tqwtl predictions

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Thanks GM,

Your model requires a 59L box, which is possible, but I'm looking for a cabinet that is roughly 110cm x 20cm x 30cm maximum external dimensions, giving me about 40L to play with.

The question is,
1) How accurate is Hornsrep?
2) If a satisfactory response is simulated, are there any other real world factors to be considered? I have heard:
3) Is Hornsrep only simulating the horn response or the horn and the drivers rear radiation, or is it the front as well? The response drops down as the frequency increases above 400hz in my simulations, this is obviously a problem if it includes sound output from the front.

I have heard several rules for making a horn, but I don't know how true they are?

1) That the horn length must be 1/3 of the Fs of the driver
2) The areas of each end of the horn must add up to the length of the horn (approximately)
3) The driver must tap at 1/3 of the way down the horn for a flatter response.

Basically, is this actually going to work? And is it a good idea?

You're welcome!

1] Plenty close enough when used correctly.

2] Yes, like can one build it in the desired space, etc..

3] It shows both horn and driver, though only up to a certain point where either the horn rolls it off or where the driver's no longer a point source. It's academic though for a tapped alignment as the phase response limits the XO point to at/just above the tap point or ~250 Hz at most according to HR. For a higher XO point you'll need a regular TQWT.

1] Never been exposed to this rule-of-thumb and makes no sense to me as presented, but a horn has to be a 1/2 WL long to properly load, though doesn't have to be to Fs unless that's the design goal.

2] This makes no sense to me either as presented, but the throat area [St] is a function of the horn's desired HF response and the mouth area [Sm] plus any local boundary loading sets how low it will load.

3] Well, it needs to be at an acoustically odd harmonic, so typically 1/5th, 1/3rd, 2/5ths, though due to the flare factor [T or M depending on the source] and frequency ['x' Hz pipe fundamental] these points are physically a little bit lower in reality, but for most folks just dividing by the fractions is plenty close enough.

Yes and mostly no as I'm not sure if you understand how hard this will be to accurately fold up and will probably have some dead space that will need filling to keep from resonating; whereas mine will be much easier with just plain boards and if it can't be made to fit the space, then use the wizard to make it smaller, but do keep the driver at the 13.94 cm point at each end.

Regardless, you didn't respond to whether or not you want a TTQWT or regular TQWT, but judging by your Sketch, you want a regular one, so your sim is no good as is or even if you change the driver type from [TH] to [OD], though if you change mine you'll have a good one to start with.

GM
 
Not going bigger will compromise the sound, and i doubt you can fit a proper TL into that box.

dave

I am willing to accept efficiency loss from a TL cabinet format at the bottom end (below 80hz) as a sub will be used, and EQ will also be available. I would like the speaker to sound reasonable without the sub turned on though.

I don't seam to get significantly more bottom end even when increasing the size of s1 to s4 by a factor of 2 or more, mostly efficiency.
 
whereas mine will be much easier with just plain boards and if it can't be made to fit the space, then use the wizard to make it smaller, but do keep the driver at the 13.94 cm point at each end.

FWIW, I shrunk a TQWT [no tap] version of mine to ~37.5 L and with a stuffing adjustment it appears to meet your needs.

GM
 

Attachments

  • TD10M40hzTQWT.txt
    973 bytes · Views: 30
Then what is the point of doing a TL?

dave

The reason for doing a TL is subjectively, I find them to sound good, and I can't help but think that standard ported enclosures must colour the sound as they keep most of the back-wave from the driver in. I have no measurements to support that this is audible, but I just can't see how large proportions of the sound doesn't bounce around inside the enclosure and simply emerge back into the room by passing back through the speaker diaphragm at a later interval.

Damping of course will help convert some of this into heat, but it just doesn't seem right to me. The more efficient the cabinet design makes the speaker, the less energy will be kicking around in the cabinet I suppose. On top of this, the larger the cabinet in proportion to the driver, the more area to absorb the backwave.
 
Last edited:
3] It shows both horn and driver, though only up to a certain point where either the horn rolls it off or where the driver's no longer a point source. It's academic though for a tapped alignment as the phase response limits the XO point to at/just above the tap point or ~250 Hz at most according to HR. For a higher XO point you'll need a regular TQWT.
I was not aware of the max frequency limitations of a TH. Simulations all show the output dropping at higher frequencies, hence my question. This explains it. thanks!
2] This makes no sense to me either as presented, but the throat area [St] is a function of the horn's desired HF response and the mouth area [Sm] plus any local boundary loading sets how low it will load.
I definitely read somewhere that the compression ratio (ratio of s1 to s4 should ideally be kept to circa 3:1. I was assuming this was to do with air velocity in the same way as port noise becomes a problem with a undersized port, or with pressure on the driver diaphragm itself from the back-wave.
I'm not sure if you understand how hard this will be to accurately fold up and will probably have some dead space that will need filling to keep from resonating; whereas mine will be much easier with just plain boards and if it can't be made to fit the space, then use the wizard to make it smaller, but do keep the driver at the 13.94 cm point at each end.
I was going on the premise that if the horn angle was constant, (link s1 to s4) that a single fold would be needed, and as the angle doesn't change, this would be easy (1 piece of wood down the middle) I can accept a moderate increase in the depth of the cabinet if necessary.
Regardless, you didn't respond to whether or not you want a TTQWT or regular TQWT, but judging by your Sketch, you want a regular one, so your sim is no good as is or even if you change the driver type from [TH] to [OD], though if you change mine you'll have a good one to start with.

Considering your earlier point about using a TH only up to about 250hz, I will need to use this driver up to circa 500-700hz crossover (JMLC 350hz horn above) So I would need a TL or MLTL then. I am unsure of the advantages/disadvantages of a MLTL though to be honest.

One of the reasons I chose the TD10M is because I wish to have the option to change over the tweeter for various alternatives, so wanted a driver that could play higher if needed.

I appreciate the major drawback to the design is the distance between the tweeter and the bass driver, but this is partly why the design has the TD10M at the top of the cabinet, so I could swap out the horn for alternatives at a later date and see if I felt decreasing the C to C distance to the tweeter was worth the trade off of loosing that up front presentation of a horn.

While I am talking about it, I assume there is no alternative layout I have missed that would reduce the c to c distance without significant drawbacks? (cant look ugly though). I thought about angling the driver back, so the magnet is much closer to the tweeter, but this just seams to cause complications with reflective surfaces in front of the driver.

I could move the driver so the magnet is very close to the tweeter magnet, and put a mini horn mouth on the front of the driver as well, but I have no idea how to simulate this, and it would probably have awful off axis behaviour unless the cabinet was very wide.
 
Good enough, but the vent has to be inputted at the bottom, though the sim calculates it as both the driver and vent is the same distance from the listening position [LP], which adds a notch in the response.

To view the response with it shifted up at an odd harmonic where some of us like to put them to deal with the notch, then you have to put in the angular offset using the OUTPUT/HORN/COMBINED in the TOOLS menu.

Attached for reference:

GM
 

Attachments

  • 25GDN-5_MLTL.txt
    965 bytes · Views: 26
Hi all,

So this design has developed into a 3 way with a JBL 2206 as the bass driver, and the TD10M moved to mid range duty.

I plan to keep the form factor (shown on the previous page) but make the tube in the middle sealed.

The 2206 will be placed side facing in the bottom cabinet, and will likely be just ported as per JBL's recommendations as it only needs about 50L

I would like to run the mid driver down to about 200hz or so. I hope this is low enough to not become an issue with beaming/directivity.

The idea is to keep the distortion low for the mid by offloading the higher excursion needs of bass to the 2206h.

I don't know where the crossover frequency will be, but as it will be digital active, this shouldn't be too much of a problem.

Aesthetically, I don't want the edge around the mid driver to be too thick as per Planet10 suggests in post 18. I will if I need too though. Can anyone provide any evidence to support Planet10s suggestion that having the box too close to the size of the driver is detrimental to the sound? And how close can I go?
 
frugal-phile™
Joined 2001
Paid Member
We went thru this when devoping the 1st mFonkens.

This was too narrow:

mFonken-protoTypes.jpg

This turned out OK (3rd try) but still not quite as “open” as the full-size Fonken:

bluegreen-mFonken.jpg

dave
 
Looking at the image, I'm guessing you have about an inch of space either side of the driver??

As the TD10M is now being used in the midrange only, and I would expect it to stay under 1mm xmax in use over 200hz, does this change anything?

I know Troels did some tests years ago on chamfering the driver hole on the back face, and he came to the conclusion that this reduced the 'cupping' sound.

However, this might actually be a desirable effect as its crossing to a horn with a narrower dispersion pattern....

I've asked for a quote for a tube to be made with a 18mm wall thickness at 330mm OD.

Another possible change to the design is to add a space to put interchangeable baffle covers on the mid driver. The aim is to have the option to play with pattern control on the mid a la JBL eon 615. I don't know if this raises distortion though.

YouTube
 
Last edited:
frugal-phile™
Joined 2001
Paid Member
Looking at the image, I'm guessing you have about an inch of space either side of the driver??

Can’t findI had to dig deep to find the original drawing (a decade old document):

Prototype

attachment.php


Final:

attachment.php


… chamfering the driver hole on the back face, and he came to the conclusion that this reduced the 'cupping' sound.

Something we always do.

dave
 

Attachments

  • mFonkenProto-edgeSpace.png
    mFonkenProto-edgeSpace.png
    6.4 KB · Views: 472
  • mFonken-edgeSpace.png
    mFonken-edgeSpace.png
    8.6 KB · Views: 491
I have found a manufacturer of card tubes that can make me a tube to specification. I have suggested a tube @ 360mm wide OD, with a 18mm wall thickness. This gives me 1.5 inches between the widest part of the frame of the driver (not the cone) and the outer wall of the enclosure. I hope this is enough not to colour the sound!!
 
I am concerned about how the sound will be affected if I go sealed with the mid driver, so plan to build the speaker in 2 stages.

The first is to run the speaker as a test rig with a removable back on the mid tube, so I can try sealed mid and ported mid/bass configurations.

And the second is to add the 2206h to the base cab and run the TD10m as mid sealed.

I should have enough room to make a semi B+W style enclosure tapering shape inside the mid tube when in the sealed format.

This will be interesting to play with, but also should give me some confidence that I'm making better choices.
 
An update:

I currently have the test rig speaker running with a digital active crossover (in JRiver)

I spent the last 2 days fiddling with crossover setting and levels. I wanted to see where I got too without using measuring equipment, just personal preference.

The TD10M is simply in a well damped (10 large towels, haha) 2m long open ended tube. Testing aperiodic, sealed and ported is very easy as I can just move a single MDF disc up or down the tube to change the volume of the box.

At the moment, the tube is open at the back (transmission line effectively) and just 50-60% stuffed. Moving the EQ bars up and down, its clear a hump at about 120/150hz is present, but not too bad.

interestingly crossovers ended up 1st order?! at:
350hz low pass to the AE TD10M
750HZ high pass to the JBL 2452.

I was surprised how much different orders of filter sounded. Perhaps something else is going on. I can hear what I assume is beaming when moving the bass driver crossover up further. Generally I am encouraged by what I hear. I don't have the 'SL' or aluminium diaphrams in the 2452, just the titanium ones, but they really don't sound bad at all! I was concerned it would have some 'horn honk', but the only time I have heard anything that I would imaging was like that it has turned out to be the audio track (confirmed by hooking up other speakers and headphones.
The TD10m delivers plenty of bass, However, my only criticism so far is I don't perceive it to be as dynamic as the JBL 2452 1.5 inch compression driver. Having said that, its still got WAY more of a punch than the 6 inch bass drivers I have been listening too the last few years! If this is improved further by going to a larger diameter bass driver, I can see why so many end up with massive 16 inch altecs, Addictive.

I plan to set up a pair of JBL 2206h drivers I have for running below 200hz and see what this does to the kick.

Tomorrow I will take some measurements to see how bad my ear tuned setup is!
 
Last edited:
The following are measurements of my JMLC 350hz horn with 3D printed throat angle adapter for the JBL 2452H.

Image 1 is unsmoothed on axis at 1m measurement.
Image 2 is with 1/48 smoothing.
The third image is (spaced) the 1m measurements at:
0 deg
10 deg
20 deg
30 deg
40 deg

JMLC 2452H 1m on axis.jpg - Google Drive
JMLC 2452H 1m 0 deg 1-48 smoothed.jpg - Google Drive
JMLC 2452H 1m all.jpg - Google Drive

The next 2 are of the AE TD10M. (spaced and unspaced) They are simply in a 2m open ended tube, not a proper tuned box.

TD10M 1m All.jpg - Google Drive
TD10M 1m All unspaced.jpg - Google Drive
 
Hi all,

I have been using a pair of Hypex Fusion plate amps with a L-pad to take 20db off of the horn driver for noise free sound. This is currently attached to a pair of JM Labs 816s speakers as the bass driver/stand just to play around with and practice with the software etc while I wait for the wood to arrive for building the cabs.

Shared album - lance bartlett - Google Photos

I have set up a rig for testing the drivers and playing with the digital EQ in the Fusion amps.

It should work out well that the plate amps are so slim as they will fit neatly on the narrow back panel pictured above.

So there is 2 concerns with the design going forward from here:

1) I'm concerned about the internal wave of the mid driver causing coloration. I know many speakers are configured in the way my design currently stands, but I just want to get it right!

2) Having tested the JMLC horn and the 2452H driver, crossed over to my regular floor standing speaker (a JM Labs 816s) somthing is not quite right. They are EQ'ed dead flat, but rock music is just not right. The top end is overly harsh. What causes this I dont know.

Things it could be:

a) The Titanium 2452H diaphragm breakup null at 14Khz is just too much, so I would need to go over to the 2452SL diaphragm (different shape and has a 'Aquaplas' coating)

b) The 2452H really starts losing phase (inverting the dome) at 7khz or so, and this is more audible than expected so I need to go over to the JBL D2430K duel concentric driver.

c) The horn profile causes HOM too much? (initially thought I had this issue before I used careful EQ and the effect I perceived vastly reduced.)

d) The 1.5 inch format needs a tweeter to take over above 7khz.

e) Phase? I reversed the phase between the JM labs and the horn as when in phase it created a deep null, but have not used re-phase at all on the EQ'ed response yet.

f) Something else.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.