Small Syns

ra7

Member
Joined 2009
Paid Member
Hi Damian,

I did a measurement (about 3ft from the baffle) today so I could work up an FIR phase eq for it. Here is the response, with (red) and without (black, no EQ) some mild PEQ I added for in-room bumps. Sorry, I only saved 1/6th octave curves, since that is what I was adjusting the FIR EQ with. The bump near 100Hz is from the speaker itself (interaction between crossover and the driver impedance), the dip is floor bounce.

I'm very happy with the sound, though it makes me even more bummed about the tweeter being discontinued. I haven't gone outside to do polar measurements, but the on-axis is very close to the same as the sealed version, as is the sound (except for more powerful sounding bass). I did tweak the crossover a little (the same should probably be done on the sealed version). C3 and R1 change a little. Not sure if it can be heard, but the midrange gets a little flatter, at least with the particular instances of drivers I used.

That looks really good! Is that bottom end without a sub? Amazing!

I see you put the coSynes for sale. How would you compare the sound of the two speakers? These are smaller, but does the smaller horn make a difference to the sound in the same environment?
 
That's without a sub, with the speakers up against the wall behind them. The room is helping a lot. More than I expected, actually.

I think the new ones are smoother sounding than the CoSynes, but that could just be that they're newer to me? The new ones don't have directivity as low as the CoSynes should, but the ceiling is so low in the listening room (basement) I don't think either has enough to matter. Fortunately, I have the volume between the wood beams above stuffed with absorber.

The newer ones are visually smaller, though, which fits better with my (and my wife's) recent inclinations. Besides, it's not surprising when big speaker boxes sound big, but it is when littler ones do. The big ones are linear phase as-is without EQ, but I'm not afraid of electronics nor the type to avoid EQ, so the miniDSP takes care of that with the Small Syns (I'm still not sure whether there'a much of an audible difference though).
 
Looking good bwaslo! Looks wise I particularly like the proportions of the sealed box (looks golden ratio).

Any chance you can provide CSD plots of the compression driver and phase and group delay plots of the completed speaker before and after implementing the FIR filter?

Next time I set the mic up in the room, I'll get some more graphs for you. Not sure the CSD of the compression tweeter is going to matter to many people, though, since the driver is discontinued. I should be trying to work in some different drivers in the next several weeks. Erich is sending me some DNA-360s.
 

ra7

Member
Joined 2009
Paid Member
Thanks! Yes, the new HD versions of the miniDSP seem to be quite useful. I think I will order some SEOS-15 waveguides. Is there are link to your final design and crossover?

My favorite 1" CD is the BMS 4550. The BMS design is slightly different than other CDs, which have a traditional phase plug. I did a 1" CD comparison once and somehow, I felt that the 4550 just sounded smoother in the sweep.
 
Thanks! Yes, the new HD versions of the miniDSP seem to be quite useful. I think I will order some SEOS-15 waveguides. Is there are link to your final design and crossover?

My favorite 1" CD is the BMS 4550. The BMS design is slightly different than other CDs, which have a traditional phase plug. I did a 1" CD comparison once and somehow, I felt that the 4550 just sounded smoother in the sweep.

The crossover is at message 158 of this thread (#numbers are Parts Express stock numbers). http://www.diyaudio.com/forums/multi-way/292379-small-syns-16.html#post4837592

There is a pretty complete set of build diagrams for the shelf ported ("Not Quite As Small Syns") at http://www.diyaudio.com/forums/attachments/multi-way/570498d1474318272-small-syns-sp-box-drawings.pdf
I haven't got to build diagrams for the sealed version yet, though they shouldn't be hard to figure out from the ported one.

Some instructions for modifying the horn are at threads 97 and 98
http://www.diyaudio.com/forums/multi-way/292379-small-syns-10.html#post4788726

Response, polar and distortion measurements for the sealed version are at posts beginning with #77 http://www.diyaudio.com/forums/multi-way/292379-small-syns-8.html#post4787283

I'll gather this all up into one guide once I get the tweeter situation resolved.

I've heard good things about the BMS 4550, but I've also heard it has a rather long path length inside the driver, which is a non-starter for this design. I need the path length from the diaphragm to the throat exit to be as short as I can get it, otherwise the quarter-wave notch on the midrange will move down too low.
 
Hi Bill,
Do you use hornresp to simulate different drivers or back of the envelope, cut and try? Thanks.

Regards
Simmonds

I use HornResponse to figure the midrange port sizes and lengths and to check whether their locations will work ok. Cut and try would make you go through drilling holes into a lot of SEOS15 horns to get it right! I don't model the tweeter response, though, since I know (have measured) how that works on an unmodified horn.
 
re' I've heard good things about the BMS 4550, but I've also heard it has a rather long path length inside the driver, which is a non-starter for this design. I need the path length from the diaphragm to the throat exit to be as short as I can get it, otherwise the quarter-wave notch on the midrange will move down too low.

