The Natangs Project Born

[mackintire]

Project Killed and reborn as "The Kadabbies" ....see page 3


Revised March 29th 2009

The twin towers will be 42" tall 7.5" wide 9.5" deep.

The origional tower design had a downward firing port, but that idea did not work out.

Do to my SPL requirements I will be using a sealed design.

They have design requirements that cross between the Natalie P's and CJD's RS150 MTM.

They are louder and play lower than the RS150s, but it remains to be seen if they can match up to the natalie P's as they quite a bit smaller.

The Natang towers use the following drivers in a Acoustic suspension design (aka Sealed box):

(2) Tang Band W5-1685 5" Underhung Midbass Driver
(1) SEAS Prestige 27TDFC (H1189) Textile dome


The Natang Center uses the same Seas tweeter and a pair of 4" tang band underhung drivers and is designed to be far smaller than almost any center currently in the DIY projects section. It's being design as hybrid passive radiator/ Sealed design. At lower volumes the cabinet will operate as a passive radiator, over that it will start to operate as a seal enclose and the bass will diminish. At max volume it is expected that the center will start rolling off around 90hz. At more normal volume levels the center should be about to reach 65hz. In my case I will be using my reciever to cross them over at 80Hz.

The size of the center channel's cabinet will be very small 6"x16"x7" featuring (2) 5.25" side firing passive radiators.


Finishing the system I've added a DIY subwoofer.

The sub has 3600ci of internal volume, is tuned to 22hz with a 3" precision port and is powered by a 500watt O-audio AMP. The subsonic filter is set for 25hz which results in a flat response to 22Hz with 20hz being -4db and 18hz being -10db.


The subwoofer I used is the Dayton RSS265HF-4 10" Reference HF. This Sub does not like to be overdriven, doing so will damage this sub. I have mine tuned to nearfield 108db output, which put the in room response at some notes near 113db.


Why this order? No reason other than..thats the way it's happening:

The sub is being built first and should be functional by the second week of March. UPDATE 3/29/2009: its running in my house now. It still needs veneered.

Update 7/10/2009: The subwoofer cabinet is being veneered and if I am lucky stained within the next two weeks.

Update 9/2/2009: The sub cabinet is being stained this weekend. Pre stained pics added.

The Twin Towers are second. Baffles and box assembly mid September

The Center channel is third and will start once the towers are partly assembled.

[HareBrained]

Although I don't have experience with any of the drivers, it sounds like a good combination for the tower. You'll need to ensure the stand off from the carpet is at least as tall as the diameter of the port. I've also seen designs that use a base below to speaker to ensure the port output is not lost in the carpet. But that's simple enough to play with after it's built.

Regarding the W4, a single Peerless 5.25 PR per driver would work nicely. The Unibox sim does have the PR reaching xmax at a low power but that will simply result in a hybrid (vented-sealed) enclosure and will control driver cone excursion. But sealed would also be fine with a F3~93Hz. The difference would be minor especially if you'll be employing an active HP at 80Hz as the PR contribution above 80Hz is below -12db compared to the driver. Without the HP filter, the PR is worthwhile.

[mackintire]

This is looking good

So far this is looking very good. My WinISD pro looks good, now I have to model both designs in LSPCAD. I'm kind of jumping in at the other end of the pool as the driver selection and the cabinet design are coming first.

Where would you suggest that I crossover the Tangband drivers. I know they have breakup at 8k but many RS150 designs have very low crossover points. In this case It looks like I have lots of flexibility.

So what would be ideal? Chris a little guidance would be appreciated.


Hey,

I modeled the box and drivers in LPSCAD and the impedance is very low at 40hz near 2.5 ohms, should I worry?

[ThomasW]

In this case It looks like I have lots of flexibility.

So what would be ideal?

This is where having something like a DCX2496 and being able to actually listen to the system comes in handy.

Hey,

I modeled the box and drivers in LPSCAD and the impedance is very low at 40hz near 2.5 ohms, should I worry?

Yes this is a concern if your receiver isn't capable of driving low impedance loads..

That said IMO it's not the best idea to ask small drivers like this to operate as both woofer and midrange. Putting a real woofer below the MTM cluster would go a long way toward improving the performance of this system.

