to get phase curve to be flatest on tweeter (after cd compensation applied). Set up polarities of drivers to get phases as near 0 degrees as possible. combine with low order filters outside of natural rolloffs, multiple sections when needed to get delay. Much driver overlap needed and a fortunate arrangement of driver depth offsets (synergy layout tends to make that more likely, there was luck in my case). Tom has a more methodical process described in his Unity patent. I mostly went trial-and-error watching response mag and phase while adjusting things in PCD.
Concerning my approach: modeled in PCD, adjusted common delay to all drivers to get phase curve to be flatest on tweeter (after cd compensation applied). Set up polarities of drivers to get phases as near 0 degrees as possible. combine with low order filters outside of natural rolloffs, multiple sections when needed to get delay. Much driver overlap needed and a fortunate arrangement of driver depth offsets (synergy layout tends to make that more likely, there was luck in my case). Tom has a more methodical process described in his Unity patent. I mostly went trial-and-error watching response mag and phase while adjusting things in PCD.
Has Tom seen this yet? Oh, I guess he has, he helped on the crossover!
I'll see Tom next week, we'll chat about it. Nice documentation over on the other forum.
I'll see Tom next week, we'll chat about it. Nice documentation over on the other forum.
Could a SH50 design work using a Beyma TPL150 AMT as the tweeter or does it not work like a compression driver in the horn throat to achieve constant directivity?
Any body know it the Beyma TPL150, an AMT tweeter, would work properly in a Danley Sound Synergy horn? Will the AMT "load" the horn properly like a compression driver? Will a waveguide control directivity on the AMT in the same way it does in a waveguide?
Thanks
Thanks
I think that it should work a treat. After all Beyma horn loads it. I want to try an AMT Synergy horn as well.
The surface area might need to be reduced a little bit for uniform polars.
The surface area might need to be reduced a little bit for uniform polars.
I'll be next to try ribbon Synergy and I just received the 12 Celestion drivers and pair of Akiplab 100db proribbons
Oh Yeah Baby! Now we are cooking. This sounds awesome Sergey. Which Akip Lab(http://www.akiplab.com/en/en_index.htm) ribbon are you using? Where did you get them and how much did they cost? Which Celestion drivers?
Would love to see how this build progresses.
This is Awesome!
Keep us up to date(you should start a thread)
FYI: BWaslo has built a Synergy with compression drivers and even recieved help on the crossover from Tom Danley.
ok, got a hold of myself now...got a little exited😉
thanks
Shane
Would love to see how this build progresses.
This is Awesome!
Keep us up to date(you should start a thread)
FYI: BWaslo has built a Synergy with compression drivers and even recieved help on the crossover from Tom Danley.
ok, got a hold of myself now...got a little exited😉
thanks
Shane
I took a very similar approach with my latest crossover version, first order filters with lots of overlap and a second electrical pole kicking in closer to the final rolloff (with resistive damping on the woofer shunt capacitor).
It models a lot like your results above, and sounds superb.
It models a lot like your results above, and sounds superb.
Badman...you are a Bad Man.
That is a beautiful monster. I notice you have her on castors...hehe. What drivers did you use and what is in the throat of the wave guide?
That is a beautiful monster. I notice you have her on castors...hehe. What drivers did you use and what is in the throat of the wave guide?
Thanks- the woofer is a BD BD15, one of the few 15"s one can trust to play a little higher (as needed for the relatively-linear-phase XO), and the compression driver is selenium d2500Ti. I modified the diaphragms with a homebrew coating of tacky glue (Foam glue, effectively the same thing as Aleene's Tacky glue, or several other white-glue products) and talc, diluted with water.
Made a big difference at the top end, commensurate with the smoothing one expects from damping. The very highest tizziness from phragm breakup seems to be gone, and a dip in the 6-8k region seems to have filled in, so it sounds more extended but less tizzy.
In the throat, is reticulated aquarium foam, a la geddes. I removed the bugscreens and smoothed the throat transition with clay, though I may do a silicone fill-in sometime, instead.
Attachments
Naw, but I decided to pull the active XO out of the rig for a while after I had something evil happen in my rig (think the switching pop from the pre set the crossover into some sort of malfunction mode, blowing the amp channel and associated woofer), and had been meaning to try the seleniums for a while, I'd previously run these OSWGs with JBL 2426h.
Xo is in the neighborhood of 1600Hz, but is first order down to about 1k, where a second electrical pole and a significant amount of acoustic rolloff begin, creating a very steep final filter function.
With the JBL 2426 this waveguide was very happy with a crossover more like 1kHz.
With the JBL 2426 this waveguide was very happy with a crossover more like 1kHz.
I'll assume this is directed at me- and this is a good juncture for me to bow back out of the thread, certainly don't want to steal the OPs thunder.
I don't have measures I can share- my measurements are based upon in-room with use of an RTA/Mic, I do sweeps and whatnot but can't capture them. I simulate with PCD, and measure to ensure that there's nothing disagreeing with it.
- Status
- Not open for further replies.