You would definitely have to lower your XO to use the BMS4550 but I think it would handle the lower XO quite well. I'm crossing it at 950 Hz in my Synergy and I measure THD at -50db throughout the XO region and up past 10 Khz at levels too loud to listen to but otherwise unquantified.

I modeled the BMS4550 at 6.5 cm path length, if I recall correctly, and then trimmed that by 1.1 cm to match null freq measurement. I know you don't want to redo your passive XO but for somebody using DSP it might be an option.
 
I was drawing up the sealed box version, and found I made an error in the drawing for the woofer apertures in the drawings for the Shelf Ported version recently built. See corrected dimension in bold here -
correction%20of%20SP%20apertures.png

Also attached is a corrected pdf of the whole box.

(BTW, I got some good news privately about the Tymphany compression tweeter - waiting to make sure it's ok to announce it or if they will....)
 

Attachments

  • SP Box Drawings.pdf
    631.9 KB · Views: 118
Last edited:
Appearance with different panel thicknesses

I'm leaning toward re-designing the sealed version to use half-inch (12mm) ply instead of 3/4" nominal ply (19mm). That would be one fewer size of pre-veneered (or not) ply to obtain, would allow for a little more internal volume for the box, and I think looks a little better, too.
Appearance%20comparison%20half%20inch%20panels.png

I don't think the wall stiffness will be affected much because of the rather narrow sizes of the top, bottom, and side panels, and the use of internal braces and stiffeners in general. The front panel construction would be the same, as that has to be done for the different layers needed on that.

I'm also making the cabinet 1" deeper, to allow for optional use of larger/deeper HF drivers (even though it looks like maybe the DFM-2535R00-08 will stay around, after all - they are on sale at Parts Express till the 30th, by the way, might be worth getting a few while they're cheap).
 
The Tymphany HF CD is no longer "discontinued"

The people at Parts Express convinced Tymphany to continue production and availability of the DFM-2535R00-08!

It turns out the driver they were going to replace it with was both (1) different and (2) more expensive, so not really a good replacement. Apparently it was selling well enough at PE and PE has enough influence to convince the manufacturer to keep it active. (thanks l, Parts Express)
 
That's great news! Glad to see the project continue as-is.

I've been getting parts together over the last few days, and furiously selling things on Craigslist to fund them. A banjo is next ;) All of the drivers are ordered and shipped, and the waveguides are on their way as of this afternoon.

I'm visiting the lumber yard tomorrow with a friend and his truck. I was going to ask you about using all 3/4" plywood, but now will just buy 1/2"!

Time to start reading up on the crossover. I have plenty of experience soldering cables and laptop parts, but my electronics knowledge is not advanced. Got a TPA3116 amp on the way, to play with and potentially mod. Reading up on the MiniDSP 2x4HD also.

I don't quite understand how impedance works at the input of the amp, and I'm trying to figure out if I should use the miniDSP before or after the preamp. I'd like to use my TV as a switcher and go optical out to the miniDSP. But using the DSP as a "receiver" seems somewhat impractical (no visual indicator of volume, use of IR remote control changes volume 3-4db per click). Anyways, all problems for later down the road, but any input is appreciated.

Looking forward to making sawdust over these next few weeks. Thanks for the consistent updates!
 
Bill, that's great to hear regarding the Tymphany CD. I already have some, ordered just in case they did go out of stock.
I have a question regarding the use of two mid ranges, would that help to increase the overall efficiency or would there be more to it?

Thanks for all your great work.
John
 
Hi Sphykik,

(can I assume that there's a great story behind that name?)

I'm visiting the lumber yard tomorrow with a friend and his truck. I was going to ask you about using all 3/4" plywood, but now will just buy 1/2"!

That would be a good plan, if you are building sealed versions. If you are going after cosmetics, look for hardwood pre-veneered ply -- I'm partial to "red oak", but there are usually several types available. Get some iron-on veneer edge tape, too.

I've been working up clean drawings for the 'sealed', and since you are already getting lumber, attached are what I came up with (so far) for cutting plans on the plywood panels. It might help. I should have the rest of the drawings up sometime today or tomorrow.

Time to start reading up on the crossover. I have plenty of experience soldering cables and laptop parts, but my electronics knowledge is not advanced.

There's not too much you need to learn, just how to solder parts and try to keep inductors away from each other.

I don't quite understand how impedance works at the input of the amp, and I'm trying to figure out if I should use the miniDSP before or after the preamp. I'd like to use my TV as a switcher and go optical out to the miniDSP. But using the DSP as a "receiver" seems somewhat impractical (no visual indicator of volume, use of IR remote control changes volume 3-4db per click). Anyways, all problems for later down the road, but any input is appreciated.

