I have several Dayton UM15-22 'Ultimax' drivers I am retasking from an infinite baffle installation. Since I already own them I want to use them in a non-IB installation in a large room (30'x40'x12'). In that pursuit I have been comparing all of the usual designs... Sealed, Ported, TL, and in this case Tapped Horn. I've built several sealed, ported, and TLs but not a TH so I could use a sanity check on this design:
I'm including screen shots. The CAD model is PRELIMINARY and I know it needs more like bracing and driver access; it is just a sanity check on practicality and folding. It is roughly 89in x 20 in x 20 in.. The cabinet section lengths and cross sectional areas match the Hornresp data.
Thanks in advance
- The peak group delay is 64ms at 27 Hz which normally (like in a BR) would be a lot. But the peak is very narrow and I do not know how this will play out in reality. I have seen some discussions on this, but thoughts?
- The response/group-delay spike at 100 Hz is eliminated with some stuffing in S12; the stuffing has very little effect everywhere else. What I attached below is the NON-stuffed response.
- I have seen some discussions on the presence of the subwoofer structure choking a throat if it is too narrow but don't have a feel for this. Thoughts? I could make the cabinet wider for example.
- I have DSP to deal with the subsonics and will taper it off (highpass) below 30 Hz or so. Upper end will roll off around 80-100Hz (lowpass).
- Corner placement
- This plot is with the driver near xmax, allowing for the subsonic cutoff (controlling the cone below 20 Hz). The sensitivity, if true, is really impressive, which is supposedly one of the great things about THs?
I'm including screen shots. The CAD model is PRELIMINARY and I know it needs more like bracing and driver access; it is just a sanity check on practicality and folding. It is roughly 89in x 20 in x 20 in.. The cabinet section lengths and cross sectional areas match the Hornresp data.
Thanks in advance
Last edited:
Literally 'at a glance' (1st thing I look at), it in theory will be very power limited due to a (Sd/S1) ~3.737:1 CR
OK I remember reading the rule of thumb is the SD/S1 ratio should be no more than 4, and I am close to that, but do not know the tradeoff. For the power limitation, is it due to air velocity or some such noise/nonlinearity?
One cycle at 27Hz is 37ms, considering the typical decay times of LF program material or room modes, 64ms is not much to worry about. The 100ms "ringing" at 100Hz may be worth damping out.
As narrow as the horn is at the transition, it would be worth considering the actual cross section of the driver in that area. In your design that could be as simple as straightening out this section a bit:
That straight section could also be used as the access point for the driver installation.
Also, the area at S2 is actually larger, you have VTC (volume of throat chamber) set to "0", the cone is concave at that point and should be considered in such a constricted cross section.
The CR (compression ratio) at S1 is not an issue for a tapped horn, it would be for a front loaded horn.
The CR at S2 when ATC is considered is actually about 2/1 at S2, not a worry.
About 6dB more sensitivity than a bass reflex is possible, and obviously way more down low than sealed.
The sensitivity is exaggerated by the 0.5*pi setting (Corner, wall & floor). To actually achieve that sensitivity would require those boundaries to be infinite, in a small room they are not. 2.0*pi (floor) will be closer to what you can expect.
Design looks like it should work OK after the above considerations.
Art
"The sensitivity is exaggerated by the 0.5*pi setting..."
Embarrassingly I did forget about that. I noticed the shape of the predicted response (in the bandpass region) was different for 2pi vs 0.5pi, so I set it where I expected the placement. Backing it off to 2pi the sensitivity is more like 125 dB. It is still a lot more than a large ported box (to your point).
Embarrassingly I did forget about that. I noticed the shape of the predicted response (in the bandpass region) was different for 2pi vs 0.5pi, so I set it where I expected the placement. Backing it off to 2pi the sensitivity is more like 125 dB. It is still a lot more than a large ported box (to your point).
OK one more question... I see many builds of THs with very little bracing and even relatively thin walls, like as low as 1/2" for something this size. I am thinking 3/4" plywood or OSB. Any suggestions?
