Designing a 2-way bookshelf speaker

[Kannan]

Since you have played with Dayton drivers, I suggest you try with the Dayton Ref. 180 4 ohm in combination with the Dayton Ref. 100 4 ohm.
This is a very good combination where the 180.4 Dayton can just handle the lows below 300 hz while the 100.4 take up from 300.
The 100.4 has an excellent open midrange and that too with lots of body and a usable bandwidth upto 12 khz
You can just run a simple tweeter (I recommend the Dayton AMT with a f/s of 5 khz) to get the airiness. Cross it at around 11 or 12 khz with a simple capacitor.

 

[tcpip]

Since you have played with Dayton drivers, I suggest you try with the Dayton Ref. 180 4 ohm in combination with the Dayton Ref. 100 4 ohm.
This is a very good combination where the 180.4 Dayton can just handle the lows below 300 hz while the 100.4 take up from 300.

I want to build a 2-way. So, this does not fit.

Secondly, why would I use a 6.5" for pure woofer duty, even assuming I was building a 3-way? If you ask me, I'd use a 10" at least. The Dayton RS270 comes to mind.

Thirdly, the crossover for a 3-way passive design would be more painful to design than a 2-way.

Fourthly, my aim is to do the design with Indian drivers as much as possible, and certainly with drivers sold in India. The drivers you mention are mostly proudly marked "SOLD OUT" with the only reseller who sells them.

And finally, I'm aiming for a low-cost design. A 3-way with an AMT tweeter is more expensive than cheaper alternatives.

But I agree that with your suggested drivers and xo points, the xo design becomes simpler and if someone likes the idea of a 3-way with those drivers, then your approach would work.

 

[Kannan]

I want to build a 2-way. So, this does not fit.

Secondly, why would I use a 6.5" for pure woofer duty, even assuming I was building a 3-way? If you ask me, I'd use a 10" at least. The Dayton RS270 comes to mind.

But I agree that with your suggested drivers and xo points, the xo design becomes simpler and if someone likes the idea of a 3-way with those drivers, then your approach would work.

Sir, I have a 3-way Tower design with the Dayton DC250 (10 inch) and have also worked with the RS180 in a ported bookshelf design.
I also have these drivers performing midbass duty in the front doors of my car in a full 4-way active setup.

The performance of the RS180 which is metal cone is aplomb and can hit lows without the need for a sub when not crossed high as in a 2-way design. That is why I mentioned a range of under 300 hz to 400hz.
The concept of wideband with the likes of Rs100.4 is very interesting.
They can do a wide octave without getting muddy on the lows or shouty in the highs and can take care of the most critical 400 to 5 khz region of the music. The RS100 goes much much higher.

I though understand your reasoning of designing a budget system, though the Daytons too are not expensive.

 

[jmascreen]

Something similar in line with GR XLS classic/encore?

Sent from my Moto G (4) using Tapatalk

 

[Hari Iyer]

Yes I was afraid I'll have to think beyond bass reflex for some of these smaller drivers. I will at all costs, aim to keep the box size small. So, things like MLTL are out. (I remember Lynn Olson's famous Ariel speakers, slim MTM TL floorstanders with 5.25" Vifa poly cone midbass. I'm not building a floorstander here, come what may.)

My recommendation is this - Design the box size for a vented enclousre using WinISD and have a small rectangular opening at the bottom front below the woofer (baffle at the bottom) of size 1/2" x 4". Suff the enclosure with loose to medium polyester fiber. Adjust the stuffing by checking the impedance in the REW. If the mounted + stuffed impedance @ system resonance is less than 40% of the free air resonance then you have got both the stuffing and the box right.

This will ensure minimum impedance @ resonance there by increasing low end extension.

Do not go for box tuning below resonance of the woofer as this could boom the sound stage.

 

[HormusPeston]

What a wonderful thread! Thank you, tcpip.
Warmly,
~HP

 

[tcpip]

Something similar in line with GR XLS classic/encore?

Yes, surprisingly close to what these speakers were trying to do. However, I'll reach less success because I'll be using the off-the-shelf drivers from a company which sells its production overruns to our retail market, while the original designers got the drivers tweaked to get better performance.

Sent from my Moto G (4) using Tapatalk

Maybe you can think of switching off this signature, to conserve bandwidth? I got reprimanded on diyaudio.com once for keeping my Tapatalk signature on. hyeah:

 

[tcpip]

What a wonderful thread! Thank you, tcpip.

Heh heh, don't embarass me. Let me get off my posterior and actually build something worth listening to, then you can give me all the kudos. hyeah:

 

[tcpip]

I have so many questions about all this:

My recommendation is this - Design the box size for a vented enclousre using WinISD ...

I'll first have to do T/S parameter measurement to arrive at a reasonable box size, right?

... Adjust the stuffing by checking the impedance in the REW.

Does REW let you measure speaker impedance? I didn't know this. I thought you'll need at least Speaker Workshop and an impedance jig to do impedance measurements.

If the mounted + stuffed impedance @ system resonance is less than 40% of the free air resonance then you have got both the stuffing and the box right.

Can you give me some numbers? If the free air resonance (Fs) is, say, 100Hz, then the speaker-in-box resonance must be below 40Hz (less than 40% of the free-air Fs)?

Do not go for box tuning below resonance of the woofer as this could boom the sound stage.

I can't understand at all how anything you do to tune the bottom end of the speaker+enclosure will impact soundstage. Soundstage is not impacted by low frequencies at these octaves at all, I thought? Or did you mean just "make the sound boomy"?

 

[tcpip]

Ok, the die is cast. Just placed the order with diyaudiocart.

The TM design will use the SM16IH-4 4-Ohm 6.5" midbass unit with the stamped steel frame, and the TL26SG tweeter. I chose 4 Ohm for the midbass unit so that I can pull out more power from the amp for the large portion of the frequency range which the midbass will cover. Tweeters are more sensitive than midbass drivers most of the time anyway, and need to have their levels cut with L-pads quite often. So having a 8-Ohm tweeter with a 4-Ohm woofer will help bring down the relative level of the tweeter a bit. Driver cost for this model: Rs.5,560/pair

The MTM will use the M13NH-08 midbass unit and the Tympany XT25SC90 tweeter. Here, the two midbass units will parallel to give me a nominal 4 Ohms per side, and will mate with the 4 Ohms of the ring radiator. Driver cost for this model: Rs.8,571/pair

What can we guess about the final speakers going by just the driver choices?

• None of these midbass drivers are going to be particularly low-distortion, irrespective of what their sales pages say. In particular, I expect the 6.5" midbass driver to be just about average on that front. In comparison, the Dayton RS drivers would have had a lot lower distortion, but hey, we wanted Indian drivers. Of the two, I expect the 5.25" pair of the MTM design to give slightly lower distortion than the big single 6.5", simply because two 5.25" drivers have a larger radiating area, hence will have lower excursion. Of course, it's likely that many of us won't notice the higher distortion.
• The higher octaves, I expect, will be quite decent. Both tweeters are pretty good, probably roughly in the same league in terms of distortion. I expect to cross both designs over at about 2KHz or so.
• The bass response from both designs will be adequate for small rooms. (The typical Bombay bedroom is 10' x 13'. That's a small room.) The MTM may have lower bass extension than the 6.5" TM design, because of its smaller drivers and higher Fs. No earth-shaking home theatre thunder here, but may be perfectly adequate for most types of music. The MTM user who wants more bass can add a sub for the region below about 50-60Hz.
• I expect the midrange to be good, within the limits of any distortion issues in Point 1 above.
• These speakers will not be good party speakers -- they won't play clean and very loud. For those use-cases, you have to start with a minimum of 2 x 6.5" in an MTM config. But they should be good to excellent for relatively near-field listening in small rooms.

It'll be interesting to see how the whole thing turns out, when all is done and dusted.


By the way, I want to make one point explicit and clear: I place no restrictions on anyone reproducing my designs for personal, commercial, business, or any other use, in any quantity. Kits and finished products made by others are most welcome. I just don't want anyone building and selling them commercially to say that they are based on my design. Basically, go do what you want with these designs, without asking my permission. But if you're doing it commercially, don't refer to me. Say you designed them yourself, or say you got the design from an unspecified source. All my speaker designs are in the public domain, unless otherwise specified.


The thing I find depressing about doing these designs is that I have no visibility about how many drivers there are in stock of each of these models. If the designs turn out worth replicating, I don't know how many of you will be able to make them before the drivers run out. And with Peerless India, once a driver runs out, it will never reappear. We're the tail end of the remaindering surplus-disposal stream of a company which does not build these models any more. In fact, diyaudiocart has more drivers listed with the "SOLD OUT" banner than drivers actually available, I think.

In fact, this was my primary reason to move entirely away from Indian drivers after the initial Asawari versions.

I bought the drivers for my Darbari in 2006-2007. Even today, I can buy a second set of those drivers. I like that.

Maybe someday, after this, I'll make one or two designs with drivers "imported and probably available in India".

 

[Hari Iyer]

I have so many questions about all this:

I'll first have to do T/S parameter measurement to arrive at a reasonable box size, right?

Yes, i generally break-in the woofer for approximately one to two hours at half octave below resonance frequency before doing my first measurement and moderate levels (free air)

Doe s REW let you measure speaker impedance? I didn't know this. I thought you'll need at least Speaker Workshop and an impedance jig to do impedance measurements.

Yes, REW has impedance and TS measurement interface. Without impedance measurements, you cannot measure the TS parameter of the drivers. You still will require an impedance jig with REW.

Infact I found the REW help much easier to understand than the speaker workshop.

Can you give me some numbers? If the free air resonance (Fs) is, say, 100Hz, then the speaker-in-box resonance must be below 40Hz (less than 40% of the free-air Fs)?

You probably mis-understood what i said or i may not have put it properly.

