MW16P-4 vs MW16P-8 & TW29R vs SB29RDC-C000-4

[ernperkins]

I think I have decided on the Satori TW29R-B to go with the MW16P for this project for several reasons:


I tried a layout that takes a bit of a design idea from Troels Gravesen's Illuminator Monitor suggested by Linesource. I may just stick with the original layout.

I think you'll be happy with the result. And you won't be constantly asking yourself how much better the TW29 would have sounded!

 

[rscamp]

...And you won't be constantly asking yourself how much better the TW29 would have sounded!

True!

This is a very small scale project I'm working on to try out making curved side speaker cabinets. If it works out well I'll try using the same approach with these drivers.

 

[GreatLaBroski]

I think you'll be happy with the result. And you won't be constantly asking yourself how much better the TW29 would have sounded!

I've owned both those drivers (neo version of the TW29 though) and they're outstanding. You'll be happy.

 

[rscamp]

I've owned both those drivers (neo version of the TW29 though) and they're outstanding. You'll be happy.

I hope you are right. I just put in the order!

 

[GreatLaBroski]

I hope you are right. I just put in the order!

I'll hang around the thread so you can have an opportunity to call me out by name if I'm wrong.

In some ways, my Satori drivers (TW29RN, MW16P-8, MW19P-8, MR13P-4) impressed me more than my ScanSpeak drivers (12m, 12mu, 18wu, 32w/4847t00, D3004 BE) due to the bang-for-the-buck factor. I expected the SS drivers to be class-leading, and they lived up to my expectations, but the Satoris were a real shocker as to how good they are.

The Satoris are 1/2 the price and yet trade blows with the SS drivers. I'd probably prefer the Satori speakers for passive crossover based speakers due to their wide bandwidth and overall good behavior which I'd expect to simplify the build. And the value is great.

 

[LineSource]

If it works out well I'll try using the same approach with these drivers.

If you are committed to translam construction, you should also review a few "big rounds" cabinet shapes before you start your Satori Build. A standard 2-template method can be used. Panel cut time should not be different, just more waste will be generated. Covering the sides is probably easier than sand and finish.

Vapor Audio has Genius Translams with high functionality...
-large radius on baffle edges for low diffraction
-tapered rears for resonance reduction
-integrated 3D bracing
-pockets for wool absorption reduction
-pockets for sand vibration reduction

 

[rscamp]

If you are committed to translam construction, you should also review a few "big rounds" cabinet shapes before you start your Satori Build. A standard 2-template method can be used. Panel cut time should not be different, just more waste will be generated. Covering the sides is probably easier than sand and finish.

Vapor Audio has Genius Translams with high functionality...
-large radius on baffle edges for low diffraction
-tapered rears for resonance reduction
-integrated 3D bracing
-pockets for wool absorption reduction
-pockets for sand vibration reduction

Thanks for the info! For now I'm just bending layers of thin sheets over the formers.

 

[rscamp]

According to parcel tracking, the order will arrive today.

Hi,
Yes a dsp can do many things other than pure filtering duties.
My point about your case is that in the case your vented 2 ways is difficult to cross to your sub one possible option is to close the port and use an LT to make things easier.
My comment about low end extension is a worst case scenario: you may (or not!) need to increase extension in the low end then in that case the trade off are the one i stated earlier ( this was what i faced for my own needs) BUT the point of an LT is to change the whole system ( closed box AND driver) fc and Q.
So it works both ways, you could use it to make fc higher with a different Q. In this case there is no trade off and you may gain in overall spl ( to the xmax limit or power dissipation of the driver).
As well it can open some possibilities to use the acoustical properties of the box+driver combo to achieve a target of high pass defined.
Example you could make a Samuel Harsch filter solution for the high pass required just using the LT to have the butterworth 2pole target.
In fact it open many possibilities and different options ( it could help in a 1 pole filter too, trying to compensate for the behavior of the driver outside of its intended usable range then include a filter to have a theorical text book response... as long as you don't try to compensate things which are non minimal phase( as break up of membrane, some diffraction, geometry related issues,...) it works quite well ime.
But for your specific case you'll have to contemplate what works for you in your room and for your expectations...
All that to say dsp filter is quite flexible and it can open ways you haven't thought about...

I just started to investigate this. Are you talking about this miniDSP biquad functionality:

Linkwitz Transform

and this sort of crossover:

https://www.diyaudio.com/forums/multi-way/277691-harsch-xo.html#post4402917

?

 

[krivium]

Yes minidsp have a Linkwitz Transform biquad fonction ( you can apply it using other ways, usually it use one or two parametric eq band, sometimes more...).
You can simulate it using winISD ( the freeware have a dedicated tab in the eq/filter section and it will give you valuable simulation including xmax and power requirement). I suggest you try it (to have an idea of limitation and compromise).
You could try simulate here too:
Calculs d'une enceinte close avec ou sans transformée de Linkwitz, 1/6
It is in french but you should understand what it is about. In case of i can help.
Yes for the crossover too. Xrk did an extensive use and study of it. For your case i suggest you stick with the definition he gives in the first paragraph (below the first OB picture). That said you may not be after a 'quasi perfect transient' crossover type.