[HareBrained]

This is where having something like a DCX2496 and being able to actually listen to the system comes in handy.

Along with a mic, professional measurement software, mixing board, multiple high quality amps, and a sonically neutral room. But some, like myself, have less resources and are willing to accept the compromises. For those who only have Rules of Thumb, shoot for a xover point that is an octave above the Fs of the tweeter and below an octave below the break-up. There are component topographies that allow you outside that range but it shouldn't be necessary in this case.

Yes this is a concern if your receiver isn't capable of driving low impedance loads..

Yeah, it's a problem. Again, having some knowledge of different component topologies can help raise this. Make sure you enter the DCR values of the coils used in the woofer's low pass inductors. As they're in series, they will raise the impedance. If the low impedance is in the tweeter, use a different attenuation circuit. The typical L-pad (series and parallel resistors) have little impact on impedance, but increasing the series resistor or eliminating the parallel will add some ohms.

That said IMO it's not the best idea to ask small drivers like this to operate as both woofer and midrange. Putting a real woofer below the MTM cluster would go a long way toward improving the performance of this system.

I agree, it's not ideal, but there are an awful lot of TM and MTM designs running "full range" that sound very good. If you're running a HT receiver, using the active HP filter, and running a sub, there is no issue. If you're running them full range, you may find the midrange to be a little less open sounding as the driver is trying to play 7.5 octaves. But unless you have the ability to try the driver both ways, you probably will never know there will be a difference. And it's still going to sound very good.

[mackintire]

"Make sure you enter the DCR values of the coils used in the woofer's low pass inductors. As they're in series, they will raise the impedance."


I will do that, unfortunatly LSPCAD crashed last night and I lost about 4 hours of work. The good news is I should have time to enter all the data in this coming weekend. I was also tinkering with a complex notch filter that was in parrellel to get rid of the FS impedance peak. I managed to rid myself of the peak but ended up with a extremely load impedance at 40hz. I may try out a compromise that has the end resulting impendance looking like a double camel hump but still keeps the load under 20 ohms. As you said earlier at that point I hadn t yet figured out where I planned on setting the crossover point. I think I'm setting on 3K using 4th order LR crossovers.

The overall impedance of the system is mostly flat with a 7.2 ohm load across most of the frequency spectrum. The impedance start falling at 100hz and bottoms out at 40hz.

I'll post the outcome so there's more to see and comment on. Hopefully I get some decent constructive criticizim. (Thanks ThomasW)


More background info...
These towers will probably be set to small with a crossover at 80hz for movies, but when we use them for music I'd much rather run them full range if possible.

I'll be driving them with the Onkyo reciever that came with my Onkyo HTS-800

Thanks for the help so far!

[ThomasW]

but there are an awful lot of TM and MTM designs running "full range" that sound very good.

There are but it's room dependent. These small midwoofers aren't going to be happy trying to fill a largish room.

I think I'm setting on 3K using 4th order LR crossovers.

That seems like a good place to start.

BTW saving data files to a USB drive or similar is a good idea...

[mackintire]

"These small midwoofers aren't going to be happy trying to fill a largish room."

The Nat-P's in tower form are a little too large for the space.


Saving data...yea I know, a hard lesson learned.

[mackintire]

The breakdown

Ok here's where we are currently:

Drivers:
(4) Tang Band W5-1685 5" Underhung Midbass Driver
(2) Tang Band W4-1720 4" Underhung Midbass Driver
(2) Peerless 830880 Peerless 5.25" Mini Passive Radiator
(3) SEAS Prestige 27TFFC (H0881) Textile dome

HP and LP crossover 4th order LR at 3k to minimize the lobing effect.

This coming weekend I will have the crossovers redone in LSPCAD along with impedance and TR. I should also have the box modeling done in unibox, and WinISD.

These are the possible remaining issues:

* Port tuning , Should I be tuning for flat bandwidth or assuming room gain. A better question is which program should I trust. Unibox, WinISD, LSPCAD? Something else?

* Crossover Refinement (Once I post it...)

