The speaker project (dialup warning)

JoshK · « **on:** 23 Jun 2009, 02:43 pm »

Well I suck at coming up with clever names like many speaker builders. So for now its just "the speaker project".

I decided to create this thread to organize the various posts I've made on the speaker project I am working on. I hope to quasi-document the process and share data, etc.

I've been asked a few times if I had a thread about the project, so here it is....

Background:
I have been enamored with large format high efficiency drivers ever since visiting Dr. Geddes and hearing his setup with his Summas. I recognize that we all have different priorities in what we target in our hi-fi systems but for me 6.5" 2ways have never satisfied my sensibilities. The lack of effort, lack of compression and raw inherent dynamics provided by "pro audio" drivers is just unmatched in any hi-fi system I've heard to date. So this wetted my appetite and eventually made me abandon my previous speaker project to persue this current one with full abandon.

I had initial sketches in my mind of what I was after but they were more puzzle pieces that needed to be carefully balanced. I had my first rough ideas about what I had planned to persue about the time that Lynn Olson's "Beyond the Ariel" thread showed up on diyaudio. There is plenty of good overlap of persuits with some key departures, but the thread has been invaluable brainstorming material.

A few goals for my project are:
1) real sensitivity (db/w/m) in excess of 95db over the bandwidth of the speaker. (some minor departure in the first two octaves is tolerable).
2) a direct focus on the power response of the system. Constant directivity and/or controlled directivity will be employed.
3) unconstrained headroom for sane domestic listening. I.e. peaks over, say 105db, without noticeable compression or distortion.
4) some modularity to the setup to experiment with different combinations of bass and/or MT setups. (DEQX will be the tool to experiment)
5) optional--integration of open baffle/dipole concepts. I say this is optional, but it has inspired a lot of thinking.

JoshK · « **Reply #1 on:** 23 Jun 2009, 02:43 pm »

The realization:
After much deliberation the current project entails the following
1) 12" Geddes waveguide and foam using B&C DE250 compression drivers from ~1Khz up
2) AE Lambda TD12M Apollo 12" drivers for midrange duties target somewhere in the 200-300hz up to 1khz
3) AE Lambda TD15X + 2 AE PR15-1400g in 4-5 cuft enclosure for ~50hz-300hz.

1)&2) are married together on a baffle that has been displayed in various threads here. Essentially, cosmetics are all that is left unless current direction is changed grossly.

Construction of 3) is due to start, hopefully, very soon. Higher priority bathroom remodeling is coming to fruition, which leaves some of my free summer weekends available to building.

I idea thus far is to use 1)&2) open baffle above the PR tuned bass bin. There were a number of threads on diyaudio that have inspired me to experiment with "muffling" the back wave of the mid to acheive a nice steady evolution of power response from the CD of the waveguide down to the omni dispertion of the bassbin.

I am hoping I am able to get up the learning curve fast enough on measuring to determine if I am successful in the above target. Certainly it would have been far easier to have used a TD15M all the way up to 1Khz to match the directivity of the WG at Fc and meet my other goals but it would have limited the LF reach and not married the concept of OB and boxless mids.

The MT baffle will not be permanently fastened to the bassbin, which will give me to modularity to experiment with, say, open baffle bass if I wish to in the future. I can also "play" with some of the other ideas I have down the road and still use portions of this speaker.

Picture of the MT baffle in its current state (i.e. still in need of further bondo/sanding/primer but ready to take measurements if desired).

JoshK · « **Reply #2 on:** 23 Jun 2009, 02:43 pm »

The bassbin alignment

I had considered OB "alignment" for this project and initially that was the goal, but it proved to be much more complicated that I'd hoped. The desire to match the sensitivity and dynamic capability of the MT with OB bass is the kicker. For now, that has given out to practicality and the fact that I have multiple TD15Xs on hand. The TD15X isn't well suited to OB without lots of EQ and then you have given up all its sensitivity and inherent advantageous qualities.

I started playing with modeling the TD15X in a sealed enclosure. I generally liked what I saw. This is not a subwoofer, its not designed nor intended to reach 15hz f3, its a spectacular midbass woofer. I plan to use it as it is intended.

