Small Syns

I posted a few plots of this speaker build on another thread of xrk971's, so thought I'd stop hijacking his thread (http://www.diyaudio.com/forums/multi-way/285030-bookshelf-multi-way-point-source-horn-post4734171.html) and start a new one.

_______________________________________
Edit Dec 18, 2016:
Since this thread is long and meanders a bit (or a lot -- including extended debugging times), here are links to full plan sets, final crossover design, polar and distortion measurements:

When the weather lets me out to work with the saw again, I'll be building a shallow version of the Shelf Ported Small Syns. A 29" tall, 20" wide and only 9" deep box for mounting near the wall with minimal room intrusion.
_______________________________________

This one is about a project to use a SEOS15 in a partial Unity/Synergy arrangement, with a midrange driver bandpassed into it and woofers ported just outside and below the waveguide through a pair of slots. The idea is to extend the horizontal directivity a little lower with the array effect of the woofers, while keeping their output locations close enough to the center of the waveguide to keep vertical polar behavior (relatively) well behaved - an essentially point-source behaviour with directivity. I posted some plots of an earlier attempt of it using a small Celestion CD for the tweeter, but it had some distortion issues trying to go too low in frequency. So starting here will be the version using the new Peerless DFM-2535R00-08 driver (which, contrary to angst and gnashing of teeth comments within this thread, is NOT discontinued, and is in fact available from Parts Express at a very reasonable price!).

Here's some pics of the cabinets to show what it looks like.

Sealed%20no%20grille.jpg


Grilles.png


One%20Sealedin%20LR.png


SP%20Box%20resized.png


What I'm going for is:
  • smaller size than most Unity/Synergy variant speakers
  • better WAF (a non-flashy understated cabinet, works well in any room)
  • designed to work close to wall behind it
  • Easy to build without exotic tools (buzz saw and jigsaw, and drill mostly)
  • Good polar behavior, particularly in the horizontal
  • Good sensitivity >93dBSPL 2.83V/1m
  • Tolerable impedance (4 ohm nominal, try to stay above 3 ohms min) (ended up with a dip to about 2.9 ohms)
  • Relatively simple crossover (got less simple, but still not bad)
  • A sealed version (will need subs for any real bass) and a ported full-response version for full-range
I'm NOT going for linear phase for these designs, as I think it is much more practical to linearise phase after-the-fact with a FIR equalizer if that quality is judged to be important (I'm not entirely sold on it after some tests -- polar pattern, single point source, and response shape seem to be a lot more important, from what I've been able to hear in different speaker designs).

The trick to this design is to blend between the midrange and the woofer pair to get reasonably flat, but dropping in level, off-axis mid-high response. It's not very easy to get! I've been simming this in XSim using separate driver objects for each driver type at each angle, connecting the matching types to balance the off-axis response by trial and error.. It takes a fair amount of time to work into something reasonable, but build results do tend to match the simulations quite well.

Harmonic distortion in the tweeter range is pretty good now with the Peerless tweeter and a little more agressive crossover slope

HD%2095dB%201m.png


Another midrange could be added on the other side of the waveguide, but I think this is ok without it and minimizes diffraction within the waveguide for the tweeter.

When I get things more complete, I'll put up some build details, including how to modify the SEOS15 waveguide for synergy operation, in case anyone wants to duplicate the design or use it as a starting point for further developments. Please stay tuned!

edit Feb 22 2017:
Next (shelf ported) version --

SPSmallSyns2_appearance.png


I hope to get started on this in a month or two
 
Last edited:
Here is a view of the drivers' contributions to the total response (on-axis). Notice how the midrange hands off kind of slowly to the woofers. As you get off axis, the woofers' responses start to interfere with each other causing a dip, right about where the directivity of the SEOS horn goes away and the mids go toward omnidirectional. The output from the mids falls away there and the woofers take over and keep some pattern control going down below 500Hz .
simulated.png
 
Here is how the midrange mounts to the waveguide. The flat mounting plate board is glued on, and sealed with, epoxy putty. I'll have cutting diagrams and dimensions, as well as hole drilling info in some later posts.
mountplate%20attached.png


Here is shot showing how the woofers hide from behind the waveguide, the baffle is milled/routed out down to two different levels. That can also be done with less tooling and no jig fabrication by jigsawing thinner wood panels and them laminating them together with wood glue.
Bafflecutouts.jpg
 
There also need to be spacers between the woofers and the baffle that they mount behind to keep them from slapping into it. I just cut thosr out of 1/4" MDF with a jigsaw. They also get laminated onto the baffle. You can also see some plywood bracing strips glued to the back panel and to the box walls to keep the panel quiet.
spacer.png


edit: by the way, do you see that patched area just to the bottom right of the waveguide mounting area? Kind of gray, like a hole was filled with wood putty and a dowel? Well, I don't want to talk about it...
 