* impedance issues at FS < 4 ohms (I'm going to try to work this out over the weekend)

* driver variation (apparently the TB drivers may not have the same parameters that are listed on the spec sheet)

* Since the tower cabinet's dimensions are identical to CJD (Chris) MTM RS150/27TDFC towers, and the differences being the Tang Band drivers and a floor facing port, do you think I will need some sort of BSC for the tangband's? Someone elsewhere mentioned that the W5-1685 behaved like it has built in BSC.

* The center channel is going to be very different. Consider it experimental for now.

The plan is to get all the info together and post it here for some constructive feedback. If the design is still within budget, come March I will order the drivers and supplies to assemble the cabinets.

Since the design has elements from CJD (Chris) MTM RS150/27TDFC towers, and the Natalie-P's I'm naming these towers the Natangs

[ThomasW]

Sorry this change makes little sense to me. The underhung TB drivers appear to be very linear. If you're going for more output why not use 2 of the 5" or 6" units and avoid the hassle of a 3-way? The cost increase of using the 2 larger drivers is likely offset by the lower cost of a 2-way crossover.

Also those Peerless PRs are designed to match their little low Xmax midwoofer. To be safe you'd need 3 to work with a single TB. If you want to use PR's get a larger diameter ones with more Xmax.

[mackintire]

I think that you may have misunderstood. The Natangs is a 2way MTM tower with a Seas tweeter and (2) 5" underhung tangbands per cabinet.
Part of the design will be trying to use the tuned 1/4 wave resonance inherent of the box design in addition to the port tuning of the cabinet to order to extend the bass output.

The center will use the seas tweeter, (2) 4" drivers and (2) passive radiators.

[ThomasW]

Ok my bad, I just looked at the parts list...

As HB posted there's little point of using the PRs for a center with an 80Hz XO.

Since you're going with a higher XO point you might consider using one of the small diameter Neo magnet 1" domes for the center. Offset it up/down depending on where the center is going to live, that way you can pull the woofers closer together...

The Modula MTM center channel is an example of this. I'm only referencing this from a design standpoint, I'm not recommending you use the $200 tweeter in that system....

[mackintire]

I can look into it, but I would like to maintain the same staging and timbre across the front speakers.

This can be considers a bang for your buck type project with sizing constaints. Compared to some of the other lower end projects this one costs a little more, but size wise is dimensionally smaller than the other towers listed in the finished projects section.

Its a combination of price, performance, cosmetics (wood vineer) and size (happy wife).

Ideally this system will occupy similar space as the factory speakers that came with my Onkyo HTS800 on stands while sounding much much better.

This is all part of the happy wife requirement. (I'm sure you understand.)

The total cost for the older HTS800 speaker system, not including the reciever was $299.99 on the Onkyo website.

My goal is to replace the front three speakers for less than $600.00 in parts.

[ThomasW]

Since this is being designed from scratch you could use the same little Neo domes for all the speakers...

[mackintire]

Can you give me a recommendation on a Neo that's as linear and as well behaved + - 15° as the SEAS 27TFFC and in the same price ballpark.

[ThomasW]

We're going to redesign the NeoDcc using this tweeter since the Vifa D26 are discontinued

http://www.madisound.com/catalog/product_info.php?cPath=45_229_324&products_id=8177

http://www.madisound.com/catalog/product_info.php?cPath=45_229_324&products_id=1466

The low cost leader is ....

http://www.madisound.com/catalog/product_info.php?cPath=45_229_324&products_id=8355

We have a pair of these on order. Testing will be posted when time allows..

http://www.parts-express.com/pdf/264-865s.pdf

[mackintire]

Keeping in mind that the design is centered around 6 ohms I am not sure how any of those will work. I 'm also a little conserned and confused about the power ratings. 8 watts nominal, but can handle 80 watts max. Who came up with those specs and why? I can only guess that they can handle dynamics but not constant loud SPL?

I'm hoping someone can anwser my concerns as my options are open at the moment.

[ThomasW]

8 watts nominal, but can handle 80 watts max. Who came up with those specs and why?

One is continuous power, the other is transient peaks

[mackintire]

Well that puts those tweeters out of the running. The current system is THX compliant in this instance it means it can sustain SPL in excess of 104db.

I 'd rather not step backwards in any fashion compared to the crappy Onkyo speakers.

When running them in HT mode I was intending to run the Natang-S electronically crossed over at 80hz. When playing stereo music they most likely will be run full range. I truth if the end result does not play nice running full range I will settle for crossing them over at 60hz, but according to the freq model that shouldn t be an issue.

In comparison the Onkyo mains sound thin, laid back and have issues trying to reach 100hz at any decent SPL, the response also audiably falls way short of 17khz. I corrected a little of the factory arrangement by overriding the audassy EQ and crossing the mains over at 150, which helped but now I can localize the sub with ease. I am looking forward to the change to say the least.

[ThomasW]

Well that puts those tweeters out of the running. The current system is THX compliant in this instance it means it can sustain SPL in excess of 104db.

Personally I think THX specs are nothing but marketing bunk.

That said we're talking about 80 watts going into a tweeter that will output 91db with 1watt input measured at a distance of 1 meter, so 104 dB isn't a problem

[mackintire]

You're right of course...

Most people do not know or understand what THX is. The THX "standard" certification is a pretty loose standard which is why it doesn't mean much.

Still it IS a standard, so we can gain knowledge from understanding what that standard is.

However there are higher standards for THX, such as the Ultra 2 standard which is a bit more difficult to achieve. If you would like to know more here is an excellent article: http://www.hometheaterhifi.com/featu...ification.html


Just in case the less knowledgeable read this....

if 1 watt = 91db
then 2 watts = 94db
and 4 watts = 97db
8 = 100db
16 = 103db
32 = 106db
64 = 109db
80 watts = approx 110db
So max continuous power is 100db
Max transient power is somewhere around 110db

[mackintire]

Found a decent sub to match up with the Natang-S

Found a suitable sub. I purchased an Axiom EP175 off ebay for $70 shipped. The plate amp was removed, but I 'm going to replace that with a 500watt O Audio amp. I plan on stripping the sub and having it veneered as soon as it arrives. Considering the origional specs and the new amp. I should get a decent sub that plays 110db+ in room down to 30hz solid if not lower. I 'll have to play with the parametric EQ to see how it all ends up. Still for $70 I can't complain.

[mackintire]

To start it all off: here are the Unibox calculations at MAX SPL. I did not use the manufactuers parameters I used Zaph's instead.

Again keep in mind that the only time these will be cranked will be when they are being crossed over at 80hz by the receiver. They will however be used full range at levels up to 103db. Another note is that the port will be firing into the floor, the distance between the port and the floor will be 2.5-3.0 inches. At higher SPL levels the port will have a higher resistance. This is one part of the reason that I am not so worried about driver Xmax. Bass performance will suffer at higher SPLs, but that is a deliberate part of the design.

The crossovers are not included in these models.



UPDATE: Operating under 103db the port speed and Xmax are within tolerances.

[ThomasW]

Running fullrange @103dB the drivers run out of excursion at 77Hz. But I guess that's what you want.

[mackintire]

The previous pics are max SPL.

These are at 103db

Running out of excursion at 30hz @ 103db for reference

[mackintire]

I 'm playing around with it and I 'm thinking that the box has too much volume. Lowering the box volume down to 20l brings the Xmax back closer to where I want it. Running full range with a port tune of 42hz Xmax doesn t exceed until 30hz at 106db. I'll keep tweaking....

[BobEllis]

I'm curious. If you have a sub that integrates properly, why would you turn it off remove the HP on the mains for some listening?

I see a lot of people doing this, often for "music, the sub is for movies only." Many modern recordings have significant information below 40 Hz where many mains give up. I find almost all music sounds better with the sub. Late in the evening I just keep the volume down.

[mackintire]

The best anwser is probably the truth. The sub hasn't arrived yet from Colorado, so I'm not 100% sure of what the end result will be. The modeling looks good but I 'm not going to hold my breath until I hear it and get it tuned up.

[mackintire]

Does is sound reasonable that using Unibox the speaker xmax doesn't start tightening up until I lower the volume of the cabinet under 20l. Is this an error on Unibox's part or is the true VAS of the TB W5-1685 alot smaller as Zaph's measurement show?

