Lab 12 Based Offset Driver - Mass Loaded - Transmission Line (OD-ML-TL) Design by Bj

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Why the T-TQWP vs Vented

I found this thread several months ago and was really intrigued by the design. So, I started modeling with other drivers and simplifying the design until basicly I ended up with a vented enclosure with a slot vent. Now, I'm really puzzled. Can someone explain why someone would want to go to the added complexity of a folded T-TQWP design versus a vented design, when Hornresp pretty much shows the same response in the same enclosure volume.

To show what I mean, I took Shadydave's current build and massaged it into a vented design that pretty much has the same SPL response and enclosure volume. What am I missing here?

-Vince
 

Attachments

  • Lab12 Vented.gif
    Lab12 Vented.gif
    7.4 KB · Views: 712
  • Lab 12 Hornresp details.gif
    Lab 12 Hornresp details.gif
    85.9 KB · Views: 709
  • Lab 12 Hornresp SPL.gif
    Lab 12 Hornresp SPL.gif
    64.7 KB · Views: 703
vnwhite,
The correct way to model a port loaded (BR) box is as shown below.(following your drawing)
.
As to your question, I am not sure.
They both seem to model quite similarly.
.
Also, the first box that I built was not really built following the plans and should probably not be used as an example.
That said, I ran the Original "final" HRSP model and the results are quite similar.
.
The BR box does have a phase issue right at the electrical impedance peak @ 36 Hz.
And, in my experience BR boxes are often quite "boomy" or show "single note boom".
.
The only thing that I can say, is the one that I built (incorrectly) sounds very good.
I have since built another following the original "final" plan, and am just waiting for some stuffing material so I can glue the last side onto the box.(and wire, sand, install feet, quarter round and paint it).
I plan on performing numerous back to back tests to see how much I screwed up my simplified version of the OD_ML_TL box.
.
I look forward to any comments by those who know more.

Dave
 

Attachments

  • HRSP Lab 12 BR.jpg
    HRSP Lab 12 BR.jpg
    145.1 KB · Views: 666
  • Lab 12 Schematic.jpg
    Lab 12 Schematic.jpg
    70 KB · Views: 632
Last edited:
Update

I have built 2 versions of this subwoofer.
I was confused by the Sketchup drawing of the box, specifically the 12.4” dimension.
The 12.4” dimension should probably read 12.908” as that is the effective length of the top of the port, but the original drawing has a flat on the port's rear top angled board to make up this difference.
This confusion ultimately caused me to build the first version with a shorter slot port.
.
I did not make the change without simulating it in Hornresponse and found that, at least in simulation, there seemed to be many small changes that would not make a difference.
I also added too much stuffing in the first design as I did not understand, I mean I screwed up.
There is a jog in the FR curve according to HRSP that changes drastically with small (1-2 cm) changes in port length but I am not so sure that the jog exists at all after stuffing in the line.
.
I just plugged the second box into my amp Monday night and plan on doing some sweeps next weekend after I break in the driver. I need to do impedance sweeps also.
I am generally very impressed with both boxes.
.
I have a pair of Infinity 1262’s in 73 liter sealed boxes driven to 420 Watts each that have a rising response (in room) down to 35 Hz before they start to roll off, but the Lab 2 subs add quite a bit to the mix that I would have never realized existed without them.
.
The Lab 2 subs can be driven to between 150 and 200 watts without their amplitude reaching more than about ½ of Xmax.
...This is not what HRSP predicts.
.
I have voltage limited to between 24 and 29 V depending on my mood and how much I will tolerate things falling off of shelves.
I broke a large picture window with the 2 Infinity boxes and one Lab 12 sub, both with strong (48 dB/octave) HP filters at 20 Hz.
.
The first Lab 2 sub I built shows an F3 of about 16.5 Hz, but I have a concrete house, both floors and walls and the 2 Infinity 1262’s are in full concrete corners.
.
I drew this box using Cadkey in numerous forms, but could not get my drawing to agree in path length with the “original final” sketchup drawing.
I spent allot of time trying to get the path length measured exactly and I am pretty sure that my drawings are correct.
The 2nd version I built is essentially the same as the “original” design except that I lowered the driver 8.6 mm and left out the 2nd piece of ply at the top end of the line.
This makes the lengths slightly different and I did this because HRSP says this will get rid of the jog at 80 Hz.
.
My initial impression concerning the longer port variation is it does sound better, but I have not done any measurements yet.
The longer port is hard to construct and this is one reason I did not build it at first.
This may have been a mistake.
.
I attached a drawing of the "Original final" design, as best as I could draw it.
The dimensions of path length do not agree with the sketchup versions.
.
I also attached a drawing of the 2nd box I built (BTN-A), where I tried to copy as exactly as possible the "Original final" design but with path dimensions that HRSP likes.
.
The hard part of the port is pictured as well as the 2nd box I built.
I decided to use 22.5 degree angles to join the top of the port parts, this both made it harder and easier.
.
I did not mount the driver in a counterbore on either version but I cut a 17 mm deep 45 degree angle on the back of the baffle in the 2nd version.
.
I do not know if it will make any difference as the driver has a large open area in the basket.
For now I included a cutlist of the box that I just built, but I strongly advise anyone to draw any box that they are building, at least on the side section of plywood to get everything to fit together.

