Why no Bass?

[PB2]

That is a nice cabinet and I'm not questioning your carpentry, but could there be a leak? I hope you didn't leave anything loose while working on it such as the input cup or something?

What does it do when you drive it hard at the tuning frequency? Is the port velocity high, cone motion low? You can check for leaks while doing this since the pressure is high at Fb.

Did you consider an active DSP XO like the Behringer? Then you'd have to triamp, or you could keep the mid to tweeter XO and bi-amp.
It would let you take out room peaks, and extend the bass with EQ but I agree that you should determine if something is wrong first.

[Loren42]

Originally Posted by PB2

That is a nice cabinet and I'm not questioning your carpentry, but could there be a leak? I hope you didn't leave anything loose while working on it such as the input cup or something?

What does it do when you drive it hard at the tuning frequency? Is the port velocity high, cone motion low? You can check for leaks while doing this since the pressure is high at Fb.

Did you consider an active DSP XO like the Behringer? Then you'd have to triamp, or you could keep the mid to tweeter XO and bi-amp.
It would let you take out room peaks, and extend the bass with EQ but I agree that you should determine if something is wrong first.

I don't think there is a leak, but it is worth checking for.

I had an active system (tri-amped) with it awhile back, but when I finished building my tube amp I went back to a passive system and upgraded the midrange driver and some cabinet fixes. One tube amp is enough when it costs $1100 in parts.

Yes, my plan is to get the speakers set up correctly first, then at some point we will rearrange the room and we can look into "room treatments" to tame issues with the room. After all that is done I can consider custom EQ to address remaining room and baffle step compensation.

I kind of like to keep things as simple as possible rather than add EQ as a bandaid, but I realize some amount of EQ will be required in the end.

[PB2]

You could try passive baffle step at line level just to make it easier, but I would agree with your suspicions that something is off given that the step is at such a low frequency. I wonder if you might have a strange cavity resonance going on there - have you measured the woofer near field just to see if there are any large deviations in the passband?

You may have multiple problems which is often the case with a new design, not enough baffle step,
a low bass issue, a few room effects, and perhaps a poorly chosen mid to tweeter crossover point.
Did you consider dropping this to 2 kHz?

You could use an inexpensive SS amp on the low end, refine the design with Digital EQ-XO,
which is very easy to tune, then move back to passive if you believe it is necessary. I'd like to
see 200W/ch on those woofers.

What are some of your test tracks if you don't mind my asking?

[Loren42]

Originally Posted by PB2

You could try passive baffle step at line level just to make it easier, but I would agree with your suspicions that something is off given that the step is at such a low frequency. I wonder if you might have a strange cavity resonance going on there - have you measured the woofer near field just to see if there are any large deviations in the passband?

You may have multiple problems which is often the case with a new design, not enough baffle step,
a low bass issue, a few room effects, and perhaps a poorly chosen mid to tweeter crossover point.
Did you consider dropping this to 2 kHz?

You could use an inexpensive SS amp on the low end, refine the design with Digital EQ-XO,
which is very easy to tune, then move back to passive if you believe it is necessary. I'd like to
see 200W/ch on those woofers.

What are some of your test tracks if you don't mind my asking?

Here is a composite of the port and the nearfield of the woofer very, very close to the dust cap:



The two plots are not correctly scaled in SPL, but you get the idea...

It is a 3-way, so the lower cross point is 400 Hz and the upper is at 3K Hz.

I think I have a lot of room effects, too.

I don't know what is really a good sample to listen t, I use Shawn Colvin's Polaroids, but most of my work right now is with sine wave sweeps. My intent is to get the ugly peaks and valleys tamed, then work on timber and sound quality once everything is matched up level wise.

[PB2]

OK, so the cone output looks reasonable no peaks in the response, notched at the
box resonance, but the port output must be attenuated since the summed output is shelved down.
Also, the peak frequency of the port is a bit high relative to
the notch in the driver output. Have you tried plugging one
port which will tune it much lower? Just to see what happens
to the total output? Might be high due to leakage from the woofer as you
stated earlier.

You might also try stuffing the upper chamber with damping material to reduce the cavity
resonances. There might be a top to bottom transmission line type resonance which will not
behave as a lumped box compliance as seen from the lower end of the box.

BTW - I really like this big beast of a speaker, nice work!

[Loren42]

Originally Posted by PB2

OK, so the cone output looks reasonable no peaks in the response, notched at the
box resonance, but the port output must be attenuated since the summed output is shelved down.
Also, the peak frequency of the port is a bit high relative to
the notch in the driver output. Have you tried plugging one
port which will tune it much lower? Just to see what happens
to the total output? Might be high due to leakage from the woofer as you
stated earlier.

You might also try stuffing the upper chamber with damping material to reduce the cavity
resonances. There might be a top to bottom transmission line type resonance which will not
behave as a lumped box compliance as seen from the lower end of the box.

BTW - I really like this big beast of a speaker, nice work!

"but the port output must be attenuated since the summed output is shelved down."

Well, that is the big question. How do I determine if this is true or false? The problem is idiopathic in that it is so hard to determine what is room artifact and what is deficiencies in the speaker system.

I can work the problem holistically, but it is better to know where the source of he problem lies, so I keep trying to unlock clues with the speaker first.

As for the possible cavity resonance, do you feel I should add additional layers of stuffing over the walls?

Thank you for your compliments. I like large speakers (as does my girlfriend) and the original design (shape) was one of my first speaker builds in high school when I did not have a clue as to what I was doing.

It still has the basic dimensions of the first build, but I sculpted the mid and tweeter section as you see it now.

This is a rendering of the first cabinet built many, many years ago. The woofer was a JBL D140 mated with some crapy tweeter made by AR.



The port was a slot at the cabinet bass and the two 2X3s that acted as legs also worked as a ducted port. None of this was done very scientifically and I don't remember where the tuning was for the cabinet. T/S parameters were only just getting developed back then!

[PB2]

Here is a link to Don Keele's original paper on the near field measurement method. Note that his measurement of a vented system shown in Figures 12 - 14 provides excellent agreement with theory. The system was similar to your situation in that the woofer was a 15" and the port was very close. The article also tells how to scale the port and cone output by area:
http://www.xlrtechs.com/dbkeele.com/...ld%20Paper.pdf

You should be able to use nearfield measurements properly scaled, and outdoor ground plane measurements for a half space measurement.
You would normally bury it in the ground for a perfect half space measurement but sitting on the ground should be good enough below 100 Hz.