Monster Massive

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Basically we see that the beamwidth is narrowing at high frequency, just as expected.

I'm kinda stoked that the dipole radiation doesn't seem to be creating and weird lobes off axis, with the possible exception of a dip at 1khz.
Patrick,

+/- 17 dB from 500 to 20kHz.
Roughly omnidirectional below 2 kHz to -12 dB 45 degrees off axis at 10 kHz .
Huge off axis dips around 1000, 1600, 3000.
Around a 20 dB drop in the top octave.
What's not to like :rolleyes:?

You are padding the HF around 4 dB, a bypass capacitor of around .25 or .5 mfd could bring back around 4 dB at 20 kHz.

On my beamy Maltese horn, HF at output at 20K is equal to 10 kHz, even with 3" diaphragm drivers.
Are the mid holes in the skinny horn killing the HF, or does the driver just lack HF?

Art
 
Patrick,

+/- 17 dB from 500 to 20kHz.

As already noted, the midranges are more efficient than the compression driver.
This response has zero equalization.
The entire crossover is one cap and two resistors.
This is hardly what I would consider "finished."

Roughly omnidirectional below 2 kHz to -12 dB 45 degrees off axis at 10 kHz .
Huge off axis dips around 1000, 1600, 3000.
Around a 20 dB drop in the top octave.
What's not to like :rolleyes:?

You are padding the HF around 4 dB, a bypass capacitor of around .25 or .5 mfd could bring back around 4 dB at 20 kHz.

On my beamy Maltese horn, HF at output at 20K is equal to 10 kHz, even with 3" diaphragm drivers.
Are the mid holes in the skinny horn killing the HF, or does the driver just lack HF?

Art

In this thread from four years ago (http://www.diyaudio.com/forums/multi-way/151376-homster-how-i-learned-how-fix-horn.html) I noticed that stuffing a crummy horn with polyfill made a much bigger difference than stuffing a good one. The QSC waveguide with polyfill barely had any attenuation; the USD horn had a LOT of atttenuation. Something like 10dB iirc.

I'm seeing something similar with this LeCleach horn. When I fill the throat, I'm losing about five decibels of output at 20khz.

That would seem to indicate that the midrange holes add HOMs.
The horn *definitely* sounds more brittle in the upper midrange now that there's four holes in the throat. (Last week there was only one.)

But the polyfill seems to nuke it pretty good. I'd say that this horn with polyfill and four midrange holes sounds 'smoother' than last weeks horn with no polyfill at all, and one hole.

Another reason I'm losing HF is that the horn surface is really rough. As I continue to refine it I'll throw some bond on there to clean up the surface, but ATM, it's just unsanded fiberglass.

BTW, all of my subjective comments are based on listening to the horn equalized.
 
This response has zero equalization.

When I fill the throat, I'm losing about five decibels of output at 20khz.

The horn *definitely* sounds more brittle in the upper midrange now that there's four holes in the throat. (Last week there was only one.)

I'd say that this horn with polyfill and four midrange holes sounds 'smoother' than last weeks horn with no polyfill at all, and one hole.

BTW, all of my subjective comments are based on listening to the horn equalized.
Unfortunately, EQ won't fix the axial response problems, which would be more visible with the response equalized on axis.
Too bad stuffing the throat is needed to remove "brittleness" with this design, you have made it clear how much you like the top octave the polyfill eats up.

How many bands of EQ are you using?
 
Patrick,
Very nice work with the LeCleach horn! The sound of the horn on the video is actually very nice - can you post some jazz with sax and bass to see how that sounds? You are the master of prototyping with whatever it takes to get it to fit. I love the tube duct and PVC chamber approach you did for your car - although the time delay was big. In my sims I see that the "dip" near 1Khz can be reduced by placing the mid driver closer to the mouth - about halfway down the length of the main expansion. Can you please provide detailed dimensions of your LeCleach horn? I would like to try it in foam core with smooth bends.
Regards,
Xrk971

Here's some James Brown, Bassnectar, John Coltrane, and Michael Jackson:

Synergy LeCleach Nov 2013 - YouTube
 
Unfortunately, EQ won't fix the axial response problems, which would be more visible with the response equalized on axis.
Too bad stuffing the throat is needed to remove "brittleness" with this design, you have made it clear how much you like the top octave the polyfill eats up.

How many bands of EQ are you using?

Art you seem to be critical of using a horn because it's beamwidth is narrower than what's used in the commercial Synergy horns, and because the beamwidth of a LeCleach horn isn't constant.

IMG_20131115_164816.jpg

Luckily, I have a pile of horns and a pile of compression drivers, so evaluating your criticisms is as simple as walking to the garage.

On the left is my LeCleach Synergy horn; on the right is a QSC 60x90 waveguide. $35 at Parts Express. Compression driver on both is identical, a Celestion CDX1-1425.

From the front, the LeCleach looks about twice as big...

IMG_20131115_164838.jpg

But from the top, we can see that LeCleach is much, much deeper.