I will post more info later
Dave

I also think it is a bad idea to write a post in Word and cut-N-paste it into the forum:mad:
 

Attachments

  • BTN-A cutlist.jpg
    BTN-A cutlist.jpg
    128 KB · Views: 232
  • Hard assembly.JPG
    Hard assembly.JPG
    170.8 KB · Views: 266
  • My 2nd box.JPG
    My 2nd box.JPG
    133.9 KB · Views: 321
  • 2nd fold I built.jpg
    2nd fold I built.jpg
    99.4 KB · Views: 299
  • BTN fold.jpg
    BTN fold.jpg
    103.3 KB · Views: 297
Last edited:
Can someone explain why someone would want to go to the added complexity of a folded T-TQWP design versus a vented design, when Hornresp pretty much shows the same response in the same enclosure volume.

-Vince

Lab12
132L BR, 23hz tune vs 127L T-TQWT, 24hz tune, 225wrms, 13mm xmax.
Both enclosures can be built with 7 boards.
 

Attachments

  • lab12ttqwt4.JPG
    lab12ttqwt4.JPG
    55.2 KB · Views: 162
  • lab12ttqwt3.JPG
    lab12ttqwt3.JPG
    60.9 KB · Views: 178
  • lab12ttqwt2.JPG
    lab12ttqwt2.JPG
    36.1 KB · Views: 134
  • lab12ttqwt1.JPG
    lab12ttqwt1.JPG
    68.7 KB · Views: 127
  • lab12br4.JPG
    lab12br4.JPG
    55.6 KB · Views: 181
  • lab12br3.JPG
    lab12br3.JPG
    58.4 KB · Views: 197
  • lab12br2.JPG
    lab12br2.JPG
    35.6 KB · Views: 136
  • lab12br1.JPG
    lab12br1.JPG
    69 KB · Views: 204
It turns out my basement leaks a little water...now that I have a perfect excuse due to one of my horns getting wet :( I'm going to build a pair of these (2nd fold version from post #85) with the Epic 12s to see what happens.

Nothing to lose, I'm going to have to toss my horn cabinets anyway, might as well try to salvage some of the wood and see what these sound like.

Edit: I realize now after writing what I wrote and seeing it next to my sig that I could have brought this on myself with whiskey and deadmau5.
 