I looked at ported alignments. JohnJ of AES suggested a 5cuft enclosure with a 3" x 16" x 26" slot port. I did a lot of modeling of this alignment and variations. Its good, but it has tradeoffs. Tradeoffs include higher group delay, difficulty in reducing chuffing from port velocity, difficulty in summing with subwoofers.

I plan to use dedicated subwoofers so first or even second octave extention is not necessary.

Then someone on AA mentioned the "God of Thunder". This was the bassbin used in the Unity kit project that Lambda, in its former incarnation. It includes 1 TD15X in a 4cuft box with 2 1400g 15" PRs. I modeled this alignment and there is a lot to like.

The FR of said alignment using Unibox.

The FR looks a lot like a sealed box but with 10hz lower F3. The rolloff is near the same as a sealed box down to Fb in the teens. The step response is much closer to the sealed box than the ported box, which implies group delay is much closer to the sealed than ported. The excursion is however quite reduced down low versus a sealed box. Very happy tradeoffs.

JoshK · « **Reply #3 on:** 23 Jun 2009, 04:00 pm »

Comparison of FR for the various alignments. Note, the bandpass alignment is not optimized, its a byproduct of Unibox that you can't select only a couple alignments for comparison.

Comparison of the Step Response

Comparison of the cone excursion

Another advantage to this PR alignment is that, according to Unibox at least, the tuning is not heavily dependent on the box volume. This is a key point. Lynn Olson has pointed out that many ported alignments are heavily dependent on the volume and tuning which makes the tuning optimal at only one SPL. Essentially the tuning changes with dynamics or SPL. AFAICT, this alignment doesn't have those problems.

nullspace · « **Reply #4 on:** 24 Jun 2009, 01:04 am »

Hi Josh,

Looks like a good start. Best of luck with the project, and I'll be looking forward to updates.

Any first inclinations on how you'll transition/crossover from the midbass to the bassbins?

Regards,
John

JohnR · « **Reply #5 on:** 24 Jun 2009, 01:47 am »

The PR alignment seems quite heavily damped. What size/Q is the sealed alignment you have there?

JoshK · « **Reply #6 on:** 24 Jun 2009, 02:07 am »

Indeed the PRs are quite dampened. That is what makes it more like a sealed enclosure. Since I am using subs I am not willing to trade transient response, group delay for extra extension. The PRs look like they buy you 10hz and lessen excursion (which lowers distortion).

To answer your question, the Q of the sealed is .56 and the box is 4cuft.

The crossover between the bassbins and MT will be active using the DEQX. The slope is up in the air, even the Fc isn't nailed down, just a general idea that it will be in the 200-300hz range as that is where the the TD12M is at it's best and the TD15X is happy up till just before 500hz where it has its first resonance due to surround.

JohnR · « **Reply #7 on:** 24 Jun 2009, 10:38 am »

Ah, I get it now. Do you have subs picked out? Love the concept re headroom and dynamics.

JoshK · « **Reply #8 on:** 25 Jun 2009, 06:14 pm »

FYI, in case you didn't see it.

http://sites.google.com/site/drivervault/driver-measurements/

A vault of measurements with a slant towards more pro-ish driver, like in my project. My TD12M and TD15X are both included.

The lambda drivers definitely do as they claim as far as impedance is concerned. Look at the Le(X) graphs. Virtually no change to impedance vs position in the passband. Super flat overall Z too. In the non-linear distortion measurements they hold their own. In my eyes the TD12M dominates in the group listed and the TD15X ranks high in its group. Looks like I didn't go wrong.

These charts certainly help in my choice of bass/mid crossover point. As I mentioned earlier before the measurements were available the TD15X is good till 500hz. The measurements confirm this and suggest its not even bad all the way up till >1khz, but not quite as good as it is below 500hz.

The TD12M is superb above 100hz. I don't know how large his test baffle is, and this will have a lot to do with where I pick my Fc in my OB mid.

Its great to have such polar measurements too! That helps a great deal with mid/treble Fc. This confirms the directivity near Fc, which based on this, should be in the 1k-1.5khz area.

JoshK · « **Reply #9 on:** 12 Sep 2009, 04:08 pm »

I had forgotten about this thread. It seems most logical to update this thread in a quasi-build thread manner. First I'll borrow some of my comments from the "what equipment are you working on" thread for logical cohesion.

