Reposting this here since I still tinker with the measurements every once in a while. Eventually I might find some sort of satisfaction with it.
1 Like
On multiple occasions I seriously look into swapping the tweeter for the DX20BF00 because it is the same OD. Though, I believe there is a difference in the flange depth. And there was a post on old MAC/DIY about the XT25SC90 that got me looking into it again.
I tried nixing the tweeter notch and bumping up to a 3rd order electrical highpass. Along with 3rd order on Mid lowpass, and lowering the crossover point a bit. In doing more massaging I was able to give one resistor the boot. Now at 11 parts +4 resistors.
Reminder: I didn’t model the baffle step here so I’m doing some guesstimating on the woofer slope. Might look like alot, but in-room, with only one speaker, it seemed a little lean in previous testing. I’m betting on it filling in with both playing.
That’s the same woofer as Paul Carmody’s “Tarkus”, isn’t it, although he used it vented. I don’t know enough to work out how much BSC he put into the design
Geoff
Think tarkus used a 10" or 12"
1 Like
Yes, my mistake - he used the 830668, 10” woofer
Sounds like a nice project, good luck
Geoff
1 Like
No problem at all man. I’m the dunce here. I’ve been working on this thing for nearly 2 years now. I realy need to get it done. Would give me more reason to show up at the Detriot gathering with this in tow.
1 Like
Sometimes when I’m in a rut I use EQ to raise and lower frequencies looking for where the problem is.
3 Likes
Alright. Yet again I think I’m at a decent result. Not my best crossover work, but I think the drivers make up for my failings. This might be the one that gets the solder.
I’m not getting the wierd shriek in female vocals that made me want to drop it in the dumpster. The measurements show a hefty room mode suckout still. I’m not thinking that is what is making it seem lean. Maybe the low distortion? I’m still listening with only one speaker. Otherwise I’d have to make another 3-way wire passthrough plate. This makes me not the biggest fan of sealed 3-ways unless I have 3-way binding post plate.
-Ungated response so has all my livingroom reflections.
-Thinking that peak just above 600hz might be baffle diffraction? IDK it isn’t realy bothering me.
-I left in a bit of a rise in the tweeter response to make up for the narrowing dispersion of the ring radiator.
-Fortunately the impedance ended up rather benign. It needs some power to party.
I’ve had a weird name for it that stuck in my head a while now. Pahoehoe as in the thick viscus type of lava that often forms a ropey grain pattern when it is still trying to flow as it cools.
1 Like
Not bad, great name. Wonder if woofer is doing something at 600, as there is a peak in its response that is causing it.
Yeah I tried chasing it for a minute. But the nearfield nixed that theory for me.
Since the peak shows up in both the woofer and mid response I wrote it off as the first baffle diffraction peak. The sides are hard edged 90s with zero radius.
you can always try taping some cardboard “wings” to the baffle to satisfy any possible baffle diffraction wiggles.But a nearfield measure should isolate it also.
I don’t think I’ve ever seen baffle diffraction that low before.
1 Like
There is often a big bulge just above baffle step, but it is odd that it is so peaky. Especially since the M/T are offset on the baffle to try and smooth out the diffraction some. It might just be a room artifact. Either way I’m not too worried about it as it is not something I can detect in the sound. I need to get both speakers playing to do the true final determination. Due to the offset I can try listening to them in either orientation too.
2 Likes
I did a big push and got the crossovers assembled and installed. The impedance measurement ended up with an extra peak at the M-T crossover. I’ve gone over the circuit several times, so I’m thinking the previous measurement may be different due to the connection issues I was having with the wago lever nuts. Both have the same blip so, if it is wrong, at least it is consistent 
Too late this evening to do any serious listening or response measurements. Will have to wait for tomorrow.
7 Likes
Those look great! Crossovers are nice and tidy, as well.
1 Like
Yeah, I second that, nice round over detail at the top. Clean set of speakers👍🏻
1 Like
I think most of you guys know these drivers well enough to know what they are capable of. And, in spite of my jackassery 
, they seem to deliver. What I found surprising was how dynamic the result seems for being likely in the low 80s db/w. Having both speakers playing definitely fills in the low end, even at times seeming a bit much in my small living room. Crossing fingers that might make for a better showing at the gatherings.
1 Like
That’s awesome you finally got these where you wanted them! Better grab a sixer and sit down for a session!
I just started cutting wood on a WWMTM using dual 830667, TC9 and DX20. What enclosure size and tuning dis you end up going with for the 830667? I keep going back and forth on how I want to tune them
I was shooting for a relatively compact bookshelf so I went sealed alignment. 10"x11"x18", 3/4" material. Bracing was a scrap piece running front to back and one side to side. There wasn’t a whole bunch of room for much else. The midrange enclosure was a plastic 1liter festival beer mug. So I estimated the final internal volume around .72cf / 20liters. Sofar just lined with acoustic foam rather than stuffing it. IMO it can get away with being a bit boosted in the voicing, as the “tightness” of sealed alignment keeps it from sounding flabby and overbearing. Scaling that up for two woofers would result in a very svelte tower.
Modeled with a guesstimated .5ohm coil DCR in series.
50hz F3 always looks a bit dissapointing for an 8", but that slight peak gives a nice extra little punch to bass drums. And the slope is still rather shallow. So, with room gain, you still hear alot of that LF stuff without it trying to steal the show.
But… ported does unlock alot of LF potential. When I model it, 2.5-3cf for a pair seems doable (f3 in the low 30s). 3" port would probably work well if dual flared. That way port resonance should be way over passband. My personal reservation is that lowers the response peak down closer to 60ish hz. I tend to shy away from that for fear it could result in too much boominess for my taste in my small living room. Though the peaking is less than 2db so probably splitting hairs in reality.
So yeah, I understand the difficulty in the decision. I suppose ports can always be stuffed. Then “uncorked” if you bring it to a large room, or you just wana unleash the beast.
I have a few of that subwoofer’s (at least what Peerless calls them) next bigger brother, the 830668 10”. I’ve modeled them six ways to Sunday and always land on a very reasonably sized sealed box with a F3 around 50 Hz. I want to build a pair of 3 ways with them in that sized sealed box. Having modeled them in ported boxes and hearing the Tarkus… I just don’t get that.
1 Like