Some more tests

So I took both OD_ML_TL boxes outside with a Crown xti 1002 amp and my laptop with REW an Art dual USB pre and an ECM 1000 mic.
I set up on the lawn with the microphone 1 meter from the front of the boxes and about an inch from the ground.
I was doing sweeps from 20 Hz to 300 Hz and usually doing 2 or 4 and letting them average.
.
All of my RTA data is unsmoothed (how do people get it to say that on top of the graph?).
.
The first graph shows my first simplified version box, in room. This was done last January. I also included similar outdoor sweeps of both boxes that I built on the same picture.
These were done with all of the same settings, just outdoors.
I realize that room response is a pretty big thing, but I had no idea how big it was, and how much low extension it can add.
.
These sweeps were all done with 512K length, 20 to 300 Hz (I did not realize this setting in REW would sweep the driver from 10 Hz to 600Hz, hence some confusion before concerning cone flapping and no acoustic output) and 2 or 4 sweeps.
The data is unsmoothed but the SPL level was calibrated a week before, so I imagine it is not all that far off but I can not guarantee it to be correct.
.
I have backed down to a 20 Hz, 24 dB/Octave HP filter for general listening and this does not cause any distress.
.
I run the limiter threshold at between 25.11 and 29.14 Volts and for the sweeps I set the attack to 10 mS and the release to 10 Seconds. (I use 10 mS and 200 mS for Music and HT)
.
In looking at both the phase information and a small difference in frequency response (not lower extension, but up near 80-100 Hz) there are some noticeable differences between the first, simplified version of the box that I built and the version that closely resembles the “final” version.
I ran the 2nd box at -8 dBFS and the first (simple) box at -10 dBFS settings on REW in the first picture otherwise the curves would have overlapped much more.
I was trying to do some impedance sweeps but I am getting some confusing results.
.
The Impedance peaks come out pretty close to what HRSP predicts, and there is a small difference between the two versions of the box.
However the difference is small with the first box (with the shorter vent) coming out at a peak of 33.65 Hz and the box that closely resembles the “final” design peaking at 34.42 Hz.
.
The result that confuses me the most is, if I do a “measure T/S parameters” with DATS it tells me that the Re of the driver is 12.6 ohms.
.
I just bought Virtins virtual instrument package and plan on testing for THD as soon as the rain stops. I would have done it Sunday morning but between the neighbors coming over (they live 1//2 a mile away) but I guess 8:00 AM on Sunday is a little early, and the rain that started a few hours later I had to call it a day.
.
One thing I will say to NWGrad, I am using a Crown xti 1002 because I liked the DSP option when I looked up the amps specs after you said you were going to drive your OD_ML_TL’s with one. You also said you might use a 12 dB/Octave BW HP filter.
.
I was testing with it set to 24 dB/Octave and 29.14 Volts limit (~ 200 W) with a starting frequency of 20 Hz, but then I did a sweep with the HP filter set to 12 dB/Octave and the amp immediately shut down with whatever protection scheme it has. This should have been less than 200 Watts rms as the measured impedance is always higher than the 4.29 ohms used in the HRSP simulation, usually quite a bit higher. I am unsure about what type of protection circuitry the amp has but it came right back on in 10 seconds or so when there was no longer an input signal.
.
A quick question, Does anyone ever use a Zoebel network on a sub?
.
I expected to see more difference between the simplified version that I built at first and the “real” design but maybe the phase information tells most of the story where the simple design has it’s 1st phase wrap at 43.5 Hz and in the 2nd box I built, this happens down lower at 32 Hz or so.
.
If I place both cabinets next to each other and set the limiter at 29 Volts in my concrete house these can definitely be used in anger.
.
All Hornresponse screens are from the input that shows the actual values for the 2nd box I built.
.
I expected to see a flatter response with the outdoors tests.
I am wondering if I did something wrong.
I used the USB pre directly into the amp and the amp had a 20 Hz, 24 dB/Octave filter, but no low pass filter setting at all.
Is this type of falling response normal for outdoor testing?


Dave

PS. DrDyna, the cutlist I included in my last post is for the 2nd box in post 85 and this box is almost exactly the design called "FINAL" by Bjorno and NWGrad.
The only difference is the line is 1.8 cm longer and the driver is 8 mm lower.
I did this because HRSP seemed to like it.
 

Attachments

  • 2nd box_34.42 Hz impedance peak_Re- 12.47 ohms.jpg
    2nd box_34.42 Hz impedance peak_Re- 12.47 ohms.jpg
    298.8 KB · Views: 96
  • 2nd box_34.42 Hz peak impedance.jpg
    2nd box_34.42 Hz peak impedance.jpg
    298.4 KB · Views: 101
  • Simple box, 33.65 Hz.jpg
    Simple box, 33.65 Hz.jpg
    287.8 KB · Views: 126
  • 2nd box_24 dB octave HP_29 V.jpg
    2nd box_24 dB octave HP_29 V.jpg
    89.1 KB · Views: 127
  • 1st box_24 dB octave_20 Hz HP_29 V.jpg
    1st box_24 dB octave_20 Hz HP_29 V.jpg
    87.8 KB · Views: 209
  • 1st Inroom and both outside_Red -8 dBFS_blue - 10 dBFS.jpg
    1st Inroom and both outside_Red -8 dBFS_blue - 10 dBFS.jpg
    113.3 KB · Views: 225
  • HRSP phase.jpg
    HRSP phase.jpg
    131.2 KB · Views: 82
  • HRSP Impedance.jpg
    HRSP Impedance.jpg
    107 KB · Views: 105
  • HRSP input.jpg
    HRSP input.jpg
    150.2 KB · Views: 111
Last edited:
shadydave,