I should draw up a braced plan and add that image here. TBD stand by
I should draw up a braced plan and add that image here. TBD stand by
Last edited:
The SS15 is a popular TH cabinet made of 1/2" plywood to be light weight for portable use. Jim (jbell, the SS15 designer) mentioned some 3/8" deflection of the SS15 front panel at high power.
That deflection could amount to almost half of the -7dB measured upper frequency power compression losses.
Plywood is stiffer than MDF, 3/4" plywood with bracing should be good.
Bracing down the center in the areas with red lines, with "t" cross bracing in the larger mouth area. The bracing does not need to be continuous, scrap cuts of 3-4" (75-100mm) could be used.
I round the leading and trailing edges of bracing to reduce turbulence, though probably not required.
That deflection could amount to almost half of the -7dB measured upper frequency power compression losses.
Plywood is stiffer than MDF, 3/4" plywood with bracing should be good.
Bracing down the center in the areas with red lines, with "t" cross bracing in the larger mouth area. The bracing does not need to be continuous, scrap cuts of 3-4" (75-100mm) could be used.
I round the leading and trailing edges of bracing to reduce turbulence, though probably not required.
Hmm, just noticed the driver specs are pure 'floobydust' 🙁 That said, are there any reliable measured specs or are the published ones good enough?
Thanks, yeah, I know better, just literally a 'knee-jerk' response while mulling over some FLH stub designs and missed the outright ludicrous Cms, Mmd driver specs altogether and see you did too. 🙁 We've looked at this samo-samo stuff so much we're down to 'seeing' what we expect to see.
Quite different response with published? specs.
Last edited:
Assuming this is measured response? From the attached posted by GM above...
Sd = 814.60
Bl = 14.09
Cms = 2.37E-04
Rms = 14.35
Mmd = 267.67
Le = 1.31
Re = 3.40
TH = 1
From Parts Express
�
Sd = 814.60
Bl = 14.09
Cms = 2.37E-04
Rms = 14.35
Mmd = 267.67
Le = 1.31
Re = 3.40
TH = 1
From Parts Express
�
Parts Express often has misprints, those TS specs are not close to what Dayton publishes for the UM15-22 15" Ultimax:
What about these though, straight from Dayton's web site? They do not match the ones posted above.
https://www.daytonaudio.com/product/1123/um15-22-15-ultimax-dvc-subwoofer-2-ohm-per-coil
https://www.daytonaudio.com/product/1123/um15-22-15-ultimax-dvc-subwoofer-2-ohm-per-coil
top is the updated , bottom is the original ,looks like VAS is cut in half too
(I grabbed GM‘s sim)
I usually use the driver model number and "PDF" in my searches:
The first "PDF" specs in post #13 has the lower FS and larger VAS, but according to Booger Weldz sims, the specs as in #11 have the same "saddle" (peak, dip, peak) response.
Oh well..
The first "PDF" specs in post #13 has the lower FS and larger VAS, but according to Booger Weldz sims, the specs as in #11 have the same "saddle" (peak, dip, peak) response.
Oh well..
Yeah, I cannot flatten that huge 10 dB saddle in the bandpass region after updating the TS parameters, at least not without going to extremes. Looking back at the OP, the TS values are same except in the original post the Mmd value was 5.99 gm instead of 272 gm. A big typo, and a huge difference obviously. Ugh... I thought I was on to something.
I think that’s a lack of motorforce and quite a high Qts so you’re just left with those ringing peaks at the two QW resonaces? What if you stretch one of those home theatre ported boxes for the ultimax into a tapered tapped pipe ? (I think they’re called ‘Marty’?
I was thinking in terms of a very strong motor with a cone made out of unobtanium operating in a vacuum. Ha!
I have tried so many TH configs with this particular driver that I don't think I can get there from here using a TH with this driver. If anyone knows differently I'm open to suggestions. I know I could get a better driver to build a TH but that defeats the purpose of using my existing drivers.
In the mean time I'm back to a large sealed or a BR as the leading candidates.
I have tried so many TH configs with this particular driver that I don't think I can get there from here using a TH with this driver. If anyone knows differently I'm open to suggestions. I know I could get a better driver to build a TH but that defeats the purpose of using my existing drivers.
In the mean time I'm back to a large sealed or a BR as the leading candidates.
Last edited:
