LORDSANSUI, your simulations are not right, this is off topic but some quick pointers :
First, that equation (Fs/Qts*0.39) is not to calculate f3, f3 has direct relation with given volume.
you cannot run different drivers with different Qts, Vas ... in same cab size, you have to change your cab size, port diameter and length (not just length) based on your driver and calculated tuning frequency, if you take a look at Paul's results you see he has different cab sizes.
you cann't have higher SPL than the driver itself, your result have 99db for 18Sound driver while the driver itself has only 97.
and to answer your question about AE driver, as been talked about before, first its Qms is not that bad and they are other thing that comes into equation as well, besides the overall high quality of the driver itself and good parameters, none of the other drivers can go that low and w relatively smaller cab, also Paul simulation shows it goes 90db which pretty damn high for that low freq and totally acceptable as a high sensitivity speaker.
I suggest to take a look at WinISD, not fancy but it calculates the volume and tuning freq for you and and you can see f3 and also you can see how changing volume and tuning freq can effect the low freq response.
Hope this helps
Last edited:
If the volume is a constrain, there is no problem to use the same volume for all simulation. The result would not be optimized for some drivers, but it could be optimize after. The major part of frequency response was flat.
The port area and port length are not linked to the driver, they are linked to the rear chamber volume and the tuning frequency to be achieve. So I just changed the length to better align with some drivers Fs. Simulating at 2,83v the port isn't a problem, Port fine tuning is done to reduce chuffing noise and compression (port velocity under 30m/s), but it's checked at Pmax or around it not at 2,83v.
I don't think the simulation are not right, it could be said that they are not fine tuned, but for driver comparison I didn't saw reason to fine tune all simulations. maybe I'm missing something, but if they didn't help, there is no reason to keep doing it. it's not my intention to mess the thread. I will keep reading it to discover what I didn't got.
Attached you can check the simulation to 18Sound. It has already been proven that Hornresp results are aligned with WinISD.
The port area and port length are not linked to the driver, they are linked to the rear chamber volume and the tuning frequency to be achieve. So I just changed the length to better align with some drivers Fs. Simulating at 2,83v the port isn't a problem, Port fine tuning is done to reduce chuffing noise and compression (port velocity under 30m/s), but it's checked at Pmax or around it not at 2,83v.
I don't think the simulation are not right, it could be said that they are not fine tuned, but for driver comparison I didn't saw reason to fine tune all simulations. maybe I'm missing something, but if they didn't help, there is no reason to keep doing it. it's not my intention to mess the thread. I will keep reading it to discover what I didn't got.
Attached you can check the simulation to 18Sound. It has already been proven that Hornresp results are aligned with WinISD.
Attachments
hi Aatto,
Me again, sory. My last suggestion, not for driver anymore, but for the internal loundspeaker design.
If I remember correctly the LF driver could have a specific chamber for it in this project, if this is right, the publication below might be interesting for you. Brian Steele develop a different way to design bass reflex he called pseudo-MLTL. He achieved some improvements in the upper frequencies response.
At link below you find his first proof of concept at home level application
The Subwoofer DIY Page v1.1 - Projects : The Boom Unit
At link below you find his proof of concept extending the concept to a PA application
The Subwoofer DIY Page v1.1 - Projects : Proof of Concept #6
Me again, sory. My last suggestion, not for driver anymore, but for the internal loundspeaker design.
If I remember correctly the LF driver could have a specific chamber for it in this project, if this is right, the publication below might be interesting for you. Brian Steele develop a different way to design bass reflex he called pseudo-MLTL. He achieved some improvements in the upper frequencies response.
At link below you find his first proof of concept at home level application
The Subwoofer DIY Page v1.1 - Projects : The Boom Unit
At link below you find his proof of concept extending the concept to a PA application
The Subwoofer DIY Page v1.1 - Projects : Proof of Concept #6
TL is a very good concept, but it needs some experience to realize it in a good way. Because of the larger area of the TL horn, compared with a basreflex port, TL is more sensitive for low medium frequency resonances. They can be avoided by dampening material in the horn and choosing for the correct horn shape.
There are good tools to simulate it like the MJ King software. Also in Leap it can be simulated, but it is not so powerful.
If a TL will be chosen, there must be someone in the team with good experience in such design. From my side I only did some study and simulations, no practical TL design.
There are good tools to simulate it like the MJ King software. Also in Leap it can be simulated, but it is not so powerful.
If a TL will be chosen, there must be someone in the team with good experience in such design. From my side I only did some study and simulations, no practical TL design.
Last edited:
Lordsansui,
That is correct, tuning frequency fB is only determined by cabinet volume and port size, not by the driver parameters.
For my simulations I did post here, I have chosen concepts that are equal or more damped than a Butterworth 4th order basreflex alignmement. Less damped concepts than B4, have more ringing and worse groupdelay response. To realize B4 both tuning frequency fB and cabinet volume VB have to be chosen as a function of the driver parameters.
Hi Paul,
Brian suggestion is not exactly a standard TL, it's more simple to simulated and to build as BR. Easy to simulate with Hornresp, he also has a worksheet in excell file. If this particualr design is welcome, he might help to optimize a proposal with the AE TD15X driver, so you and the others can compare the solutions. He use to help a lot in this forum.
I see your point, you are designing taking into accout not just the frequency response but also the transients linked dampening. This makes things more complicated, mainly regarding personal taste. I have no pratical knologe in this area but Qts, Vb, filter type and filter order are the main rule.
In addition, Hifi and psychoacoustics are a long battle.
If it is possible to meet the design goals using bass reflex rather than TL (or pseudo-MLTL), I would be inclined to say it would be better to stick with bass reflex. I think this because, if I remember correctly, it is also a design goal that this speaker be intended for someone newer to speaker building to pick up and build, and bass reflex is a more common, easier design to build and understand for the new DIYer, particularly if pseudo-MLTL is as picky and finicky as I've read that TL designs are with respect to stuffing.
Just my thoughts as a newer DIYer myself, but I would be more inclined to pick up a project that is simpler to get right at this point. Keep it simple!
Just my thoughts as a newer DIYer myself, but I would be more inclined to pick up a project that is simpler to get right at this point. Keep it simple!
off topic.
just thought of something, I have a WAW(FAST) w SBA woofer, maybe I can start a thread to design a TL or pseudo-MLTL around that, but for now i have to wait and see how this goes.
Sorry to disagree but a clarification is needed, the sentence above is wrong because technically is possible to achieve higher efficiency based on two things:
- Space radiation
- Enclosure type
Specifically for the 18Sound 15W750 simulation, the datasheet value was measured at free space (4pi) using 125L sealed box while the simulation was done using half-space (2pi) and bass reflex.
The 18Sound 15W750 has a sensitivity of 99.5dB in half space if calulated out of the TSP parameters.
It is B4 basrelflex for Vb = 175 L and tuned at fB = 42 Hz, then F3 = 42 Hz.
ok, thought we r done w this, we are not talking "technically", yes technically we can go active and achieve 100db from a 90db driver ! while at it we can also use dsp and eq and get the flattest respond too but it is not about this project and I assumed we are talking within this project and same goes with 18Sound driver, one would assume you are talking full space unless mentioned other wise, and if this helps and makes thing go faster yes you were right, technically that MAYBE possible and yes we can achieve 99db half space from a driver w 97db full space
hope we can let go and move on now.- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.