Pretty Persuasions - InDIY Coax Build Thread

[xmax]

Why? I just use off the shelf textbook filters designed for a resistor of 8 ohms with
exactly a Fc 3K for both drivers, anybody can do it.

No, I want you to go listen to some of my speakers. You might be pleasantly surprised what can
be done with filters designed for clean, coherent, natural and accurate reproduction.

In all seriousness I was offering my personal experience just in case the OP had not compared
the transfer function between the mini DSP and passive filters, I was actually quite surprised how much
easier it was to get flatter responses taking advantage of say the transfer function of Butterworth
alignments, playing with the Q and or minimal shaping circuits. Some of the dark magic that is
passive XO design that is not possible with Mini DSP (I don't care how much you EQ it, my opinion).
Because I have been doing it the old fashioned way for so long

[dcibel]

You might be surprised to hear when you come over that I already have a clean coherent natural sound and don't actually need your help.

[xmax]

Too bad, I'm going to help you whether you like it or not!

[JavadS]

Beautiful! The frame of that mcm woofer looks identical to the Tang Band W8-2096. What drivers are on the waveguide in the other speaker? Any luck with that blown amp?

Thanks! Yea I seriously considered that TB woofer, I’ve been breaking these in more and they are incredible for $40 so I’m feeling better about them, really nice bass from this 8” in this alignment. Not sure what blown amp you’re referring to?

[JavadS]

On another note, I'm sure you are probably aware of this and I imagine you are not going 48dB slopes in the passive domain. My experience with the mini dsp; the transfer function is nothing like the passive version of the same filter, E.I. a Butterworth filter will have no bump at Fc and more obviously the phase tracking is completely different. I only found it useful for acoustic center alignment and roughing in how low tweeters could go before sounding like screwdrivers in your ears and where cone break up could be avoided. Beautiful builds!!!

Oh definitely this active effort has nothing to do with the passive, passive is all 12db, I’m just waiting on some components, I’ll post more about that soon thanks!

[JavadS]

Pretty Persuasions passive crossover update, I have my initial measurements done, after getting my inbox FRD (frequency response out of Omnimic) and ZMA (impedance curve out of WT2) for each driver, I loaded them into Xsim and began some modeling. I modeled a 6db, 12 and 24db design with several variations an iterations of each, overall about 12 different models. 24db between tweeter and mid didn’t net any improvement over what I could get with 12, and 6db slopes just weren’t enough to tame a compression driver tweeter or mid that goes break up crazy at 2k, 12 gave me the best response and phase characteristic with the least number of components.

I am playing with a series (parallel components)LCR (EQ attenuation circuit) at 2600hz to bring down a hump in tweeter response, I can get really close w/o it but I’ll see how it plays in actual measurements I do later this week. As luck would have it I had less than half the parts I needed and since I could distract myself with some active crossover while I was waiting, I didn’t bother piecing together components.

I did wire up a 6db version just to hear the drivers initially, it sounded ok, then I did my active setup and that made the 6db version sound terrible, was an interesting back to back effort to see truly how bringing response and phase inline transformed the sound of the horn from honky and bright to smooth and balanced.

See attached charts for a little of what I’m seeing, as you can see some of these curves are quite challenging to deal with compared to the average hifi dome tweeter and mid but things fall in line nicely. Raw response of all drivers in box



12db response curves, note dip at 4K is in natural tweeter response



Tweeter reverse phase null



Xsim model

[xmax]

These are absolutely beautiful cabinets and I love the classic looking wide baffle, I think people are finally getting
away from the belief the baffle has to be very narrow or "small" for the speakers to have good imaging. I deal
with wider large baffles with some of my designs and the first comment I always get from people is the imaging
is like a "movie" or having the "artist in the room with you".


The curves are looking nice! Sorry I got you mixed up with another cat on the amp comment. I don't mean to offend anyone I have no
way of knowing everyones level of XO design and I'm trying to be informative for all readers.
One of the reasons I mentioned the difference with DSP and passive, it seems to be kind of a lost art,
but when modeling crossovers there is a lot of different classic "ratios" with just
second order like Bessel, Butterworth, Linkwitz Riley. I try to take advantage of this aspect whenever
possible. (people seem to really hate talking about anything textbook on this forum) I noticed when trying to mimic this
with DSP it gets it completely wrong. Somebody will say something about pure resistive loads or some other crap,
these observations were made with ribbon tweeters or drivers with some sort of impedance compensation.
Regardless of what others think I do believe there are advantages to keeping cap and coil ratios within
classical ratios to a certain extent to prevent "ringing" and other nasty artifacts. If nothing else it's
a hell of a starting point. If you don't believe me just do some calculations with say Zaph or
troelsgravesen.dk designs you will start to see they are more than not "textbook" especially
in the high-pass section. I don't think it is all coincidence...

[dcibel]

I'm sure you heard a song in the last 24 hours my filters were involved with.

Ah, you must be the one who mixed my SNFU and Dead Kennedys albums . Please do us all a favour and turn down the compressors so our hifi systems can actually reproduce the dynamic range that they are capable of.

Sorry Javad, I'll stop invading your lovely speaker thread now.

[xmax]

The closest thing I worked on to Dead Kennedys would be Erase Errata, but I'm no longer a recording/mix/mastering engineer.
What you (and I) have a problem with is digital limiting, not compression.

[dcibel]

Limiter is just a different compressor, but yes we are in agreement here. Anyway, turn the gain knob down a bit so the recording can contain some dynamic range. Don't worry, I have a gain knob of my own to turn it back up on my end ;)

[EarlK]

Nice! Javad,

One question;

Do you use amplifiers that are stable down to such a low impedance ( ie 2 ohms ) ?

[JavadS]

Nice! Javad, One question; Do you use amplifiers that are stable down to such a low impedance ( ie 2 ohms ) ?

I do, but that crossover isn’t finalized and far from it, that was my first initial model (hope that was clear), my goal will be to get it above 3ohms at that point which I’ve subsequently done in some of my initial crossover testing. Thanks! Javad

[KEtheredge87]

......bringing response and phase inline transformed the sound of the horn from honky and bright to smooth and balanced.

Hey Javad... This question may border on the subjective, but can you help describe "honky" a bit more in this context? I've heard of horn honk before, but never really been able to associate that with a sound. It just makes me think of an old cartoon bicycle horn or something.

Thanks!
Keith

[Wolf]

You're not far off. It adds that honky coloration or 'forced' sound into the lower female register for sure. When I dialed the resistor value down recently on Steve's "Jaws" in the horn's LCR, it was clear that the honk had subsided using Jennifer Warnes.

Later,
Wolf

[JavadS]

Hey Javad... This question may border on the subjective, but can you help describe "honky" a bit more in this context? I've heard of horn honk before, but never really been able to associate that with a sound. It just makes me think of an old cartoon bicycle horn or something. Thanks! Keith

The issue is probably more related to the Horns boosting the frequency range in the 1 to 4kHz then it is the sound of the horn itself, but yeah that’s the sound you get from a speaker when it’s too hot in that range, it’s Sylmar to the sound of an old clock radio something that doesn’t do you any trouble or anything below 500 Hz. That’s a tough range ain a horn to control as your typical dome tweeter is quite flat thru that range, if you can get it right then you can neutralize a lot of the sound of what people don’t like about tweeter horns in general.