15" hornloaded coax

Apr 4, 2017
32
5
8
#1
Maybe nice to share some stuff I found out recently.

With my new setup I'm currently using a sort of D&B C7 top with hybrid subs underneath (more like TSE sub, but a bit bigger).
Last weekend I had some time to do some testing, and a bit of measuring.
I have only one screenshot available at this moment (I was stupid enough to NOT save al the data).


There is clearly some trouble starting at 600 Hz with also three phase wraps.
My first thought was reflections coming back from the rather big XR1464C horn, and this was also backed up at another forum.
Also to be clear, measuring was done outside, top cabinet stacked on two subs, and measuring distance was about 2m.

So now I have about two options:
-Add some sort of foam behind the horn to minimise the reflections.
-Or use a different driver (BMS 4554 now) to cross lower. Current crossover is at 1200 Hz.

I'm going to start first with placing some foam behrind the driver, and see what the results are.
 
Apr 4, 2017
32
5
8
#3
I have to add that I'm pretty new to the measuring part, and probably make some rooky mistakes..


Used crossover is the Beyma FD212.
 
#4
Those look like comb filter reflections in your measurement space or set up. Your crossover is unlikely to be a part of that problem. A measurement of individual passbands is always instructive but if you've mounted that large a horn in front of a 15" driver that is operating up to the horn crossover then you are likely to have issues. Might be time to consider making it a three way box and move the mids to an 8". The classic Altec 604 uses a much smaller horn.
 
Apr 4, 2017
32
5
8
#5
Measurents where done outside in an open field, so I don't think that's the problem.
Moving to a 3-way design is not an option, at the moment I definately don't have the recources.
And wan't to try and (maybe solve) reduce the problem.
 
Feb 24, 2015
237
8
18
#6
I'd go back outside and measure each of the drivers individually.

I'm guessing either the 1464 horn in the throat of your 15" is giving you the ripples like you're thinking,
or you have a timing issue between the 15" and the CD.

If the 15" alone still measures with the ripples, well.....dammit...... huh?
But if it measures cleaner, you have to think it's driver interaction problems.....which I'm guessing is no more than a timing problem...

For digging into timing/xover, I'd compare REW's Estimated Impulse Response shift for each section. (assuming you're using loopback or acoustic ref)
If the Estimated IR shifts are not the same, they most likely need to be (with the use of delay)

To see what getting equal IR shifts does, you can measure the 15" again, and apply the same impulse response shift used for the BMS CD.
Then you'll have individual phase traces for both sections, and you can compare how they would overlay when time aligned.....

To dig further without having to add delay, you can then sum the two measurements (that have the same Impulse Response shift applied) under All SPL,
and see how that computer summation compares to the whole speaker measurements.........
The All SPL summation should be much cleaner than the measurement of the entire box (if timing is a prob)

I see the xover is only 2nd order so I guess maybe it's possible summation could be an issue down to 600Hz apart from timing???...
but those ripples look pretty deep...my bet is timing...

Hope this helps...good luck

Edit...I said to use the Est IR shift from the CD on both....I should have said use whichever section gives the longest Est IR shift on both.
(with the idea you'll end up delaying to longest...)
 
Last edited:
Apr 4, 2017
32
5
8
#8
Update!

The problem is definately the reflections from the horn. I've ordered the RCF ND-950 with HF950 horn for some testing.
It will probably be next week before i can get my hands on them.
 
Apr 4, 2017
32
5
8
#11
So the RCF arrived! Making measurements with the HF950 as with the XR1464 is the plan. Probably the HF950 woud be the winner in this case because of the better loading down to 600 Hz, but finding out in practice is always fun. After that i'm going to try and let it work together in the cabinet.
 
Apr 4, 2017
32
5
8
#12
So yesterday was testing day, but despite the rain I had to go inside. Yet again these measurements took place in a (to) small room, so the results aren't reliable. Also a side note, I'm totally a rookie when it comes to measuring


-Used amplifier whas Crown Itech-8000
-ND950 1.4 on the 18sound XR1464

With one PEQ applied around 10k:

I used the IR window function to get rid of the worst reflections.

Measured response without the PEQ looks almost the same as the specsheet
Hopefully the weather will be a bit better next week
 
Apr 4, 2017
32
5
8
#14
I used +2dB Q=1 @10kHz at the previous pic. Next challenge is to build a bracket for the horn/driver. The rcf weighs a bit more then the 4554, so I'm afraid the wood work at the front get's to much stress at the front. After that, I'm going to get the 15NW76 and the ND950 working together.