The very low budget DIY synergy horn build - Page 3

I have finished the rear enclosure for the horn and am now working on measurements / crossover design. Building the enclosure took much longer then the horn itself, many more pieces to cut and glue.


Here I have some of the brackets to hold the enlcosure on the horn. It took me quite a while to figure out this step, how would I cut the brackets to fit on the horn? If I where to cut the brackets on an angle that matches the horn the corners would be a mess with angle chages or seperate small squares on the corners like I did on the v1 horns. Then I figured it out I don't need to cut them at all I need to cut the sides of the horn mouth square. Here you can see the sides and the first two brackets for top and bottom mounted.



Here I am gluing the side brackets on.



Starting the frame.





Dry fit of side panel to check driver clearance and determine location of removable access panel.





Gluing the sides in place.



Some bracing.



All done with access panels removed, the top and bottom are also removable.




Drivers installed.



Now I start my work on the crossover.

The mids on this one have been sealed you can see the strips of black cardboard on them. So no rear wave interaction between woofers and mids.

Here are new measurements of the woofer/mid/tweeter in the finished horn. I had ran out of gasket material so 3 of the 4 mids and all the removable panels do not have any gaskets so the enclosure is very leaky.




The woofers are not as sensitive as I thought they would be but the mids are more sensitive then predicted. I think there is somthing going on with the woofer ports and the relationship between the diameter and the depth. I think when the diameter is much greater then the length of the port it acts like the port length moves towards zero because there is no real air slug in the port. You can see this in my woofer plot as all the high frequencies escape, it does not act like a bandpass filtering them out


I have yet to do any crossover work yet, will post plans once I start on it.

I may wire the mids in series to bring them down some without using resistors but I need to see what that will do to my crossover, if I need to throw a ton of copper on them then that idea is out.

bwalso - I'll be in Italy during the MWAF so thats a no go

Paul -
I think some of the dip at 120hz may be room related causing the woofers to look like they have a larger peak at Fsc then they should. The woofers are mid Qts ~4.5 but have a high Vas around 5.8 cu ft which causes the high Qtc of the enclosure. I have already played around with the idea of porting the enclosure for a 6th order bandpass (can do this in winISD because drivers act as direct radiators at this frequency) I get a peak if I tune it anywhere above 30hz the only real benefit is the reduced cone excursion from 22-60hz when played fullrange.

I'll try and play around with it, I can only add 3/4" more material in front of the driver before it hits the top or bottom panel, so I may try that. I can also tape some 2" pvc on the outside of the ports to see what effect it has when I increase the port length.

When I saw the woofer graph I was , what a mess.

Ok I have finally got the woofer response I wanted, took alot of effort but I have it.

First moving the woofer back with a spacer ring would not give me the needed volume in the front chamber to get the HF filtering that is needed. What was going on in the woofer response was that I was not getting as deep a 1/2 wavelength cancelation notch at 500hz (woofer port to throat and back) as was in my hornresp sim. This led the peak at 600hz to get through which is about the tuning frequency of the woofer ports.

Here is a graph showin the hornresp sim, winISD sim, and actual measurements of the first iteration.



I had to find a way to increase the front chamber volume enough to get the bandpass filtering to start lower. So here is attempt #2:







You can see I split the horn internals into two seperate parts, one in the front of the horn and one in the rear where the ports are. I figured this would give me the bandpass filtering that I wanted. The woofer fronts are a little cramped under the top and bottom removable panels but I thought there should be enough breathing room for them. I also made a slight modification to the ports.

I don't have the graphs for this one because it turned out just as bad as the first design for filtering HF. It also turns out I had a huge notch at 200hz that I did not know where it was coming from. Gold star for someone who can figure it out from the photos.

After I saw that all my work made the response worse I closed the REW session of my second attept without saving out of rage .


I'll post my finished result after someone guesses the correct reason for why I had the 200hz notch in my woofer response in the second attempt (clue it was not a measurement problem).

I used SPLtrace to get the winISD response into REW.

Things did improve greatly in the third iteration.

A correct guess to the 200hz notch problem in the second iteration and I'll put up my new woofer graphs for the third iteration.

I'll do some outdoor measurements at some point, I don't have a PA pole but I think I can get the speaker a good distance off the ground for those.

ding ding, bwalso get the correct answer, they were about 1/2 wavelength apart at 200hz inside the back of the enclosure causing the cancelation. Gold star for bwalso




This did cause some issues at higher freqencies but not the 200hz notch.



Ok here is my third iteration, woofers moved closer together so not to cause any canceling issues within the passband and the horn enclosure is much better braced then the other two also.


Original woofer placement/design vs new


woofer/mids/high



The dip at 120-130hz is a floor bounce, It goes away when I measure on the ground. Good response to 80hz, about 5dB more output the the original woofer design from 80-180hz. Reduced cone excursion over the original design, able to take close to 200 watts input before reaching driver Xmax.

Now that I solved the woofer problems I can finally get to crossover design.

LFM2, my carpentry skills are not as good as some other members on this forum but thank you.

I don't think I will port the other enclosure, it will make the response peaky and would be very hard to do in the current form.

Here are the requested pictures inside the enclosure with foam removed. I have the enclosure split into 3 seperate chambers, each woofer has its own chamber that is ported into the horn and they share the other sealed one.



Here is the shared chamber, the woofers are mounted very close to the mids.





Thanks John,
I looked back at all my tweeter sweeps just to make sure that was not a driver thing because that selenium's FR gets kind of crazy and I do see it on the sweeps with the domes that I tried. Now its gets very interesting, I accidentaly clicked on a mid response while in REW when selecting all the tweeter ones and the slight mid peak I get is at the exact same spot, now I'd like to see someone explain that if in fact those are related and not just coincidence.



Also some crossover work has been done, not the flattest response you'll ever see, but done with parts on hand so I'm not complaining until I hear it.


Mounted on boards not yet soldered, I could not mount in all onto one board and fit in in the enclosure so there is a woofer/tweeter board and a mid board.

Here is the combined mid tweeter in/out of phase and individual responses.



Off to do the sweep with taped up mid ports

OK here is the tweeter response with and without mid ports covered.




Some interesting results the dips at 300 and 850 are gone but little change at 3.6k. There is improvement at 2.5k and 4.5k and overall it is much smoother.

Got outside and did some measurements today, 0-60 degrees spaced apart a little bit:



The response looks best at about 15 degrees off axis.



My thought is that on axis all the ports from the mids and woofers are at an equal distance and any reflections that occur and cause iregularities in the response will sum. When you move off axis those distances become different and the relfections no longer combine to give large peaks or dips.

I still have some work with the woofer and woofer/mid x-over, going to stuff the sealed woofer chamber becuase right now it is almost bare, only two peices of foam in there. I hope that will help flatten the peak at 85hz.