thank you for posting these details. its very interesting the impedance behaviour heavily altered in this box. i think this was the case with the original LABhorn, as i remember the impedance on the box will be around 8ohms with the LAB12 6ohm.

in this case, you can save the 1ch from your amp by driving the 2 LAB12 with single channel with similar power and SPL, isnt it?

the room response is very flat, is that EQ'ed?

cheers
henry
 
Crazed and Confused

shadydave,

thank you for posting these details. its very interesting the impedance behaviour heavily altered in this box. i think this was the case with the original LABhorn, as i remember the impedance on the box will be around 8ohms with the LAB12 6ohm.

in this case, you can save the 1ch from your amp by driving the 2 LAB12 with single channel with similar power and SPL, isnt it?

the room response is very flat, is that EQ'ed?

cheers
henry

No EQ.
I was driving the amp with the output of the USB pre.
All sweeps posted since Jan. of this year are with a 20 Hz BW, 24 dB/octave HP filter and no low pass filter.
It is unsmoothed, or 1/48th octave which is what REW calls unsmoothed.
.
My house has an inroom hole at about 34 Hz and I found that moving the speaker around the room while maintaining the 1 M mic distance would change this hole.
I attached an example of 2 sweeps, one showing the hole prominently and the other with the hole somewhat reduced.
Also, I am rather sure of the SPL calibration on these sweeps.
I did the cal using an EBAY new (NOS) "Shack" SPL meter.
.
I think I may have hurt the USB port that I am using for DATS as I have seen several anomalous results from DATS lately, including the in box Re of over 12 ohms. Both Lab12 sub boxes read this way.
.
The thing that I am worried about is the falling low end in the outdoor response.
I thought it should look closer to the output of HRSP SPL chart.
.
Am I correct in thinking that the outdoor SPL sweeps should be allot flatter?
.
I attached 2 sweeps made indoors (in my concrete house) done in Jan where I moved the cabinet around the room. One shows the prominent hole at 34 Hz, the other shows a different looking "hole response".
It is definitely a room mode.
.
Also
Brian Steel says: Something is way off with those impedance measurements. There's no reason why they should look so ragged, unless the box is very poorly built and full of resonance modes, which I doubt.
I agree that something is wrong.
I do not think it is the box construction, 18 mm 13 ply BB glued with Urethane glue, clamped and screwed together using #6 square drive screws and stuffed with long hair wool (this time to Bjorno's specs).
.
I even pulled out some stuffing and the Speakon connector to use larger wire (was 16, went to 12 gauge appliance wire) when I saw the 12.4 ohms on the first box but this changed nothing.
.
I expected the Impedance curve to look similar to the HRSP chart with maybe the magnitude of the impedance off a little and the frequency of the peaks to be close to the HRSP prediction.
.
I am really stumped by the lower peak being so much lower in magnitude, so much wider and the general waviness of the trace.
.
I don't seem to have them with me but the bare Lab12 driver makes a single peaked impedance plot that looks just like what I would see on a spec sheet.
I need to do some more testing and figure out what is going on.

The only thing I can say is they, both, sound impressive

Dave

Time for more testing
PS..I assume that testing the impedance through a 4 pole Speakon connector, both sides using 12 gauge wire, and only about 12 ft. of it is OK..??
 

Attachments

  • Jan 12 sweeps.jpg
    Jan 12 sweeps.jpg
    117.3 KB · Views: 570
Last edited:
..I assume that testing the impedance through a 4 pole Speakon connector, both sides using 12 gauge wire, and only about 12 ft. of it is OK..??

It should be fine. A few times I've gotten similar crazy graphs from my WT3/DATS system and the only resolution was a PC restart and ensuring that I use the same USB port every time for measurements.

