First VituixCAD Design - Please check

[zinger084]

Hello,

So I began this thread somewhat of a continuation of the previous but would like some constructive comments on my design.

Again, I'm using the following:

https://www.parts-express.com/Peerless-DA32TX00-08-1-1-4-Corundum-Dome-Tweeter-264-1678

and one of the:

https://www.parts-express.com/Dayton-Audio-DS270-8-10-Designer-Series-Woofer-295-432

I want to make sure i'm not missing anything major on the design. I think this is the best I can do with the crossover and RLCs, but I also want to make sure my impedence curve isn't too wacky because i know that's important (what's the limits i should stay within, etc?

Thank you all!

[PWR RYD]

You don't have any baffle step loss correction so these will sound really thin unless they are going to be mounted on the wall. The woofer's breakup is only down 15 or so dB. That will be audible.

That is a great tweeter that can be crossed pretty low, but not that low.

Constructively I would say you should adjust (increase) the woofer's inductor and play with the shut capacitor to add some BSC. Might have to add a resistor in series with that shunt cap to soften the knee. Not sure if you need the RLC. You might get away with just a bottomless trap (small cap in parallel with the inductor).

The crossover point will need to be raised. I was able to cross that tweeter at 1400 Hz with a 4th order LR acoustic slope. That required a 3rd order electrical filter with the second cap being much larger than the first.

Hope this helps.

[Ugg10]

Are these your measurement or traced/downloaded from the manufacturer ?

If traced/manufacturer data then these are measured on the axis of each individual driver so you need offsets in at least the Y and Z directions - use 0,0,0 for the tweeter and the woofer is likely to be something like 0,-200,+35 at a guess (Z is +ve away from the listener which is different to other modelling systems). This will affect both the frequency plot but also the phase trace (top right graph). Ideally the phase plots of the two drivers should be coincident (on top of each other) fro at least 1 octave +/- the crossover point e.g. crossover at 2khz, co-incident from 1khz to 4khz. Also the phase plot fro the tweeter looks a bit off, may be worth re-tracing the two data set (assuming thats what you did) using the Vituix tool and seeing how it looks.

Also agree that crossing over the tweeter at 900hz is not a good idea and 1500-200hz is more usual. Plus a 10" woofer will start to beam pretty low (around 1500hz maximum) so again this may not be the best combination. You may be better off looking at the RS225 which is a very good driver but has a larger cone breakup I think, lots of project around using this. If you want a ready made solution the RS225+10f8424 by XRK971 on DIYaudio is a good place to start with at least three box options (sealed, bookshelf TL and floorstand TL).

Hope this helps.

[Chris Roemer]

I can't confirm any of your component values (too tiny).
A "nominal" 8n(ohm) design will typically show a Zmin around 6n.
If you used a "4ohm stable amp" you could run the minimum down to about 3ohms.

In "general", you'd like to avoid a "hump" in the filter before it rolls off. BOTH you filters (HP & LP) show a rise before they roll off (this is "hi-Q" and generally won't sound all that great).
(I haven't used Virtuix Cad, yet, but) It looks like you are ignoring the driver offsets (X, Y, Z values - yours are at 0,0,0). This is not trivial.
The offsets determine phase, which determines how the drivers "sum" around the crossover point. That means that you have to generate "minimum phase" files - NOT the same as simply having a "0" in the final column of your F/Z files. (You'll end up with zeros when you trace.)
PE-supplied (Dayton) files most likely are not min-phase either. VCad must have some way of generating min-phase files. It's just a step you're not aware of, and are therefore skipping.

Did you read Paul Carmody's "sticky" I told you about?
Did you play w/Tolvan's "Edge" software to see what you're trying to accomplish with BSC (baffle-step)?
Did you try MY (suggested) XO?

[zinger084]

You don't have any baffle step loss correction so these will sound really thin unless they are going to be mounted on the wall. The woofer's breakup is only down 15 or so dB. That will be audible.

That is a great tweeter that can be crossed pretty low, but not that low.

Constructively I would say you should adjust (increase) the woofer's inductor and play with the shut capacitor to add some BSC. Might have to add a resistor in series with that shunt cap to soften the knee. Not sure if you need the RLC. You might get away with just a bottomless trap (small cap in parallel with the inductor).

The crossover point will need to be raised. I was able to cross that tweeter at 1400 Hz with a 4th order LR acoustic slope. That required a 3rd order electrical filter with the second cap being much larger than the first.

Hope this helps.

I appreciate the response. I'm looking into how to address the BSC in VituixCAD now.

Its definitely a leaning curve with items like "the knee" and "bottomless trap" but i'm working to get there understanding all the lingo...

[zinger084]

I can't confirm any of your component values (too tiny).
A "nominal" 8n(ohm) design will typically show a Zmin around 6n.
If you used a "4ohm stable amp" you could run the minimum down to about 3ohms.

In "general", you'd like to avoid a "hump" in the filter before it rolls off. BOTH you filters (HP & LP) show a rise before they roll off (this is "hi-Q" and generally won't sound all that great).
(I haven't used Virtuix Cad, yet, but) It looks like you are ignoring the driver offsets (X, Y, Z values - yours are at 0,0,0). This is not trivial.
The offsets determine phase, which determines how the drivers "sum" around the crossover point. That means that you have to generate "minimum phase" files - NOT the same as simply having a "0" in the final column of your F/Z files. (You'll end up with zeros when you trace.)
PE-supplied (Dayton) files most likely are not min-phase either. VCad must have some way of generating min-phase files. It's just a step you're not aware of, and are therefore skipping.