I am saying this. If the resonance of the woofer is say 70Hz and the impedance at resonance is say 100 Ohms, if you are making a Aperiodic loading enclosure say tuned at 70Hz, after mounting the woofer in the box and stuffing the enclosure the impedance at 70Hz should be around 60 ohms for it to be proper. i.e. there should be atleast 40% drop in impedance @ tuning frequency in a properly tuned Aperiodic enclosure. You can measure the impedance using REW / Speaker workshop before you even build the crossover on the mounted + stuffed box. Hope this clarifies.

I can't understand at all how anything you do to tune the bottom end of the speaker+enclosure will impact soundstage. Soundstage is not impacted by low frequencies at these octaves at all, I thought? Or did you mean just "make the sound boomy"?

[/QUOTE]

Typically for woofers having resonance above 35Hz its better to tune the box at or above the resonance frequency. You can tune lower when the resonance of the woofer is below 35Hz. In this peerless driver if you tune lower, there would be significant cone movement which this driver would find difficult to handle and control and at times may tend to boom when setup in your room. To avoid this tune it at resonance or slightly higher to avoid this problem.

All my 2 paise, rest you are the designer at your best.

Cheers,

 

[tcpip]

Hey, thanks a lot.

Yes, REW has impedance and TS measurement interface. Without impedance measurements, you cannot measure the TS parameter of the drivers. You still will require an impedance jig with REW.

Infact I found the REW help much easier to understand than the speaker workshop.

Good to know about REW. And I agree that Speaker Workshop has a steep learning curve. I just was lucky; I had a very, very patient guru.

Considering how Speaker Workshop is now no longer maintained, and how Windows versions have moved on, I'm thinking of upgrading my toolset to use ARTA for all measurements, and maybe a combo of XSim or VituixCAD for simulation and xo design. For box simulation, I will continue to use Unibox -- I really like it.

Maybe someday, I'll switch to SoundEasy -- don't know yet.

I am saying this. If the resonance of the woofer is say 70Hz and the impedance at resonance is say 100 Ohms, if you are making a Aperiodic loading enclosure say tuned at 70Hz, after mounting the woofer in the box and stuffing the enclosure the impedance at 70Hz should be around 60 ohms for it to be proper. i.e. there should be atleast 40% drop in impedance @ tuning frequency in a properly tuned Aperiodic enclosure. You can measure the impedance using REW / Speaker workshop before you even build the crossover on the mounted + stuffed box. Hope this clarifies.

Thanks a ton. I now understand that you were referring to the impedance, not the Fs. I'll try this.

Typically for woofers having resonance above 35Hz its better to tune the box at or above the resonance frequency. You can tune lower when the resonance of the woofer is below 35Hz. In this peerless driver if you tune lower, there would be significant cone movement which this driver would find difficult to handle and control and at times may tend to boom when setup in your room. To avoid this tune it at resonance or slightly higher to avoid this problem.

I'll check this out. I'll take a look at my simulations (all done in Unibox) to see where the Fb is compared to the Fs. If I remember right, all of them have Fb maybe 50% to 100% higher than Fs. Even in those cases, I find that F3 is lower than free-air Fs, and F10 is way lower. I guess the in-room bass extension is somewhere between F3 and F10, no? (In all this, I'm referring to bass reflex enclosures.)

 

[Hari Iyer]

Have you tried LA (Leonard Audio) for box design/ simulation. Check out if time permits. Its cool. I designed my Meteronome MLTL using it and the outcome was pretty good.

 

[tcpip]

Have you tried LA (Leonard Audio) for box design/ simulation. Check out if time permits. Its cool. I designed my Meteronome MLTL using it and the outcome was pretty good.

No. Never did an MLTL, so never looked for any software for it.

 

[Hari Iyer]

No. Never did an MLTL, so never looked for any software for it.

You can simulate other types too other than TL.

 

[captrajesh]

The thing I find depressing about doing these designs is that I have no visibility about how many drivers there are in stock of each of these models. If the designs turn out worth replicating, I don't know how many of you will be able to make them before the drivers run out. And with Peerless India, once a driver runs out, it will never reappear.

For the likes of Peerless who being in EPZ, are bound to cater predominantly to overseas buyers, people like you don't exist. May be you should look for local manufacturers who are not driven by increasing turnover and profitability and happy with their existing clientele. They would be more accommodative of passionate DIYers.

 

[tcpip]

For the likes of Peerless who being in EPZ, are bound to cater predominantly to overseas buyers, people like you don't exist. May be you should look for local manufacturers who are not driven by increasing turnover and profitability and happy with their existing clientele. They would be more accommodative of passionate DIYers.

I would love to look at any sources available, but those smaller manufacturers are even less process oriented and accessible. I have, at one point, bought a set of 6.5" drivers from Bolton, from their factory near Delhi. You get zero technical details and wide variability of measured T/S parameters, even more variable than Peerless India. You get Qts greater than 1.0 sometimes. [emoji2]

So, Peerless India is probably better than dealing with those manufacturers.

 

[jmascreen]

How about YP?

 

[tcpip]

How about YP?

What is YP?

 

[jmascreen]

What is YP?

Another local brand, I had used it during my school days.