 

[krivium]

How high in freq your sub are acceptable quality wise?

 

[rscamp]

Yes minidsp have a Linkwitz Transform biquad fonction ( you can apply it using other ways, usually it use one or two parametric eq band, sometimes more...).
You can simulate it using winISD ( the freeware have a dedicated tab in the eq/filter section and it will give you valuable simulation including xmax and power requirement). I suggest you try it (to have an idea of limitation and compromise)...

WinISD FR plot attached. The excursion is scary so it looks like this must be used in moderation to extend low frequency response. The transfer function looks good. As a first thought, might it be best to make a critically damped enclosure alignment and start from there?

 

[rscamp]

How high in freq your sub are acceptable quality wise?

Well that's a good question. I don't currently know!

The crossover is at 90Hz but this will be re-evaluated when I start using REW. There are two subs. I have some room treatment (a Helmholtz passive absorber for example) and have spent hours and hours finding good locations for the subs with respect to obtaining smooth audible LF chirps. The re-positioning effort resulted in a large improvement - so large it seemed like I had just invested $$$ in a new system. It was amazing how this alone improved the overall sound - seemingly not just the low frequencies. But anyways, the result is the subs are not close to the main stereo pair and will be localized if I cross over too high and/or have a shallow low-pass slope.

 

[krivium]

Rscamp,
Ok you discovered the limits of LT in case of low end extension which is a worst case scenario.
Now lets contemplate this case: a satellite and sub use satellite down to 80hz and sub from there.
There is reason for that: below 80hz we have difficulty to locate sound ( and this is even more true in a room and given you are in modal zone it can even be true up to 150hz ime), so if you can locate your 'satelite' within 1meter from the sub center* everything should be fine ( as we are in the enviable position to be within 1/4 wavelength center to center where there is acoustical coupling, sub and sat behave as one and only source).

So try to simulate the satory 7" for 80hz and 0,577 Q and check for spl and excursion. You'll see this should give you much less scarying results and this will be all needed for the high pass part for a Harsch solution ( you'll need a bit of delay for your 'sat' and a 24db butt lp on the sub). The 60hz 0.5Q example is the response if you block your port on the 17l box without a LT.

* imagine a sphere 1 meter diameter from center of your sub: this will give you the possible locations for your 'satellite'.

 

[rscamp]

Rscamp,
Ok you discovered the limits of LT in case of low end extension which is a worst case scenario.
Now lets contemplate this case: a satellite and sub use satellite down to 80hz and sub from there.
There is reason for that: below 80hz we have difficulty to locate sound ( and this is even more true in a room and given you are in modal zone it can even be true up to 150hz ime), so if you can locate your 'satelite' within 1meter from the sub center* everything should be fine ( as we are in the enviable position to be within 1/4 wavelength center to center where there is acoustical coupling, sub and sat behave as one and only source).

So try to simulate the satory 7" for 80hz and 0,577 Q and check for spl and excursion. You'll see this should give you much less scarying results and this will be all needed for the high pass part for a Harsch solution ( you'll need a bit of delay for your 'sat' and a 24db butt lp on the sub). The 60hz 0.5Q example is the response if you block your port on the 17l box without a LT.

* imagine a sphere 1 meter diameter from center of your sub: this will give you the possible locations for your 'satellite'.

There are a couple of potential problems. I have a combination stereo/HT setup and I suspect the LP filter on the AVR for the subs is 4th order. I also can't position the subs within 1 metre of the stereo pair.

 

[krivium]

There is no problem for the lp to be 4pole (in case of Harsch setup as this is what is needed for the lp on sub).
1m is text book once your speakers done done you could try it to see if you experience issue or not.
But this was just a suggestion from my part ( Lt and or Harsch filter) , there is many ways to skin a cat...

 

[rscamp]

There is no problem for the lp to be 4pole (in case of Harsch setup as this is what is needed for the lp on sub).
1m is text book once your speakers done done you could try it to see if you experience issue or not.
But this was just a suggestion from my part ( Lt and or Harsch filter) , there is many ways to skin a cat...

No problem. I really appreciate your input! It helps me discover new possibilities.

The shipment just arrived! I now have most of the items needed to build and test this project. I have to be disciplined though and finish off my current speaker project or it won't get done. I'll be able to play with REW a bit while the glue/paint is drying.

 

[rscamp]

Well these practice speakers ended up sounding better than I expected.

I call the finish "MDF au naturel".

 

[Bigun]

any progress ?

 

[rscamp]

any progress ?

Yes! The pair is working subjectively well so I sort of paused the development to make time for other projects around the house. I need to objectively confirm and tweak with better measurements. Basicly what I have at the moment is 2nd order LR crossover at 1900Hz and baffle step superimposed with a bit of a Harmon curve.

 

[its bh]

Hi rscamp - great write up. Really enjoyed reading through your thread. RE: crossover.. Did you design this yourself or model off something else? Am currently in the process of planning a build using MW16P with the SB29R and am keen to avoid designing my XO from scratch. Would be super grateful for any help you could provide.
Thanks heaps!
Brenton