Getting Bass from TD15M in a 5.5 ft³ Cabinet

For whatever reason, I decided to start a separate thread for this issue. I've got this ewave project going on with TD15M-4 woofers. The actual project (an ewave type) has been working out just fine, although very slowly because I'm using it more for learning that trying to get it done. Anyway, from the start bass has not really been there consistently, and this thread is intended to be focused on just that.

It could be baffle step, it could be other mysterious box problems, it could be a bass reflex design fail, it could be all of the above and more. Feedback appreciated! Here comes the data:

The cabinets are:
26.25"W x 30.25"H x 15.75"D externally. I don't have the internal dimensions now, but they are exactly 5.75 ft³ before any subtractions. Here is a drawing/photo of the front. The ports (yellow) are on the back, and are 5" diameter by 5" long. This results in a measured 40Hz tuning freq., which is way different than WinISD predicted. More on that in the next post.





The one under test is currently quite heavily-lined, as you can see. This seemed to have little-to-no effect under 100Hz when I did it, although this will soon get confirmed when I get the 2nd one assembled.




First a few full-range measurements on crossover. The orange/green line is the same measurement smoothed/unsmoothed on woofer axis measured a little too close (around 2') with a giant pile of insulation to minimize the first major reflection (it worked a little). Yellow is the >200Hz gated system response at 1m for comparison. Don't worry about the +/-1.5dB ripple in there over 1k - it will go away later.



I think this is pretty close to the real bass response I'm getting, as far as the overall trend (orange line). The 40Hz notch is confusing to me, since the tuning should give a minimum there in nearfield, but I thought I was far enough away to get to the port output represented? Maybe not, or maybe there's a room null happening at the same freq?

Next up, nearfield measurements while messing with ports.

Here are the published parameters for my woofer (it's the 4Ω version, if you didn't catch that). I have also attached the format WinISD uses, if you're into that. Change the file extension back to .wdr to use it.

Fs: 34.7Hz
Qms: 5.09
Vas: 312L
Cms: .3mm/N
Mms: 70g
Rms: 3kg/s
Xmax: 6mm
Sd: 855sqcm
Qes: .35
Re: 2.9ohm
Le: .15mH
Z: 4ohm
Bl: 11.25Tm
Pe: 500W
Qts: .33
1W SPL: 97.8dB
2.83V: 102.1dB

Here are the nearfield measurements. Sorry the key is a little hard to read, but check it out.

Just a thought, but like JBL are the wires suppose to be reversed? Try switching wires on woofer alone.

There's only one speaker here - polarity makes no difference for FR. For anyone that isn't familiar with this stuff, the nearfield measurements are supposed to show a dip at port tuning, i.e. the port output is hidden because it's overwhelmed by the level of the woofer output measured from 1cm. The dips in the 35-45Hz range shown above are the resulting port tunings.

However, what's happening here is that the ports (a) seem awfully far from simulated behavior and (b) don't really seem to be doing anything more than approximating the sealed response when evaluated by ear.

The subjective "most bass" from the above curves with music is the purple one (two 5" holes in the back, no port tube, insulation cleared from around them), and it should be that one since it should have the highest output 50-120Hz, BUT it's all very subtle. Basically, all the curves shown above sound almost the same, which is to say rather anemic. A few tracks will sound like they have bass (probably the ones where it happens to hit a room mode right), and actually deep bass, like 30Hz, sounds very present, if unimpressive. It's 40-80Hz oomph that's really missing here.

I don't want to insult your intelligence by starting at the wrong point so I'll ask up front, have you measured ported FR in the near field fully (Meaning adding in adjusted port output and all that jazz) before?

I would expect the best subjective response in the near field between all of those to be the naked holes as it's the least restrictive of the woofers at low frequencies. It sure sounds like you're just getting the bass from the front and nothing from the ports.

RayW said:

I don't want to insult your intelligence by starting at the wrong point so I'll ask up front, have you measured ported FR in the near field fully (Meaning adding in adjusted port output and all that jazz) before?

Click to expand...

I'm not sure what you mean? I have measured the port output at the exit plane of the a port (for the original 5"x5" alignment, not for all the stuff above). I have not done the sort of thing where you merge response plots to try to represent the real bass response of the system, if that's what you mean.

RayW said:

I would expect the best subjective response in the near field between all of those to be the naked holes as it's the least restrictive of the woofers at low frequencies. It sure sounds like you're just getting the bass from the front and nothing from the ports.