The output impedance of the 2x4HD should be ok with about any amplifier or receiver input, it is pretty low. In the living room here, I drive optically from a TV (which is connected to a media center box via HDMI) to the miniDSP 2x4HD, and from the miniDSP right into power amps (I'm using two ICEPower 200ASC modules at the moment).

The lack of volume indicator is an inconvenience, but a partial workaround is to set the TV's internal speakers to 'OFF' and program the miniDSP to use the same volUP and volDOWN remote codes that the TV uses. On my TV, that causes the on-screen indicator to still show the volume increasing/decreasing, though the TV doesn't make any sound itself. The miniDSP sees the codes and does the actual volume control. You don't really know the volume setting, but you can see that the remote is sending something out at least and what direction it is going. I haven't checked but I think the volume steps are less than 3dB (doesn't sound that course, seems kind of slow, actually).

Some people would tell you (and I would have even told you, too, till I actually tried it) that using "digital volume control" into the DAC (which is inside the miniDSP, right before its outputs) will lose resolution when below full volume. But I don't hear resolution being lost even at low volume, I think any loss is down below normal domestic room noise. I think it would depend on gain structure, and if your amp has about the typical gain level of ~26dB, you should probably be fine.
 

Attachments

  • Sealed Box Panel Cuts.pdf
    375.8 KB · Views: 107
I have a question regarding the use of two mid ranges, would that help to increase the overall efficiency or would there be more to it?

It wouldn't affect the efficiency in general, that would be set by the least efficient drivers in the system (which is almost always the woofers, and is in this system, too). Two midranges would allow for higher volume in the midrange before distortion, but again, given that the woofers would be more likely to run out of gas before even the one midrange would (assuming music signals), I don't really see the point.

But if you do want to go dual midranges, you'll need to change the crossover (both impedance and needed attenuation change - not difficult, but needing attention). Also, the baffle would need to be taller so that the extra midrange lump sticking out of the top of the waveguide has somewhere to fit behind.
 
BTW, I think I never got around to mentioning it, but the "Durham's Rock Hard Water Putty" worked very well for filling the volume between the driver and the waveguide. "Rock Hard", they aren't kidding, it's like a piece of ceramic.

The only down sides are that it takes adjusting to get the consistency right (add water -- too much!, so add more powder -- too much! so add more water....repeat...). And that it is kind of messy. The stuff is yellow and leaks down inside the waveguide, so you have to make sure to clean there with a wet rag before it hardens (which happens before long). After it hardened, I took a black felt-tip marker to blacken any of the yellow that showed through the midrange ports.
 

Attachments

  • Durhams.jpg
    Durhams.jpg
    64.4 KB · Views: 655
I forgot I did this last night :)

-- here's an "alternate" panel cutting plan for the half-inch ply, based on 24"x48" sheets. For those of us who are panel-size challenged (I can't get a 48"x96" or even a 48"x48" into a Honda FIT!).

Actually, this may be a better (better use of wood, easier to cut) plan even if you can transport larger panel sizes --

Alternate%20Half%20Inch%20cutting%20plan.png


-- so, disregard the first page in the just-posted pdf from this morning

Bill
 
I use HornResponse to figure the midrange port sizes and lengths and to check whether their locations will work ok. Cut and try would make you go through drilling holes into a lot of SEOS15 horns to get it right! I don't model the tweeter response, though, since I know (have measured) how that works on an unmodified horn.

Thanks Bill. How about the "array effect of the woofers"? Is it possible to simulate it ?

Simmonds
 
How about the "array effect of the woofers"? Is it possible to simulate it ?

Simmonds

Hi Simmonds,

There are a couple of ways to do that. You could use PCD (using the same driver type for "woofer" and "tweeter", at same height, with varying horizontal offsets) and watch the off-axis response vs frequency. Or calculate what the relative delays will be due to different distances from drivers (via Pythagoras, from 8th grade geometry, and the speed of sound) when you are off at at some angle, then put those delays into XSim. You could even use a spreadsheet, calculating the sum of delayed sine waves. In any case, assume the drivers are omnidirectional (which they should be for the most part at lower frequencies, though really, aperture and baffle sizes will still have some effect-- but omni is good enough for approximating). You also have to know about where you want the array to start kicking in, since you want to blend into that from where the waveguide's directivity starts to drop out.

Then, put woofers and waveguide on a baffle and take measurements of waveguide and woofer array at several angles. Finally, design the crossover for flattest falling response at those horizontal angles from the baffle, consistent with what ranges the drivers can cover with decent overall response smoothness and power handling.

I also did the same process at first for vertical off-axis, but it turned out that the trick for that was to get the woofer apertures as close to the waveguide center as possible and then make sure that the phase responses of the woofers and midrange waveguide track each other though the crossover range. Most of the crossover design time was spent getting that phase tracking along with the requirements for horizontal off-axis responses.
 
Last edited: