My Dream Project: Planning a TL with Fostex Full Range Driver

[Hari Iyer]

Yes, I know but my laptop does not have line in hence cannot measure impedance. I think the jig needs Line-in + Line-out to measure, right? At least that was what I'd concluded when I looked it up a long time ago.
But my point is - the measures need not really be done - if one uses MJK's sheets - you're set. There are enough people who have verified this. Of course one can always go OCD in which case this never ends.

All this measuring, tuning etc only comes when you use the classic alignment tables or Shultz's paper for designing boxes for TL's - those are a bit of hit and miss.

IME the measured and simulated (e.g XSIM) is very close to each other using REW and my simulations provided you have measured the TS parameters of the driver accurately else not.

I also strongly believe that you need to validate your simulated response with the measured response to be in sync else you will not be able to tweak only with subjective listening.

@keith - can you post your simulated response of your speakers? You can post only the impedance curves. I would love to see your measured response too, but you mentioned you dont have a sound card to measure nor do your laptop have line-in (stereo).

 

[keith_correa]

IME the measured and simulated (e.g XSIM) is very close to each other using REW and my simulations provided you have measured the TS parameters of the driver accurately else not.

Let me try this stating this again - Accurate TSP's are a must - no two ways about it. Now what good will accurate TSP's get you if you input these accurate TSP's into dodgy [for want of a better word] sim software/whatever? Now on the flip side - what will the same accurate TSP's get you when you input these accurate TSP's into accurate sim software? Enough said.

I also strongly believe that you need to validate your simulated response with the measured response to be in sync else you will not be able to tweak only with subjective listening.

OK, Case 1 - the case which you state - I agree with this IF you are tweaking or want to tweak.
Case 2 - what if you do not want to tweak because your output is in sync with the sims that you did. The point that my friend is trying to make is that he NEVER feels the need to tweak because the stuffing density that he inputs to simulate and the stuffing density he uses [Bonded Dacron] to stuff is accurate. Because the software he uses is so accurate, sims match reality. He says he initially used to measure but gave it up subsequently because sims using MJK's software and measured responses were so accurate.

@keith - can you post your simulated response of your speakers? You can post only the impedance curves. I would love to see your measured response too, but you mentioned you dont have a sound card to measure nor do your laptop have line-in (stereo).

Yes, I do not have a soundcard nor does my laptop do line-in. Nor do I have a calibrated microphone. I will invest in this ONLY when I will start to build the definitive slim [I have a thing for slim floorstanders and besides they will be relatively unobtrusive in my relatively small living room] TL floorstander for my room. Till then, I can't justify the expense to myself.

 

[venki7744]

Wow!

This is turning out to be an interesting thread . Keith sire, I have not used MJK spreadsheet hence I cannot comment on its accuracy. However I still think that no software is 100% accurate and even if it does, there would be lot of error introduced while building the enclosure from very small things like passive component tolerance to error in volume introduced by bracing driver displacement, unequal stuffing density (different from what was designed), error in calculation of SD etc. Hence measuring the final response I feel is the one of the most important part of a loudspeaker design. I believe every loudspeaker manufacturer measures their loudspeaker unit once assembled for this very reason.

Hari Sire,

One curious question, is the xSim response including the enclosure? The reason why I am asking this is most of the TL would have a long tail of ripple (you can tame it with stuffing but IMO you cannot make it disappear). If this is infact the FQ of the system very curious to know how you tamed the ripples.

Cheers,
Venki

 

[Hari Iyer]

Let me try this stating this again - Accurate TSP's are a must - no two ways about it. Now what good will accurate TSP's get you if you input these accurate TSP's into dodgy [for want of a better word] sim software/whatever? Now on the flip side - what will the same accurate TSP's get you when you input these accurate TSP's into accurate sim software? Enough said.


OK, Case 1 - the case which you state - I agree with this IF you are tweaking or want to tweak.
Case 2 - what if you do not want to tweak because your output is in sync with the sims that you did. The point that my friend is trying to make is that he NEVER feels the need to tweak because the stuffing density that he inputs to simulate and the stuffing density he uses [Bonded Dacron] to stuff is accurate. Because the software he uses is so accurate, sims match reality. He says he initially used to measure but gave it up subsequently because sims using MJK's software and measured responses were so accurate.


Yes, I do not have a soundcard nor does my laptop do line-in. Nor do I have a calibrated microphone. I will invest in this ONLY when I will start to build the definitive slim [I have a thing for slim floorstanders and besides they will be relatively unobtrusive in my relatively small living room] TL floorstander for my room. Till then, I can't justify the expense to myself.

I have nothing to debate with your points - i just asked you for a "simulated" impedance curve. You could have just replied a plain - yes or no.
Yes - i have it and here is the response.
No - i dont have a simulated response.

 

[keith_correa]

I have nothing to debate with your points - i just asked you for a "simulated" impedance curve. You could have just replied a plain - yes or no.
Yes - i have it and here is the response.
No - i dont have a simulated response.

Was it a debate? Not to my knowledge. Did I ask for one? Not to my knowledge.

If you wanted a plain yes or no for a question then you should have asked ONLY a question in response to my post and not other stuff to which I felt obliged to respond to and did.

Since you are getting "snarky" - I'll leave you to your thread and get out of here. Good riddance to me!

 

[Hari Iyer]

Wow! One curious question, is the xSim response including the enclosure? The reason why I am asking this is most of the TL would have a long tail of ripple (you can tame it with stuffing but IMO you cannot make it disappear). If this is infact the FQ of the system very curious to know how you tamed the ripples.

You can measure the TS parameter with the driver in the box and use that for your simulation. That way you have factored for the box too and also the internal cables.

With some more simulations i am coming with the following observations,

1. Tapered line - difficult to find the exact driver position. The driver position comes too close to the floor and hence not practical.
2. Expanding line - Length of the line is too impractical to build in a small room like mine.
3. Straight line - Good enough for my room, but still longish. There is a potential to build a slim profile box with this design.
4. MLTL - very practical size of the box and i may finally lean into this as I have to consider WAF also as a design parameter considering my room size.

Will finalize the box dimensions within a week and also verify if its possible to tweak further with the simulator.

 

[venki7744]

You can measure the TS parameter with the driver in the box and use that for your simulation. That way you have factored for the box too and also the internal cables.

To clarify, I modelled a Tympany 3"driver in a tapered enclosure (80cm long) in Hornresp then imported the FR into x-Sim and then checked the step and square wave response and this is what I got. If you just give the Ts specs of the driver the x-Sim gives a perfect square wave response and step response but not with FR is given as input to the x-Sim.

For clarity attaching the xSim file (please rename it to .zip file).

View attachment Tympany_3Inch_TaperTL.pdf

 

[Hari Iyer]

After many days / weeks of tweaking my simulation, i will finally build a straight line TL. It allows me a slim and small profile and with more accurate tuning, better terminus and system FR, better TL Acoustic Impedance, and better stuffing control over other box geometry.

I am in the final stages of preparing my drawings will ask the carpenter to build a box (this is the most tough part in my case).

 

[kapvin]

You need not measure the FR response to check the TL measured and simulated response. The FR response will not show much about the tuning and stuffing. This will be shown by the impedance curve.

You need to measure the impedance of the speaker with stuffing (two humps), low stuffing (one hump low and other hump high), adequate stuffing (one clean hump without any distortion on the curve ie. critical dumping), over stuffing will be indicated by the impedance @ resonance close to Free Air impedance.

In critically damped TL the impedance @ resonance (tuning freq of the line) is atleast 40% to 50% lower than the Free Air impedance.

my 2 p.

Hari, this is very Intriguing. could you please clarify what you mean. both by the first sentence. and the second. thank you so much.

After many days / weeks of tweaking my simulation, i will finally build a straight line TL. It allows me a slim and small profile and with more accurate tuning, better terminus and system FR, better TL Acoustic Impedance, and better stuffing control over other box geometry.

I am in the final stages of preparing my drawings will ask the carpenter to build a box (this is the most tough part in my case).

Hi Hari,

we had discussed this offline. the t/s parameters of the driver do not really lend themselves to a straight TL. and the sims suggest that. we could do a MLTL, TQWT or Mass loaded TQWT. yup the boxes will be large. but at least it will be meaningful in terms output. anything else will still mean a "not small" box but a bit meaningless in terms of output.

Or you could choose another driver.

 

[Hari Iyer]

Hari, this is very Intriguing. could you please clarify what you mean. both by the first sentence. and the second. thank you so much.

There is a typo error in my statement

original:
You need not measure the FR response to check the TL measured and simulated response. The FR response will not show much about the tuning and stuffing. This will be shown by the impedance curve.


Its difficult to measure the FR in a normal room for the unstuffed line without room effect. Hence its difficult to ascertain the results. Its very easy to measure the impedance curve as this will not have any room effect.

First measure the unstuffed line impedance curve to check if the impedance of the measured and simulated line is similar.

After adding stuffing you can again measure the impedance to check if they are similar to the simulated response.



original:
You need to measure the impedance of the speaker with stuffing (two humps), low stuffing (one hump low and other hump high), adequate stuffing (one clean hump without any distortion on the curve ie. critical dumping), over stuffing will be indicated by the impedance @ resonance close to Free Air impedance.

modified:
You need to measure the impedance of the speaker with NO stuffing (two humps), low stuffing (one hump low and other hump high), adequate stuffing (one clean hump without any distortion on the curve ie. critical dumping), over stuffing will be indicated by the impedance @ resonance close to Free Air impedance.

Hope this clarifies the second sentence.

Hi Hari,

we had discussed this offline. the t/s parameters of the driver do not really lend themselves to a straight TL. and the sims suggest that. we could do a MLTL, TQWT or Mass loaded TQWT. yup the boxes will be large. but at least it will be meaningful in terms output. anything else will still mean a "not small" box but a bit meaningless in terms of output.

Or you could choose another driver.

Thanks for your inputs. I will try and post today what i got from the LA software with MJK's alignment tables. I have tuned close to driver resonance @ 54 Hz.

If you can, you too can post MJK's mathcad response. This will also help me to validate the LA software vis-a-vis Mathcad.

 

[kapvin]

Thanks for your inputs. I will try and post today what i got from the LA software with MJK's alignment tables. I have tuned close to driver resonance @ 54 Hz.

If you can, you too can post MJK's mathcad response. This will also help me to validate the LA software vis-a-vis Mathcad.

I can do that for sure. Please share the enclosure geometry. Please note that for sims I took t/s from the website, as the Specs you gave in the mail were inconsistent possibly due to typo.

I must add that the license that I have for the software is only for non commercial use. If you plan to commercialise the design, you need to let me know.


Sent from my iPad using Tapatalk HD

 

[Hari Iyer]

I can do that for sure. Please share the enclosure geometry. Please note that for sims I took t/s from the website, as the Specs you gave in the mail were inconsistent possibly due to typo.

I must add that the license that I have for the software is only for non commercial use. If you plan to commercialise the design, you need to let me know.


Sent from my iPad using Tapatalk HD

No, I do not intend to commercialize this speaker in any way and is for my personal use only.

Here are my input values as per MJK's alignment tables, I too have used the published parameters to arrive at this simulation.

Tuned @ 54Hz (Approx).
Line length = 61.1 inches
Cross section = 54.625 sq.inches
Driver off-set = 21.28 inches
Terminus = opens at the bottom (No ML)
Geometry = SL
Preferred stuffing = 0.92 lb/ft3

Below are my simulated results,
Unstuffed:

Stuffed:

As per Keith in his earlier post, mentions that the LA does not simulate correctly if stuffing is added. He has confirmed this with his other friends.

 

[kapvin]

thanks for the assurance. more than happy to assist in anyway possible.

edit.. if you crop and then post your sims, it might be easier to see

will either post or mail you the sims in a bit.

here is the unstuffed FR as per your enclosure geometry

there is a bit of a rising response which means that you can tune lower..

the ripples are worrying

and this is the impedance curve

with stuffing

that kind of is counter productive, right?

 

[Hari Iyer]

Yes, very much counter productive indeed. I realized today morning that there was a major error on how i calculated the box. My new dimensions are as below,

Tuning Freq - 54Hz
S0/SL - 98 sq.in
Offset - 4 in
Stuffing - 0.8 lb/ft3
Line length - 58 in

I have used Qtd = 0.3 (measured, published is 0.25).
I have used Vas = 31 Liters (measured, published is 37 liters)

I am in office today for some work, will post simulations after cropping later in the evening.

Thanks for your assistance.

 

[kapvin]

and here's slightly different way to look at the whole thing..

And another. remember these thought starters rather than refined sims, just to get a sense of direction. this one builds on your requirements a bit better - higher efficiency..bass response tapers in sync with room gain for a medium sized room

 

[Hari Iyer]

This looks cool, can you also post the stuffed response too. Is there any ML of the line. The tuning is a bit higher than the resonance of the driver around - 65Hz again the two humps are identical - great. Also offset should be ideally 0.341

Also post the TL Acoustic impedance too as that needs to be a smooth decay.

Also try simulation with my calculated values from MJKs alignment.

Thanks,

 

[kapvin]

This looks cool, can you also post the stuffed response too. Is there any ML of the line. The tuning is a bit higher than the resonance of the driver around - 65Hz again the two humps are identical - great. Also offset should be ideally 0.341

Also post the TL Acoustic impedance too as that needs to be a smooth decay.

Also try simulation with my calculated values from MJKs alignment.

Thanks,

this is with the original t/s params. low qts drivers need to be tuned a bit above resonance. this is the stuffed response. see the other sim as well. both are stuffed.

 

[Hari Iyer]

this is with the original t/s params. low qts drivers need to be tuned a bit above resonance. this is the stuffed response. see the other sim as well. both are stuffed.

Ok, the stuffing seems a bit less to me 0.48 lb/ft3 as the first hump is not damped at all. Also the system response is not able to reduce the odd harmonics completely. I love to have a single hump on the TL electrical Z with no distortion. But the Z @ resonance should be 60% lower than the free air Z.

May be increasing the stuffing will say a complete different story.

Also try with my MJK's alignment calculated values. I have seen that the odd harmonics of the calcualted and simulated values are 100% accurate. So is the TL acoustic Z, group delay and velocity.

 

[kapvin]

Ok, the stuffing seems a bit less to me 0.48 lb/ft3 as the first hump is not damped at all. Also the system response is not able to reduce the odd harmonics completely. I love to have a single hump on the TL electrical Z with no distortion. But the Z @ resonance should be 60% lower than the free air Z.

May be increasing the stuffing will say a complete different story.

Also try with my MJK's alignment calculated values. I have seen that the odd harmonics of the calcualted and simulated values are 100% accurate. So is the TL acoustic Z, group delay and velocity.

sure. Like i said, these are thought starters done in 5 minutes each rather than any extensive design output. I am not designing this. you are I will just be happy to help you validate your results.

best wishes

 

[Hari Iyer]

sure. Like i said, these are thought starters done in 5 minutes each rather than any extensive design output. I am not designing this. you are I will just be happy to help you validate your results.

best wishes

Thanks so much for helping validate my design with mathcad.

I am now trying an expanding taper design with a Taper Ratio of 2. Below are the other parameters for your consideration,

S0 = 77.116 sq.in
SL = 154.232 sq.in
SL/S0 =2

Leffective = 64 in
Lactual = 61 in

Tuning freq = 60Hz
Offset = 22.567 in (tweaked) calculated = 22.274 in
Preferred stuffing = 0.78 lb/ft3

This is what i get from the LA for the stuffed and unstuffed line:

Unstuffed:

Stuffed:

Eager to have a look at mathcad too