Click to expand...

Yes, that's the purple line like I said, and that's what it sounds like, but it also sounds only just a little different from sealed. Seems like the difference should be more dramatic. I keep wondering if the port location just isn't working. Recently I thought maybe it was the path from port to rear wall and back to woofer that was too long, but now I've got it up ~6" from the wall and it seems to behave just the same.

Do you think a little bit of baffle step plus a little bit of port ineffectiveness could be the most likely culprit?

Could you be over stuffed? Your response should be much lower and stronger than what it is. I don't know about the Lambda, but all new drives I've ever bought needed 100 hr =/- to break in and produce bass. Could this be the case here? I am mounting my lambda TD15X this weekend in 4 cu ft cabs and I hope to see better bass that what you are looking at. Needless to say, I'll be following this thread.

I mentioned that, but I don't blame you for missing it in my massive string of posts there. I certainly could be "overstuffed", but the only change I noticed from lining the walls is that the perceived midbass hump went away, and a dramatically dried-up midrange overall. I will A/B with an unlined/less-lined cabinet later. I need some cables I don't have to set that up.

Very curious to see what you find as well djnagle. Audiozaz's build has a large BR enclosure than mine in a small room, and he(?) says the bass sounds about right, but he hasn't measured very much yet.

dumptruck said:

I'm not sure what you mean? I have measured the port output at the exit plane of the a port (for the original 5"x5" alignment, not for all the stuff above). I have not done the sort of thing where you merge response plots to try to represent the real bass response of the system, if that's what you mean.

Click to expand...

Yeah, that's what I meant. You have to measure the woofer near field, then the port at the baffle plane, then adjust the port response (in dB) using the ratio between the total port area and the woofer area, and then add the responses. It's a huge PITA and I hate doing it but it might help tell the whole story here.

Yes, that's the purple line like I said, and that's what it sounds like, but it also sounds only just a little different from sealed. Seems like the difference should be more dramatic. I keep wondering if the port location just isn't working. Recently I thought maybe it was the path from port to rear wall and back to woofer that was too long, but now I've got it up ~6" from the wall and it seems to behave just the same.

Do you think a little bit of baffle step plus a little bit of port ineffectiveness could be the most likely culprit?

Click to expand...

It's certainly possible. Using an eyeball estimate on the baffle width and a quick guess at the baffle step starting point I would guess around 200Hz and the far field measurements seem to show something in that neighborhood.

It's kind of late now, I'll run the numbers through BB6 tomorrow and see if they match up with your WinISD sims or with the measurements or...?

Baffle step should be more like 120Hz, IIRC. The gating on the far field measurement is hiding what it really does - it's pretty much flat to near 100Hz just like the nearfield if you extend the gate longer. I'm going to bed as well, but I'd appreciate your thoughts any time.

Oh, PS, as far as break-in, I ran it 24 hours at about 1/2 xmax sweeping from 15Hz-30Hz in-cabinet, and it's probably had about 8 hours of playing music and test signals. I mean to imply exactly nothing - just stating the facts.

Hey DT,

Some of the stuff below might have already been stated by me a bunch of times in either my TD15M thread or yours, I'm getting mixed-up with so many posts/pages to them! Anyway, it'll be here all together.

From applying a ruler to the computer monitor upon your above front baffle layout, I estimate that it is about 24" to 25" wide, is this correct? That would give it a ~185Hz baffle step. That's one thing. I think it should be close enough to where room gain kicks-in to not absolutely necessitate BSC. My 22" baffle and relative closeness to back and side-walls mostly negates baffle-step.

WinISD is good and all, but I always find that the perceived bass response is less than that predicted. I think that since what WinISD shows is really just a transfer function based on an electrical filter analog, there are factors it fails to take into account, such as the loss of sensitivity as the wavelengths get larger than the radiator. I'm not certain at which point this factor should be considered, ½-wavelenght of the cone diameter, maybe ¼? I feel there is perhaps a 6dB/oct or so slope somewhere around this point that does not show up on the resulting WinISD calculations. Maybe someone more experienced than I will chime one this, it'd be educational for me. That's another thing.