So if the read VAS for the speaker is 6.1L then I would have to reach near 12L to achive a sealed box with as much stiffness as the designed suspension?

[---k---]

Found a suitable sub. I purchased an Axiom EP175 off ebay for $70 shipped. The plate amp was removed, but I 'm going to replace that with a 500watt O Audio amp.

Hopefully Axiom doesn't do any EQ in their plate amp. Many companies do.

[ThomasW]

Is this an error on Unibox's part or is the true VAS of the TB W5-1685 alot smaller as Zaph's measurement show?

Don't know.

Problems occur when tuning so far below the driver's Fs.

One wonders how you're going to limit the transient peaks going to the drivers....

Typically when a system is going to be run 'fullrange', SOP is to design close to the rated Pmax and raise the Fb until the excursion limit isn't exceeded. That provides inherent protection above Fb. Were that the case in this instance, the Fb would be high enough to require a high-pass filter set just below Fb to keep the driver from unloading, given the spectral content of music.

[mackintire]

I have two questions?


If I put an impedance comp network in to take care of the FS impedance change will the effect the Xmax at FS?


What if I used two impedance comp networks one for the FS peak and one for the port FS peak, both filters being narrrow bandwidth?

[mackintire]

OK, I ve back pedaled and gone to a Sealed design with a SCAN SPEAK FLOW RESISTANCE VENT

F3 is showing 66Hz at MAX SPL, I also had to apply a 1st order Butterworth HP at 45Hz to limit distortion. I should still have usable bass into the high 40's with this design while still being capable of running full range.

I 'll post the results as soon as I have them modeled.

[cjd]

FWIW...

Not sure I'd call the Axiom a sub, but it'll do a nice job filling out the low end.

My 38Hz tuned RS150 MTM's sound weak and thin to me any more... I guess you get spoiled...

Nifty looking project so far. More stuffing in the box (or a box size change) appears to be called for with the little rolled-up lip at the end of the graphs...

C

[mackintire]

I m getting annoyed at this point.

The SPEAK FLOW RESISTANCE VENT idea just, didn't work out the way I was hoping.

I have two variations worked out.

The sealed version can be driven to 106db with the F3 at 66Hz without room loading.

The ported version can be driven to 103db with an F3 at 30Hz, but will require a very steep X-over, something like 48db per octave, to maintain X-max below 30HZ at 103db.

Since I have to stay more or less in budget I am going with the sealed design. The good part of that is the QTC is going to be very close to idea. The bad, is that the design will not play as low as I would want. On the other hand... Asking a single driver to do so much is never a good idea as there is only so much that a single driver can do well without additional compromises.


I will post the seal version later this week or possibly this coming weekend. Hopefully the Axiom sub shows up by then. I planned on driving the sub direct and getting some in room measurements before I purchase the amp, just so I know what I am dealing with.

[ThomasW]

I m getting annoyed at this point. I have two variations worked out.

The sealed version can be driven to 106db with the F3 at 66Hz without room loading.

The ported version can be driven to 103db with an F3 at 30Hz, but will require a very steep X-over, something like 48db per octave, to maintain X-max below 30HZ at 103db.

The bad, is that the design will not play as low as I would want. On the other hand... Asking a single driver to do so much is never a good idea as there is only so much that a single driver can do well without additional compromises.

The best bet is to go sealed and always use them with a sub, since 'you can't fool father physics'....

[mackintire]

Good News the Axiom EP175 did not seem to have a internal EQ. If....When I blow the driver I 'll be replacing it with a Dayton RSS265HF-4 10" Reference HF Subwoofer, but all in due time.

The Axiom driver appears to be pretty solid down to 35 and only slightly down at 30hz. I 'm hopeful that the 25Hz subsonic filter will allow me to get a little more out of it.

The online reviews claim that it can play 110db+ at 38Hz and was down to 98db at 25hz, with the factory amp.

I've been reading Zaph's page and is helping me to better understand how phasing and crossovers should work properly.

Now I have a new question..

What opinions do you all have on the most reasonable equipment to purchase to do your own Inpedance testing?