Quote

Making the baffles for my bass bins. Whow buddy, they are big. Somehow 30" x 20" x 22" doesn't sound that big.

Quote

Started gluing up my bins today. One is all clamped up. Biscuit jointers rock! It was SO much easier to put a big box together while keeping everything mutual orthogonal with the biscuit joints. Not to mention they make the joints stronger.

The baffle hasn't been glued on, that will be after the driver hole is machined and the oak cross braces have been added. Babies are going to be HEAVY! Glad I used 1.5" baltic birch on all sides but the baffle (2.25" BB/MDF sandwich) as the BB is lighter, although still heavy.

Wife commented that she saw my boxes and they are huge. Her reaction was better than I thought it would be. It wasn't a "no well in hell."

JoshK · « **Reply #10 on:** 12 Sep 2009, 04:10 pm »

Clamping the bins up.

JoshK · « **Reply #11 on:** 12 Sep 2009, 04:12 pm »

Dry fitting the baffle, threw in a woofer to see how it is gonna look.

JoshK · « **Reply #12 on:** 12 Sep 2009, 04:14 pm »

I nick-named the above bass bins, Donner. Thor has been done before. I received the PRs and cut the holes today. I bought the 2" OC 703 to line the cabinets with. Inching closer, little by little.

JoshK · « **Reply #13 on:** 12 Sep 2009, 04:14 pm »

To date, finishing is my least favorite part of DIY. Its probably cause I am not yet good at it and therefore it doesn't go smoothly. I am taking my dear sweet time with my MT portion, learning as I go. I picked up some single tube glazing putty yesterday, thanks to Dr. Geddes's thread.

I put in braces and lined the bass bins with the 703 and made up internal wiring harnesses today. All this takes much more time than it looks. Of course, I fiddle about. Oh yeah, and I drilled holes for all the speakers to mount. I countersunk the hurricane nuts and chamfered the back. Drill presses are nice for adding these nice touches. I also countersunk the speakon connector. I am going to sink hurricane nuts for the feet, so that they can be removed/replaced.

I went a little nutty professor with my bracing. Its all done in golden ratio bi-sections. I braced all three axes and all three are mutally golden ratio bisectors of the panels in both dimensions. Basically, you draw a diagonal on the wall you join to. Measure this diagonal, call it D, divide by 2.618 (1+phi). Measure and mark this distance on the diagonal. This point bisects both sides of the panel at golden ratios.

JoshK · « **Reply #14 on:** 12 Sep 2009, 04:22 pm »

Some progress pictures...

JoshK · « **Reply #15 on:** 12 Sep 2009, 04:30 pm »

So if the bass bins are named Donner, german variant of Thor, God of Thunder, then should the MT be named Blitzen, german for lightening?

JoshK · « **Reply #16 on:** 13 Sep 2009, 05:45 am »

Glued one baffle on tonight. Not enough clamps to do both at once.

Then I'm onto finishing...my favorite part.

poseidonsvoice · « **Reply #17 on:** 13 Sep 2009, 06:38 am »

Josh,

Given that Mike Galusha, you and I have very parallel interests regarding wave guides, narrow controlled directivity and high efficiency loudspeakers, I thought this link from William Cowan, very accurately describes at least my requirements for a loudspeaker. I bet you will agree with these observations/conclusions as well.

Best,
Anand.

JoshK · « **Reply #18 on:** 13 Sep 2009, 02:53 pm »

I share a lot of his ideals, but maybe not 100%. I believe that speakers and room are 95% of the equation. However, I still like my tube amps, but I am not dogmatic about it. I've heard and owned SS amps I like too.

JoshK · « **Reply #19 on:** 21 Sep 2009, 01:09 am »

Bass bins are ready to be finished. I did the spot glazing of the MT baffles, so those are ready for last primer coats and painting. Its coming together.

I decided to build the MTs just like the regular boxes without the back baffle. I will shoot for a cardiod mid via stuffing the rear wave. The idea is to provide a gradual transition in the power response from the CD waveguide down to the omni-pole woofer. There was a thread on diyAudio that talked about this and a cardiod mid was the conceived solution.

If I can find my memory card I'll my up pics of the bass bin pre-finishing step.