If anyone's curious why the QSC is painted white, it's because I'd tried to scan it with a 3D scanner. My original plan was to simply use the QSC, but make it bigger to lower the cutoff. Figured I'd 3D scan it, scale it, then print it.

cleach-11-8.jpg

Here's the LeCleach from last week, with one midrange, one hole that's 2cm in diameter, and a compression driver.

monster-massive8.jpg

Here's the LeCleach now, with four midrange holes. In this measurement, only the tweeter is playing. Looks like I have a wee bit of a reflection :O

monster-massive11.jpg

Here's the response of the same compression driver, but on the QSC waveguide. Gotta admit, that looks pretty good. The LeCleach horn gives me 8dB more output at 750hz, and 12dB more output at 500hz, due to the greater depth. (The LeCleach is 30cm deep, and the flare rate at the mouth is about 400hz.) But the QSC waveguide is smaller, it's a part you can buy 'off the shelf', and since it's made in a factory the finish is perfect. All of which help with quality control, consistency, and high frequency output.

monster-massive10.jpg

monster-massive13.jpg

Here's the impulse response of the LeCleach and the QSC. Here's some things I notice:
1) I flipped the second pic to reverse the polarity. (Didn't notice the compression driver was wired out of polarity, so flipped the image to compensate.)
lp_impulse.png

2) According to theory, LeCleach should have superior impulse response to conical or oblate spheroidal. This measurement seems to be consistent with that theory. The LeCleach isn't perfect, but it's closer to perfect. The image above shows what an 'ideal' impulse would look like. See how the LeCleach and the QSC look similar for the first millisecond or so, but then the QSC diverges from ideal behavior? I'm guessing this is because the LeCleach is designed to optimize the impulse response, and also because of the additional depth of the LeCleach. It would be interesting to compare a conical horn of the same depth to LeCleach, and you can do that in HornResp.
 
monster-massive12.jpg

monster-massive9.jpg

CSD_PL_000446_GPhorn.png


Here's some comparisons of cumulative spectral delay.
I think that's an important test to do.
Unfortunately, my measurements are particularly crappy when it comes to CSD, because I'm doing them in my living room. So it's hard to get a clean decay measurement, because the walls, floor, and ceiling are so close.

Anyways, here's some observations.
1) First measurement is the LeCleach
2) Second is the QSC
3) This is the same as the second, but the measurement is from Brandon's page. It used the same compression driver and waveguide, and gives us something to compare mine too.
4) To me, it looks like the QSC is a *little* better than the LeCleach. But both measurements are similar, it's hardly night and day. The QSC appears to decay a little faster, particular at high frequency.
5) The Celestion has an obvious resonance at 18khz, as shown by the ridge in the CSD
 
Founder of XSA-Labs
Joined 2012
Paid Member
Here's some James Brown, Bassnectar, John Coltrane, and Michael Jackson:

Synergy LeCleach Nov 2013 - YouTube

Thanks for posting that, I listened to it with my pc and it works fine. Sounds really good - you did a great job on this horn. Very natural, realistic sounding. In case you are interested, here is a sound clip from my latest horn that uses only one type of driver (4 of them) to cover from bass (60 Hz) to highs (~18 kHz). Naturally, the phase and time alignment are automatically satisfied, and because they are in a synergy-like horn, are a point source spatially. So here it is, a 1-way full range synergy. Let me know what you think of the sound. What is interesting is that I actually can feel the punch of a kick drum on my chest from four 3.5 inch drivers.
 

Attachments

  • Quad-Driver-Horn-2.zip
    797.1 KB · Views: 95
Unfortunately, EQ won't fix the axial response problems, which would be more visible with the response equalized on axis.
Too bad stuffing the throat is needed to remove "brittleness" with this design, you have made it clear how much you like the top octave the polyfill eats up.

How many bands of EQ are you using?

monster-massive16.jpg


I don't have a way to equalize the response on axis, because my soundcard driver doesn't have an EQ.
So I did the next best thing, which was to tweak the passive crossover so that we have flat response on axis.
And this is what I came up with:
1) There are three components
2) The L-Pad is 10 ohms in parallel and 2ohms in series. According to L pad calculator - attenuation dB damping impedance decibel loudspeaker speaker voltage divider - sengpielaudio Sengpiel Berlin ,that will give me 4dB of attenuation.
3) The cap is 1.5uf
4) The angles are 0, 22.5, and 45 degrees
5) The frequency response fits into a window of +/- 3dB from 0 to 22.5 degrees

monster-massive15.jpg

Here's the response of the LeCleach horn that I built, with the exact same crossover

I'd like to insert some witty commentary, but the QSC seriously walks all over my horn :( :( :(
 
I'd like to insert some witty commentary, but the QSC seriously walks all over my horn :( :( :(
True, seems to be little merit to the latest horn, but were your Celestion CDX1-1425 measurements done with the mid drivers shorted?
If they were not shorted, the monster massive dip around 1200 Hz may be due to the mid drivers creating a tuned "mid trap".

The mid trap could also cause some of the other problems too.

By the way, my estimate of 4 dB attenuation using the pad configuration you employed was from the same calculator as you used. In testing out actual attenuation with real drivers, found the suggested values seldom were within 3 dB of the measured results.

Testing with the mids shorted and no HF crossover will probably yield an entirely different looking response.
 
Patrick,
What do you think about using a commercial horn like the QSC and modding it with injection holes or slots connected to mid drivers? That might make a very nice DIY Synergy.

Yeah I think that's probably the wisest route.
The LeCleach has some decent looking impulse response, but the polars are dreadful. This is mostly due to my crummy horn construction, and the fact that the midranges holes are 'tapped in' at a section of the horn where the area isn't very high at all.

Right now what I'm trying to figure out what type of directivity I want.

I used open back midranges on this Synergy horn mostly because I wanted to minimize the 'giant headphones' effect.

But it's not as simple as running them dipole; I also have to consider that dipole radiation requires different crossover points and requires different EQ than monopole does.

So before making any sawdust, I need to figure that out.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.