Advice on XO MTM WR125S and 27TDFC?

[Matthew P.]

Hi There,
I've been playing with an MTM design with a SEAS 27TDFC and the CSS WR125S and wondering what people think about the crossover design. So far, this is just modeled, but I'd like a good starting point from which i can measure when i get that set up at my place.

I"ve aimed for a LR4 acoustic crossover point and seemed to achieve that with an LR2 electrical. I was trying for the 3rd order butterworth, since that seems to be a standard MTM design XO, but i couldn't get a terribly flat response. The crossover point is 2000hz. I don't need an L-pad since the tweeter sensitivity is equivalent to the WR125S in parallel. The impedance compensation network is not needed because of the slight change in impedance through the crossover point. I chose the XO point of 2000HZ since my speakers right now have this, and i can cannibalize some of the parts, unwind the inductors to the appropriate length and don't have to spend a lot. I've been working on this in speaker workshop, so if anybody wants the file to examine and tweak, i can post that. The distance between my drivers is 11.25" in a cabinet that is 8 X 11 X 18. I haven't modeled in the cabinet yet, so don't have diffraction or dispersion data.

In the file so far, it shows a big hump between 100-200hz, but i wasn't sure if i should look at that or not. there is some jaggedness in the FR at the upper end of the XO point - between 1800-3000, but a series notch filter does not seem to help. but it might need some contouring. I don't want to introduce too many components for cost reasons, but it seems that the network doesn't need that many in order to get a decent response.

anyways, I'd very much appreciate any comments, criticisms, suggestions, and wisdom since I am very wet behind the ears with this stuff.

Thanks,
Matt

http://picasaweb.google.com/matthewpunger/MTMSeasAndCSSDesignFiles

sorry i wasn't able to post pictures for some reason, so the link above will take you to the XO and response graphs. when i am able to post the SWD file on a web page I'll do that as well.

[Matthew P.]

Here's the link to download the SWD file for the MTM design. It's a little large and the server is slow.

www.ualberta.ca/~matthewu/cssseasmtm.swd

thanks again,
Matt

[Matthew P.]

Ok..sorry, it seems I can't edit my posts, otherwise i would put all these in the original. so I've made some adjustments and the new file thats posted above seems to be quite a bit better. In the end, I added impedance compensation, and a series notch filter to help even out the bump at the XO point and raise the overall response, along with tweaking the figures of the other crossover bits. It is now also phase aligned.

Does anybody think that this is a good place to start, in the sense of it being worthwhile to start experimenting with in the enclosures using this as a base XO? I would prefer to not use so many components, but this has given me the best response out of all the things I've tried.

any advice is very appreciated,
Best Regards,
Matt

[sreten]

I don't need an L-pad since the tweeter sensitivity is equivalent to the WR125S in parallel.

Hi, yes you do if the sensitivities are equal and you have some BSC, /sreten.

[Matthew P.]

Hi Sreten,
Thanks for the heads up...I've been trying to account for BSC, but have gotten a bit confused concerning seemingly different theories out there. There's one that is based on the F3 of the baffle, which means that I need to account for a possible 3 - 6db gain at about 530hz. On the FRD tools baffle step simulator, the rise starts about 200 hz, peaks with a 7.5 db gain about 1200hz then evens out with a 5 db gain around 2500hz. also, my room is significantly less than ideal for audiophile listening, and will be listening off axis almost 99% of the time (since i live in an apartment) and the reflections I'm guessing are pretty wicked.

when modeling then, how should I compensate for all that?

thanks,
Matt

[sreten]

Hi,

your modelled BS sounds right, try following these methods :

http://www.rjbaudio.com/Audiofiles/FRDtools.html
http://www.geocities.com/woove99/Spkrbldg/

/sreten.

[Matthew P.]

Thanks Sreten,
Those websites are very helpful.

I think I've found a very simple crossover that's just a LR2 electrical without the extra filters, but it models really nicely that would appear to account for BS. It has slightly rising top end, but that might be good for off-axis response. The DCR of the inductors are accounted for in the diagram.

C1 = 9.5 uf
C2 = 30uf
L1 = .95mh
L2 = .7 mh
R1 = .5 Ohm
R2 = 1 Ohm

Sorry if these are too big,
what do you think of this response and XO?

thanks for the help,
MAtt

[Jay_WJ]

Do you know what kind of baffle was used to get the FR measurement you use now? Without knowing it, you can't model your xo properly.

[Matthew P.]

Hi Jay,
The measurements are from Zaph's test on the WR125S and RJBaudio's SEAS 27TDFC. Both are done on infinite baffle.

looks like I have a bit more work to do...

I'll report back once I have subtracted the infinite baffle and added my own enclosure simulation.

thanks for the help,
Matt

[Jay_WJ]

Where did you find Zaph's WR125S measurement? I can't find it on his web pages.