5 way horn speaker system project - tapped, bass, mids and tweeter passive active

Attachments

  • 3D29378B-CACF-44CD-97B9-9B8569089321.jpeg
    3D29378B-CACF-44CD-97B9-9B8569089321.jpeg
    182.7 KB · Views: 332
Here is a REW measurement I did on the JMLC-Iwata 220 with Umik-1. I am pretty satisfied with that!
Apparently the Iwata-300, the little brother to this one, also goes down very low, but is less good for midrange and treble. Well, I knew that, so I am using a smaller horn for everything above 2500Hz now, but soon I will use another set of compression drivers in a Iwata 1400. Maybe I should go round horn, but I just love the shape of these.

And here is what it sounds like on my cellphone:
More V-Fet goodness. - YouTube

Meant Iwata 600, not 1400. Will be crossed at 1400Hz.
 
I could, but thw difference of running it at 270Hz VS 440Hz is like having the cello in the middle of my livingroom VS having backrow seats.

Yhe secret is to use 48dB Highpass at 270Hz.

About the Iwata 600, I have not received it yet. Maybe I like it better at cutoff, who knows.
 
Last edited:
Some serious wood working skills to say the least. Maybe you can concur on some perspective I've received about using large horns for the tweeter horn...I was told that over about 350hz size horn, it starts to sound as if the voice is coming from within the horn, instead of out front, sort to speak....after reading your testimony about Xo vs fc I am starting to wonder if the gentleman who made the comment was not victim of playing a horn too low via low Xo without the help of a steep slope, like the 48db you suggest.
 
The steep slope helped with to push the curve down to 270Hz, and then it dropped like a stone, just like I wanted, and anything less than 48dB crossed much higher even though I set the crossover point at 270Hz. The gentlemen was maybe listening to round horns. :p

It is an effect more than it is correct audio reproduction. When ACDC are playing maybe you don't want enough resolution to here the ganitor drop a pencil in the next building.
 
  • Like
Reactions: 1 user
Once again, not all experiments yield long term proven benefits.
What I sometimes perceive as something worthy, new and interesting, in the long run gets reversed[emoji846]

This week I don't know why but I started questioning the dynamics of the upper mid Vitavox S2 channel.

The introduction of the Le Clèac'h horns took more time to adjust to perhaps than I'd reckoned on. They are still king over the tractrix..

You may recall I used 3uF caps on the S2s.
These miraculously smooth the freq hump at around 2KHz and give a beautifully smooth 1st order cross in.
I've tried various caps, trialled some expensive ones, have a pair of superior Clarity caps, but kept going back to the raw, less smooth, more viceral, cheap and cheerful Monacor caps.
On paper these should sound inferior, but they don't.

Anyway, I swapped from Najda doing the X/O to being OFF, and the caps doing their thing. Najda still does the time delay for alignment.

Bam! That sound is back.
No amount of playing with X/O points and slopes to seemingly do the same in Najda, sounds quite like it.
Powerfully involving and engaging, draws you in with its subtleness, delicacy, yet raw dynamics and speed when needed.

I tweaked the X/O point of the JBL mids a little to my taste too.

I do like the flexibility of this system for changing stuff and sometimes reverting.

Perhaps I should try passives on the JBL mids (bit more involved to do).
I probably have all the caps and inductors needed for 2nd order..
 
Streaming problems and Qobuz replaced by Tidal - at least for now...

I've been using Qobuz for 3 years.

Most of that with a RPi + Allo Kali reclocker, Volumio and feeding native I2S into my dacs (Najda)

Reccently trialled a couple of off the shelf Streamers.

Ifi Streamer

Screenshot_20231017_163637_com.quoord.tapatalkpro.activity~2.jpg


and Sonore Ultrarendu.

Screenshot_20231017_163618_com.quoord.tapatalkpro.activity~2.jpg


The latter with a Gustard Reclocker.
Screenshot_20231019_133621_com.amazon.mShop.android.shopping~2.jpg


Screenshot_20231019_133437_com.android.chrome~2.jpg


The Ifi had their purifier in already!
Certainly an upgrade from the RPi.
More resolution and dynamics.

I settled on the Ultrarendu etc over the IFi. Just sounded more me!

Only trouble is that I've been having problems with what appears to be the Ultrarendu hanging up.
It's documented elsewhere,
Won't start playing a track, hangs at 0.0 seconds progress flashing, nothing playing. When viewed in the Playing now screen in Bubble UPnP or mConnect for that matter.