Last edited:
Here's where I'd like to be doing measurements - sadly, the crossover being measured in the picture was a miserable failure from about a month ago, horrible vertical pattern. I live right between two elementary schools, so I can't do outdoor measurements during weekdays till summer break. I'm pretty sure the scary alien noises made during speaker testing would have them sending a SWAT team out on me.... I got a couple of troubled visits from the neighbors the Saturday I took this pic. It's near 100F here this weekend, so I'm not up to dragging out all this stuff until the weather cools again.
Outdoor%20measurement%20range.png


The jist of saying all this is that the measurements shown earlier were made in a somewhat cramped basement, so they don't go very low in frequency and have some unavoidable reflection interference....
 
Last edited:
Very cool project and an interesting approach with the woofers outside of the actual horn, but still behind a band pass hole. It would be hard to resist putting in one more mid though. At least for me it would be.

What is going to be the bottom end for this? You mentioned a ported enclosure that would require a bigger box. Is the drop off to steep to try a Linkwitz Transform?
Would depend on the woofers too I guess.
 

Eldam

Banned
2012-09-13 12:25 am
France
Very Interresting : cheap + off the self horn = :)

How is performing the Tymphany compression driver VS the very good plastic diaphragm from B&C or Beyma please ? As non fatiguing but more detailed ?

Can we consider it's 20° horn seing the polar map ?

the woofs are ok to around 100 Hz (so sub-woofers plated amp are needed ?) or a Linkwitz transform without subs ???
 
Last edited:
What is going to be the bottom end for this? You mentioned a ported enclosure that would require a bigger box. Is the drop off to steep to try a Linkwitz Transform?
Would depend on the woofers too I guess.

The ported enclosure is planned to be just a similar box of about double the volume (deeper box and a little taller to keep it looking small in a room). Same drivers, predicts to be about 42Hz f-3.

I've been running these in the living room (Celestion tweeter version) and an (aggressive) Linkwitz transform works pretty well for that application here, but it doesn't get played very loud. Pretty sure it would get rough if we tried loud hard rock on it. I also had quite a bit of other EQ applied (for the midrange bump and wall effects, as well as FIR phase eq), and with all that it works out pretty nice (big sound except for low bass) but not too elegant a way of getting there.

Downstairs, away from walls and with well-placed subs they sounded very good without EQ, though. The worry about the distortion with the Celestion tweeter is mostly intellectual, they didn't really sound bad at all. I also wanted to play with the new Peerless driver:p
 
How is performing the Tymphany compression driver VS the very good plastic diaphragm from B&C or Beyma please ? As non fatiguing but more detailed ?
Too early to say.. I've only heard it much with test tone sweeps:eek: . I actually found the plastic Celestion CDX1-1445 to sound smoother to the ear than the B&C plastics, but it doesn't handle going low in frequency very well (measurably, though sound doesn't seem bad to me).

Can we consider it's 20° horn seing the polar map ?
The polar map shown earlier is for the VERTICALS, and it's a 40° horn vertically (+/-20°).

Horizontally, it's a 90° horn, I didn't show the polars for horizontal, but here they are --
Horiz%20Polar%206th%20octave.png


the woofs are ok to around 100 Hz (so sub-woofers plated amp are needed ?) or a Linkwitz transform without subs ???

See the post just above about bass. As-is, these sound kind of anemic without either subs or LT. Probably these particular boxes will go back downstairs (where subs are ok with the management) and shelf-ported full-response versions will take up residence in the living room.
 
Last edited:
Did a crossover tweak today, just component values (will post it when I post the rest of the build details, in case anything changes still).

Here are the horizontals du-jour:
Horizontals.png

Horizontal%20Polar.png

It don't got squarewaves, to get that you'd need a miniDSP 2x4HD (I'll have a file available for that when I get to making it). Not enough delay on the tweeter and too high order filters on tweeter and woofer.

Going to try the 'near the wall' crossover tonight if I can, but I just heard company arrive, so probably not....
 
Allen,

It would be a lot of work to do with only mids playing, but here are 0. 20. 30. 45 and 60 degrees curves (without the crossover) that were taken to do the crossover design with
midrange%20no%20crossover.png



Here is a plot showing the same angles (I think) for all three drivers (tweeter was the Celestion when this was done), to show the array effect in the woofers.
Driver%20Curves%20Bare%20Smoothed.png
 
Last edited:
I've been trying to put a cut into the midrange drive in the crossover down around 300Hz, where it doesn't add much to the response but seems to be distorting more than the very clean (now) tweeter. I haven't found a way to do that, though, that still keeps the mid and woofer in phase with each other over the range where they trade off. When those don't track, the vertical pattern goes all to hell....