BTW, In my (albeit limited) experience, a lower impedance peak that's much lower and flatter than expected can be caused by too much damping material in the box. If the center frequency is shifted upwards, it could also be caused by a leak near the driver.
 
The thing that I am worried about is the falling low end in the outdoor response.
I thought it should look closer to the output of HRSP SPL chart.

Am I correct in thinking that the outdoor SPL sweeps should be allot flatter?

I expected the Impedance curve to look similar to the HRSP chart with maybe the magnitude of the impedance off a little and the frequency of the peaks to be close to the HRSP prediction.

I am really stumped by the lower peak being so much lower in magnitude, so much wider and the general waviness of the trace.
I think I may have hurt the USB port that I am using for DATS as I have seen several anomalous results from DATS lately, including the in box Re of over 12 ohms.
Though I have not done impedance sweeps on my cabinets, I am familiar with Hornresp frequency response peaks and dips looking larger than those measured, I would not worry much about the amplitude of the peaks since they are at the correct frequencies, and stuffing is not in the Hornresp program.

The in box Re of over 12 ohms would make sense as an average that includes the 18 and 38 ohm peaks and the 5.2 dip.

As far as your outdoor measurements, (not sure which they were) "cabin gain" in a small sealed room can make a big difference in LF response.
 

Attachments

  • Impedance.png
    Impedance.png
    152.3 KB · Views: 545
12gauge wire should be MORE than adequate for this setup. All connections are secure correct?

Driver bolted down firmly? Gasket in place?

10-24 Socket Head Cap Screws and I used the 1/8" X 3/8" speaker sealing foam to seal the drivers.
As far as secure I always wonder about spring loaded wire clamps, they sure are easy but...well I guess they must be proven to work well by now.
.
I used a new USB port and also noticed that I had the input level set at 0% instead of 1% as per the DATS instructions (not sure what that all means, 1% ?)
Anyway, it now makes a much cleaner looking impedance measurement.
 

Attachments

  • Lab 12 box_2_4-25-13.jpg
    Lab 12 box_2_4-25-13.jpg
    222.5 KB · Views: 519
I used a new USB port and also noticed that I had the input level set at 0% instead of 1% as per the DATS instructions (not sure what that all means, 1% ?)
Anyway, it now makes a much cleaner looking impedance measurement.
Dave,
Looks about 1% "cleaner" to me ;).

Looked OK before, and virtually identical, other than the 33.65 Hz peak being about 10% higher in the 1% test.


Art
 

Attachments

  • 1% cleaner.png
    1% cleaner.png
    123.7 KB · Views: 514
I finished one prototype yesterday and was looking forward to doing some outdoor measurements today, but its been raining, of course.

Stupid tablet, I guess its just gonna be sideways.
 

Attachments

  • 20130426_075151.jpg
    20130426_075151.jpg
    315 KB · Views: 522
Last edited:
Alright, so a few listening impressions since I've had a chance to spend some quality time with this enclosure. First, I've only completed the one made out of scrap left over from moisture damaged tapped horns, and I'm in a different room until repairs are made to the normal one. In case you haven't seen any other posts of mine, I'm using TC Sounds epic 12 rather than the Eminence driver. I've wired the single driver with each coil having one channel of an EP 2000.

This design seems to have more in common with my other T-TQWT than any of the others. I'm not sure if its due to the front mounted driver, but this seems way, way easier to integrate with mains than the tapped horn, or even the T-TQWT. It seems like it behaves more like a standard vented enclosure in that regard. It also seems like it has more "grip" on the air, but I'm not sure if that's in my imagination or not. It seems like it keeps pace with the music better with less...uh...(sorry for all the subjective junk) flop.

Where the tapped horn was just "blam boom pow bass here I am in your face" this is quite a bit more musical, and with some careful level matching seems to have disappeared behind my mains, until something needs it. "Yello - The Expert" means something completely different on this design.

While not quite as efficient as the TH, it's growing on me. I think the trade off is well worth it. Yeah, its not going to knock pictures off my neighbor's walls, but it can still grip your guts and give you a good shake when it needs to.

I'm looking forward to a sunny day tomorrow, where I can dolly this thing out in then front yard and try to get some proper measurements, but overall, I'm very happy with it and I'm most likely going to be building a proper pair out of BB Ply in the next couple of weeks.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.