[ThomasW]

It's a pain but it can be done with a frequency generator (stand alone or PC based), a DMM, and a voltage divider. It's easier to do using a PC with Arta or Speaker Workshop and a DIY test jig. Understand these freeware programs are more than a little user hostile and have minimal user support.

One plug-n-play option is Woofer Tester 3 from PE, problem with it is it's very limited functionality. Woofer Tester 2 is a better option if it's in the budget. (Neither of these options do frequency response plots)

BTW, having an impedance measurement is nice but it's not nearly enough information to design a crossover.

[mackintire]

Axiom EP175 Killed

The Axiom sub arrived....what a piece of crap.

The ports are tuned to 47hz.

20Hz was alot of air noise.

30hz was almost nothing -10db , If I closed off one of the ports I managed to get -3db from my testing refernce.

40hz was -6db down from my reference. If I blocked one of the ports I was at my reference.

They are flared on the outside but not in the inside. Atleast the cabinet was sturdy.



So I talked to my cabinet guy. We are building a 2.1 cf cabinet with (1) 3" precision flared port with a Dayton RS HF 10" subwoofer fed by a 500 watt oaudio plate amp. The port will be tuned to 22hz.

The modeling shows 108db at from 24hz- 150hz The F3 shows 20hz. All these numbers are without room gain.

That's alot of boom for a single 10" driver :twisted:

[mackintire]

Everything is still on schedule.

I'm working on the subwoofer and choosing what kind of finish I want on the entire set.

I 'll dive in and purchase the drivers early March. I may even have the baffles finished by then.

The center channel comes last.

[mackintire]

We are moving along, we just got a few sheets of MFD in and we are thinking about using cherry veneer. Subwoofer first. Towers second

[mackintire]

I was playing around with the center channel design today. After modeling a few different choices and comparing data on Zaph's site I've decided to use the Vifa XT19TD00-04 ring tweeter with a LR4th crossover at 2K along with the 4" underhung tangband drivers and (2) passive radiators. This should take care of most of the lobing issues and maintain the remaining design specs.

We will have pictures in March, so be patient.

We finally decided on a natural cherry with a brown stain for the baffle and the veneer.

I'm note sure if I will need baffle step compensation in either design. So once I get things partly assembled I will be taking measurements and seeing where everything ends up.

What this project is ending up to become, is: someone who wants the Natalie-P's but can not have speakers that large. The Chris's RS150 MTM design is a closer fit, but I want more SPL!?!? John Marshes Modular MTM is out of the price range. So... here we are.

In room the final outcome should be clean output in excess of 109db 22-20khz with a sub in a wife compatible package. Its a tall order, but I'm of the mindset that it's completely do-able.

I sure hope I can count on you guys to help me finalize the crossover once I get the measurements.

[mackintire]

Towers started

OK campers...

We are starting to assemble the towers now.

So here we are:

42.5" high
7.5" wide
9.5" deep

Dimensionally Identical to CJD's RS150 MTM as a sealed design.

Each cabinet has one Seas 27TDFC

Each cabinet has two Tang Band W5-1685 5" Underhung Midbass Drivers in series.

The Midbass spacing is identical to CJD's RS150 MTMs

(Location) These cabinets will be placed 12" from the walls.


I'm open to crossover suggestions at this point. I could model this after either the Natalie-P or CJD's MTMs

What are the pro's and con's of doing so.

On a side note it would be helpful is someone explained what is what in the Natalie-P's crossover network. I have figured out about 40% of whats going on in that circuit. I see the 3rd order highpass, something that looks like a RC baffle step comp in parrallel with a cap. I have no clue what the parallel LC is doing.

Any advice would be useful at this point.

[ThomasW]

The NatP crossover is one of Jon's weird series/parallel hybrid designs. It's specifically designed to kill the driver output quickly so as to avoid the metal cone break up.

A design like that is unnecessary for a driver like the TangBand.

[mackintire]

Thanks Thomas,

Unless someone has a better idea I was planning on 4th order LR crossover somewhere between 2.5k and 3K.

Are there changes that I can make to limit issues that may occur from having these cabinets 12" from the walls.

The TV stand is a corner unit. So the cabinets will be less then 12" from the walls but the walls will be 45 degrees in respect to the cabinets.