Did you read Paul Carmody's "sticky" I told you about?
Did you play w/Tolvan's "Edge" software to see what you're trying to accomplish with BSC (baffle-step)?
Did you try MY (suggested) XO?

Thanks for the heads up on the Q... i missed that bump.
Yes, i've read the website and watched some learning curve (or cliff) videos. I'm attempting to model the BSC in VituixCAD now.
Yes, i'm attempting to build yours right now in VituixCAD and i'll see how it goes.

[zinger084]

Are these your measurement or traced/downloaded from the manufacturer ?

If traced/man.....

I downloaded the Dayton Audio woofer files and used Vituix for the trace on the spl and impedance charts. While i think i did them right, i'm not sure. One thing i'm sure of is how hard is it to learn this software without having someone in the room teaching you! I'm also working on the diffraction tool in Vituix, but i can't seem how to "create" the baffle itself once i have everything in there (do i need the half space resonance file to make this work???)

[zinger084]

I can't confirm any of your component values (too tiny)....

...Did you try MY (suggested) XO?

If i interpreted your posts correctly, then the Vituix build from your values can be found here (and something isn't right to me...)

[DeZZar]

That 1.5uf cap and 4ohm resistor go across the first 2.5 inductor...in parallel with it. The 8ohm and 10uf cap go from + to - ahead of the woofer, not in series with it.

[LOUT]

If i interpreted your posts correctly, then the Vituix build from your values can be found here (and something isn't right to me...)

That resistor and 10uF capacitor at the far right should be somewhere else...like next to the 4ohm and 1.5uF (placed in the same way, but farther right or left)....OR the 4ohm and 1.5uF should be attached to the 2.5mH inductor (like a little circuit that bypasses parallel to the inductor) while the 8ohm and 10uF cap go in the place you currently have the 4ohm and 1.5uF cap.

[zinger084]

corrected... i think this looks right. however, i''ve still got a pretty good impedance around 800... any reason to be concerned? i'm also not seeing as flat as line as Chris uploaded on page 2... not sure what's up...??? hump

[DeZZar]

Chris has included BFC in his recommendation. But as you can see this is a classic example of why real measurements and proper modelling are required in the theoretical domain before you even decide to purchase drivers and build a box, in order to achieve the right results.

No online calculator is able to anticipate the misbehaved woofer here and its lumpy 1K characteristics - or the choppy bottom end of the tweeter particular when diffraction is modelled in.

[LOUT]

I'm guessing it should be like this:


EDIT: sorry, yes, you nailed it that second time there. Nice.
Maybe Chris's's's' response is flatter because it has the BaffleStep figured into it, and I'm guessing you're still having a bit of a battle VS the baffle-step simulation (so it's not in your graph)?
I think your impedance looks fine at 800hz. I might personally change the 4ohm resistor (on the 4ohm and 1.5uF thingy) to a 16ohm instead to keep the impedance high above 10Khz, but I think that only matters for some really finicky amplifiers that don't protect theirselves up high. And it also shouldn't matter either way if your amp is fine at 4ohm, obviously.

[Ugg10]

For the y axis position it is usual to base it on the listening position. Therefore if you intend the speaker tweeter to be at ear height then this will be 0,0,0. The woofer would then be 0,-334,30. The z offset is the distance between the tweeter acoustic centre and the woofer acoustic centre which is approximately in the plane of where the dust cap meets the cone therefore behind the tweeter by approx 30mm. Vituix uses a positive value for rearward dimensions in Z which is different to other modelling packages. This may help with the phase plot. Also worth setting a realistic speaker position in the room calculator as this again will affect the phase plot a little.

[DeZZar]

I'm attempting to model the BSC in VituixCAD now.

This will give you a full set of diffraction modelled response files for your woofer. Start with the FRD file for on-axis (0 degrees), setup your baffle and driver position as follows...

[DeZZar]

Just an extra tip with this - I generally don't decide on the tweeter position and build the baffle before modelling diffraction. IMO tweeter position is critical to producing a smooth result that doesn't create headaches for the crossover. You'll notice that the diffraction effects on the tweeter change significantly as you move a tweeter around on the baffle. Edge radius also plays a major part in this. When designing a speaker system from scratch expect to go back and forth between baffle dimensions, speaker positioning and crossover numerous times...

[rpb]

I downloaded the Dayton Audio woofer files and used Vituix for the trace on the spl and impedance charts. While i think i did them right, i'm not sure. One thing i'm sure of is how hard is it to learn this software without having someone in the room teaching you! I'm also working on the diffraction tool in Vituix, but i can't seem how to "create" the baffle itself once i have everything in there (do i need the half space resonance file to make this work???)

Is there some reason you could not use the frd, and zma files without tracing?

[zinger084]

Is there some reason you could not use the frd, and zma files without tracing?

The Peerless DA32TX00-08 Tweweter didn't have frd or zma files that I could readily find... maybe I'm missing someones secret???