DIY 4520 bass horn build

[NickH]

Hey Nick, really nice build you got going on there. In your earlier picture with the two different horns in the room, how big is that room? How far back do you typically listen?

Hi Luther,

how's it going? Ya that an interim picture. I built one bass horn at a time. I finally got the second up there last night. The room is a pretty good size. Its a little over 30 feet deep. But its still pretty empty at the moment. I'm sitting about 8 feet back. They are not centered along the short wall

The bass definitely improved with the second 4520. I'm going to do some re arranging today though.

I did get a d140 in yesterday. And wouldn't you know it, the frame is just a tad too big. So I'll have to open the holes in the horns a little. Maybe only by 1/8" though.


I'm looking into the 150-4c and the e145 since that the driver that went in the c550 originally and the 145 was the predecessor of it. I just need to find the dimension for the extension ring for the front flange.
And then winding field coils for them next. Should be fun. But not easy, lol.


I'll post some pictures soon.

Nick

[NickH]

I'm guessing the closest I can get to the original drivers for these are some d130's. I'll use put d130 cone kits in the frames I have. Then I'll put a coating of pva diluted with water or damar varnish. Then glue a ring of copper wire under the dustcap to simulate a copper voice coil.

That will be the closest I can get to an original 130a. Since there no way in hell I'll ever own 2 pair of 150's, lol.

Let me know if this is a stupid idea or not.

Nick

[NickH]

I coated the drivers in one horn this weekend. I used mod podge which is a pva glue for puzzles. It has a matte finish so you can hardly tell I did anything. Im hoping it will just smooth things out a bit. I couldnt tell if I noticed a change. Only did one coat and it was very light at that.

I also did some measurements of the horns. Now these are very on the fly measurements and with a non calibrated mic. But it seems these horns dont rool off until 200 hz. I had the mic in the mouth of the horn. It also seem that the dont unload until about 25 to 28 hertz. No these are about 4: taller then the c550/4520's. Im wondering if the little increase in mouth area is whats accounting for this. But other then that the horns seem pretty flat for scoops. Theres also a bit of a dip at 80 hz.

Im not really putting much stock in these as I did it indoors. I did do a couple measuremtns in the room. It wasnt pretty. Which I knew it wouldnt be. But I also had a lot of background clutter/noise.

Nick

[Mostlydiy]

Hey how are things going?

All caught up in listening sessions I guess

/Mostly

[NickH]

Hey how are things going?

All caught up in listening sessions I guess

/Mostly

Lol. Working on it. I've been toying with things here and there. Thank to a fellow forum member I've got the last alnico frame I need for the horns. Still not entirely sure what cones I'll use though.

Of the list of options, e130s are at the top. I've also been trying to iron the bugs out of my tube active crossover. Still have some hum in it. Need to figure out if its from the supply or a ground loop. I'm thinking ground loop though. I clamped on a 680 mfd capacitor on the supply and there was no change in hum. So maybe I play with it but I doubt it. I spent my whole day in a pipe yard inspecting pipe. I just want to do nothing right now, lol.

I'll post the progress when I've got some.

[NickH]

Last night I finally got one of my newly designed passive crossover in place. I hooked it up with clip leads as a temporary setup. I wanted to test it out first before I made a more permanent setup. It was kinda spooky what happened when it started working, I put it on the right channel. The left channel still had the jbl crossover on it. The woofer has second order linkwitz alignment. The tweeter is just a 3mfd capacitor in series with the driver.

When I first hooked everything up it sounded like crap. The mids where gone and the horn was extremelyly honky. At first I thought it was just the pads where out of whack with the necrossoverer. I kept tweaking them but could'nt get it right and got irritateded at it. So I walked away from it for a bit. Started thinking that I had something hooked up wrong in the band pass. When I went back to it and I sure enough did. I was acting like a notch filter instead of a bandpass. Fixed that and started listening.

This is when it got kind of spookey. The right speaker disappeared in the image. The jbl croosover on the thing never imaged woth a crap before. Maybe a little bit but not much. But with the new crossover the image instatntly shifted to center. But its smeared from center to left. It like there was no sound coming from the right channel even though there was. Like I moved the speaker over by a couple of feet.


All and all I think these corossovers will work so I just need to build a more permanent setup now.


Nick

[Lee in Montreal]

Phase problem? 12db (180° dephasing ) meeting 6db (90° dephasing) in a non-time-aligned drivers system?

That's where digital crossovers come very handy

[NickH]

Phase problem? 12db (180° dephasing ) meeting 6db (90° dephasing) in a non-time-aligned drivers system?

That's where digital crossovers come very handy

Ah nope. The bandpass filter hooked up backwards. It was acting like a notch instead. So it was filtering out evenrything between 800hz and 7khz. Im planning on trying a active cross. I had one Ive been working on but its still has some hum in it.


Nick

[more10]

I am using miniDSP 4x10 HD on my 4530. I invert the bass horn and delay the rest by the horn length. Much better kick!

I cross the 4530 below the backloaded dip at 150. I am using the miniDSP parametric EQ to fix the rising response on my 2220A as well as CD compensate the 2370.

[NickH]

I am using miniDSP 4x10 HD on my 4530. I invert the bass horn and delay the rest by the horn length. Much better kick!

I cross the 4530 below the backloaded dip at 150. I am using the miniDSP parametric EQ to fix the rising response on my 2220A as well as CD compensate the 2370.

Ive often thought about going this route. But Im trying to keep things as simple as possible. It might come to this but Im going to stick with passive crossovers for the moment. Ive put one in and it seems to be doing ok. I thought about crossing the horns over where they roll off but that would mean going to a 4 way system and that just another layer of complexity that I dont want. I guess you could say this is a poor mans attempt at a vintage jbl theater system.


Nick

[more10]

As an experiment, reverse polarity of the bass horns and place your tops one horn length in front of them.

[Lee in Montreal]

Reversing the polarity of the woofers doesn't affect it's response by itself. Therefore, feeling "more kick" is most likely substance-induced The horn part of the cabinet is a filter. 120Hz and down goes thru the horn and the wave is reversed as it comes from the back of the woofer. 120Hz and up comes from the front of the woofer and must be in phase with the medium driver. The 120Hz dipin response is the phase cancellation between the woofer and the horn.

[more10]

The intention of reversing the polarity of the horns is to use the horn as the front, and then time align this front with the top systems. Oh, my mistake, you need to place the horns one horn length in front of the top systems to time align them.

[Lee in Montreal]

... you need to place the horns one horn length in front of the top systems to time align them.

On top portion of the spectrum, drivers shall be physically aligned and in phase. So, yes, the sounds starts at the same time and same location. But that's only the theory. Add passive crossovers and you add delays. 6db filter is 90° delay, 12db is 180° delay. I did the testing adding and removing caps on my 2360 and checking timing corrections with my active crossovers. So, on top of the theory, you may want to add some empirical experimentation. Test by ears and with a mike.

[more10]

Filters add phase shift, not delays. The shift is mostly in the stopband.

And of course, the time alignment I proposted only works if your top systems go down to the first dip in the response, 120 that is.