Bass measurements and room modes - that's yet another thing, one which I hesitate postulating on due to lack of knowledge and experience there. I've gotten decent results measuring my Karlson's bass response with ground-plane mic placement. I've had somewhat dubious success with various other enclosures and my TD15M was giving me the worse time with gating. I'd like to be more help seeing how I pretty much have the same driver (other than being 8ohm) in a comparable enclosure but don't really know what to do in this case to get reliable data. My taller form-factor (internal golden ratio actually) + bottom vent probably push my enclosure towards being an MLTL and give it a bit of extra oomph around 90Hz, using the vertical ¼-wave mode. My small room + low tuning (~32hz) make it all work IMO and avoid boominess. I'd call the resulting sound pretty tight. Others may find it too dry.

I'll possibly be looking to upgrade my amplification with a nice DIY tube amp build and will then have to dive back into measurement hell to design a passive XO. That might not be for some months still.

It might sound silly, but I'm starting to get a decent handle on how a speaker sounds based on hearing pink noise through it. I think it can be most revealing of "material tone". Or maybe I just had too much snake whiskey. :yuck:

Anyway, good luck on your continuing project here, I'll be following this here thread.

az

When I run this through WinISD, I get just about what your measurements are showing. WinISD shows that each port tube should be about 8 to 10" long FYI. If you mess with the port tube length, you can boost the lower end just before roll off. There are a couple guys around here that use the "M" and need subs. That is why I ordered the "X" and that is still questionable whether I'll need subs for organ music.


P.S., I used 5 cu ft due to the space all the bracing, the woofer itself, and the horn/CD combo take up.

TD15M won't play sub-40Hz at rated sensitivity, that's for sure. Maybe unless one builds a truly humongous cab, but we're not quite there with our 4-6cu.f soap-boxes. Mine goes into the 30ies in a manner I'm satisfied with.



The above is for a 5.5cu.f box, mine is actually 5cu.f, and the tuning a few Hz lower, but the transfer function is not too far-off. I'm above the f10 point for most of the 30ies. WinISD does now show the quarter-wave action that's going on around 90Hz though; but this is not just an amplitude thing. I would tune it higher were I to use a subwoofer.

I remember reading something about vents located right at the floor boundary, such as mine, having their efficiency increased. Vents on the rear of the cab may be a bit more polite-sounding too.

I'm sure the great GM would not mind coming up with an old-school MLTL for TD15M. Vb for this might gravitate around 9-10cu.f though. This might be about the best bass possible for this driver.

az

I gotta agree, the TD15M isn't known to be a bass monster. With higher volumes, it does start coming out. You have to remember, you're playing with a pro driver that was designed to play loud to get the listed specs..

You're seeing why I gave in and built subs for lower volumes... Once I start getting say around 110, 115 db, then the bass really starts working like it should..

I do gotta wonder about that dip around 40hz... that's kinda strange..

similost said:

I gotta agree, the TD15M isn't known to be a bass monster. With higher volumes, it does start coming out. You have to remember, you're playing with a pro driver that was designed to play loud to get the listed specs..

You're seeing why I gave in and built subs for lower volumes... Once I start getting say around 110, 115 db, then the bass really starts working like it should..

I do gotta wonder about that dip around 40hz... that's kinda strange..

Click to expand...

While I don't usually get to play at such levels, I agree that the woofer "wakes-up" with some more drive.

re: 40Hz dip, even a couple of feet away you'll measure the tuning dip, where the woofer cone barely moves and most all output is from the port. DT's measurement only shows it a few dB's deep. The notch will be much deeper when doing a near-cone trace.

az

Djnagle, I think I'm getting closer to -6dB at 50Hz, like shown in the first measurement. I'm not trying to get by without subs here guys, just trying to get a full, emphatic sound above the primary subwoofer range.

eh... good luck with that.... I do believe the M stand for MID... these are more a mid bass woofer than anything...

If you wanted deeper bass, the H and S are more meant for that... Even at around 40hz, they can be on the anemic side unless you're pushing a bunch of power to them.

Simi, I think you may have misread my last post.

audiozaz said:

re: 40Hz dip, even a couple of feet away you'll measure the tuning dip, where the woofer cone barely moves and most all output is from the port. DT's measurement only shows it a few dB's deep. The notch will be much deeper when doing a near-cone trace.

Click to expand...

My measurements above are 1cm from the cone, except the first one, but both show around -15dB at tuning. The noise floor is relatively high around there and may limit that notch effect somewhat in both/either measurement.