The Dac light is permanently on.

Exiting and force stopping Bubble no help.
Reboot Untrarendu alone, no good.

Have to reboot Ultrarendu and switch the Gustad reclocker off and on again.

Then all returns to normal.
I've tried all the settlings.
Proxy in both places.

Still happens about once a listening session!

I'm on V2.9 of Ultrarendu SW. They sent it.

Sonore have been responsive.

I've sent Dac diagnostics, screenshots etc.

I'm still gathering data.

I thought going down from max 24/192 in Qobuz settings was helping, but nope!

24/96 same problem from time to time.

Usually it's when I load playlists, ▶️ all and then skip about a bit!

Anyway, my Qobuz sub was up and I thought I'd try Tidal (hi-fi only).
No MQA for me.

Since using Tidal it's not happened.

The tonal balance of Tidal is slightly more restrained in the treble, I found.

I actually cranked up the Upper mid (Vitavox S2s) tweeters (Raal lazy ribbons) a dB or two.

Since doing that I've no issues with the SQ.

Sounds nightly fine actually.

Let's see how long it lasts!
 
Last edited:
  • Like
Reactions: 1 users
You might recall I've been using the upper mid 1200 - 11000Hz'ish channel on a 1st order passive (3uF cap), for years and years.
Najda DSP does signal (of course) and time-alignment.

Today, a rainy day, curiosity got the better of me!
I made a 2nd order bandpass X/O for the mid channel, 350 to 1200Hz'ish.

I had the inductors and caps necessary squirreled away since the early all passive days. Back then I used Bessel - peaky, nowadays I favour L-R.

Lash-ups made on breadboards, X/Os in Najda removed, polarity reversed.
I've been listening..

First thing is that the volume is louder - same thing for the upper mids on cap vs Najda X/O.
Never got to the bottom of why that is?

Once reduced by a couple of dB, I really could not tell the difference between active X/O and passive.
Both excellent dynamics, control, tone, blend in etc.

I haven't checked the time alignment yet. Probably out by some small ms number..

So at the moment Najda does the time alignment for all, does the 6th order L-R X/Os on tapped horns at 90Hz, 2 room mode PEQ cuts on the left channel and 3 on the right channel.

On the mid bass Najda does 2nd order L-R 90Hz and 350Hz bandpass

The other 3 X/Os are passive.

I'm slowly migrating beck to all passive😂

Stumbling block would be time alignment for all and the PEQ cuts on the tapped horns.

I guess a DSP amp could take care of the X/O and PEQ?
Could physically time align everything but the tapped horns. Did that way back before Najda.
Wonder if that would be actual stumbling block to all passive, or rather non central DSP?

The Crown XLS-1002 amp I have on mid bass could do the X/O as it is.. maybe I'll try that next, just to see..
 
This is where I am. But I opt for 1st order passive above 500Hz. I tried a 17uF cap on my JA6681B in a Iwata 300 from about 600-11000Hz and I liked it better than my attempt at 2nd order. Sounded much cleaner.
I think below 500Hz I won’t bother with physical time alignment. I never heard a difference when playing with the delay on my minidsp. But above that, in theory, I can get problems with the midrange cancelling out or adding to the tweeter horns, especially super tweeters eith high crossover point. But I suspect this is more of a problem for direct radiators, because when I asked Mr BA to measure his horn system with or without the supertweeter at from 8 kHz we did not see anything negative. I hope I get the same result.
To experiement with time alignment on my tweeter horns I have prepared a simple slide on a metal shelf. This is how it looks like now.
Midbass horn is under development. I will make do with four Seas W26FX002.
 

Attachments

  • IMG_9250.jpeg
    IMG_9250.jpeg
    372.9 KB · Views: 127
  • Like
Reactions: 1 users
Yes, I think it's possible to over do the slopes between some drivers / horns.
I experimented with steep slopes years ago,, just to hear them, but find 2nd and 1st order on the sharpest frequencies of human hearing much nicer.
Might even try 1st order on the mid horns.

Tapped horns apart, as the craziness above 100Hz needs killing of fast!

I've often wondered what I'd plug in a Najda sized hole if it packed up?
With amps that can do X/O for mid bass and deep bass, plus some PEQ correction on deep stuff, it'd be little hassle.
I still have a modded Behringer 24/96 DCX that could do the tapped horn X/O and PEQ. Extra box though..