I guess the other option is to use two midranges. Cost on doing that isn't a problem, the drivers are rather inexpensive. But mounting them onto the horn is one of the more time consuming parts of this, I hate to double that without good reason. Also, the box would have to get a little higher to accommodate the midrange driver bump sticking off the top of the horn. For a ported box, though, it has to get taller anyway.

With the single mid, the midrange distortion still looks lower than the woofer's (as it usually goes....). I still like the idea of minimizing port holes into the waveguide.

If anyone wants to try doubling midranges, I think it should be just a matter of adding the other driver the same position and sizes as the first, and wiring them in series. Impedance doubles so power absorbed halves, but efficiency doubles too, so things should break even? Except the crossover components, have to cut capacitor values in half and double resistor values. Second harmonic distortion should drop by about 6dB (to 1/4 the percentage).

First test on the 'against the wall' indicate that I really screwed something up in my sims or, more likely, measurements. In the basement, with the Small Syns against a wall, the response shape is much better controlled than I measured upstairs. I suspect the wall behind the mic might have been getting into the act too much (don't want to design taking that into consideration). So, I'm doing a new crossover for "against the wall", it looks like it will be more similar to the "out from the wall" crossover. It would be nice if it could be made switchable without using too many extra expensive parts..
 
Last edited:

AllenB

Moderator
Paid Member
2008-10-18 11:31 am
Thanks. These (horizontal along vertical 0 degree plane?) polars appear to show a reasonably clean resonance at the notch. Can I assume the ports are at just under 2" axially from the throat where the horn height is a little under 3"?
 
Hi AllenB.
Those are at the 0 degree vertical (and are horizontals). The rest I'd have to check....

The ports are as close to the horn throat as I could reasonably get them. The notch is at 2.15kHz. I don't run the midrange very close to the notch as I didn't want to use difficult-to-document and duplicate frustrums on the port holes I'd need to get the high frequencies up to it). The ports are simple holes cut with a wood drill bit through the 0.35" thick horn walls, though I do fill the volume in the mounting plate with putty, with finger-shaped paths from center to the holes (those don't seem to be be critical at all).
 
(not) fun with distortion

Well, I figured out a crossover that cut out the drive to the midrange near 300Hz where distortion was indicated, while still keeping the woofer and midrange in phase with each other where they trade off, while still behaving off-axis, and without getting impedance too low anywhere. It even makes the response flatter.

So, I measured distortion, and the distortion near 300Hz is nice and low, but there's now a narrow 2nd harmonic distortion peak at ~450Hz! I spent several hours fighting this as well as other distortions that crept in during the process. The more I looked around, the worse it seemed to get. Like a switch on my power amplifier that decided to go cruddy then (I of course took apart the horn assembly, removed woofers, etc, thinking it was something in the speaker). And a too-long screw when I put the horn back in, that started buzzing against the cone of one of the woofers. But in figuring out and fixing all that, I found out that the 450Hz distortion peak isn't even coming from the midrange (and now I wonder whether the 300Hz stuff was?).

It's from the woofers, both of them I think! (but not sure... hard to tell and I was worn out from all the fun by that time). It's there even with the midrange disconnected (disconnection was caused by the crossover hodgepodge falling off the table onto the floor). I guess I need to take a woofer out and see if I can measure it by itself outside of the box.

Does anyone have any ideas what might be causing it? It shows as a 2nd harmonic distortion when driven at 450Hz and the woofer's impedance curve shows a wiggle right about 900Hz (the published curve on the data sheet also shows that). A resonance of the surround? Could it be getting excited when the woofer is doing 450Hz, and reading as distortion? The peak is very high-Q, so I think it has to be something mechanical.

The distortion curve with new crossover -
hd1.png

The impedance curve of the woofers (in the box)-
woofers%20impedance%20in%20box.PNG

The HD2 is on the order of a percent on the peak, it mostly just bugs me wondering where it is coming from.
 

AllenB

Moderator
Paid Member
2008-10-18 11:31 am
Their impedance curve does show some irregularity at about 900Hz (resonance in the rubber surround?), would that be expected to result in a measured narrow disortion peak when driven at 450Hz?
Spider maybe? If I understand you are not measuring distortion per se, but the level of the harmonic? I was also thinking something to do with the cabinet?