If this is your first visit, be sure to
check out the FAQ. You may have to register
before you can post: click the register link above to proceed. To start viewing messages,
select the forum that you want to visit from the selection below.
because of highish Qts, I've decided that this driver is best to be used in a dipole configuration. As most of the guys know, my preference in general are line array systems. And so, here's a perfect contender for a ultra cheap build.
The problem with a build like this is finding a good enough tweeter without breaking a bank. Seas TDFC, Dayton RS28A would probably be slightly better choices. I don't have these on the shelve and didn't want to buy another pair just for this project. But I did have a pair of Vifa buyout tweeter. Basically it's a DX25 that is used by ProAc in their 5k loudspeaker models. Not a perfect tweeter but at $26 it's really good AND I got it at like $19. ProAc version is very close to the DX25.
MLS FR with 4msc gating. Mic is about a meter and a half away, centered on the array. High frequency loss, due to the driver coupling is pretty evident here. As with anything else, no free lunch. This is why I was looking in to robust tweeter, capable of keeping up with an array in to 2000hz area.
Again, it's not unusual for a line array.
Now, the added "oopsy" is the fact that it's a dipole and it rolls of much earlier then a conventional speaker in the box.
Oh and cone resonance, even with high frequency loss is ear shuttering.
So, we got our work cut out.
I've attached the tweeter, did some MLS and modeling. Just a few first tries. You may say: "Aha! Roman, you are crazy! Nothing good will come out of it." But wait. Looks like a 3.2mh inductor should flatten the woofer response and still have enough output in to 2k area to blend with the tweeter flat enough. Alas, I only have 2.7mh. But, I figure I try and it's not the end of the world.
Tweeter has a 2mfd on it, polarity is not reversed. So, this on axis FR is very ragged and surely not something I can live with. Tweeter rolls off too early, the knee is incorrect. Woofer needs to be corrected as well. But surprisingly, the speaker didn't sound as horrible as it looks.
Then I decided to make the tweeter work harder and lower. 5mfd capacitor.
Mind you, this is not precise modeling or even controlled measurements. I am only throwing parts together to see which way to go and how far I can get waving the sword and riding a wight horse.
The individual roll offs of the drivers give more incite of the sound.
Vifa needs to work even lower and drop down at a higher slop. Second order filter looks like a good idea at this point.
Sony may or may not work with the second order but a cone resonance appear to be audible. Something is bothersome about the sound. It's nasal and chesty at the same time.
Second graph shows individual roll offs, FR of both an array and the tweeter and overplayed analog sweep.
At this point the response does not look so terrible and the speaker is actually starting to sound reasonably well. Of cause it needs more work. I wanted to look in to harmonic content of an array. The sweep is analog and contaminated but interestingly enough, cone break up is there and shining. This is one of the reasons, this array does not sound very pleasing. I'll have to design a notch for that resonance. I may also try to apply coating to the dust cap.
So from here, the plan is to get a couple of coils for the woofer array.
Notch out cone resonance and design a second order filter for the tweeter so it can work in to 2k and then drop down in to nothingness.
I'll be also trying a sidewall to increase apparent front baffle width and see how that affect the knee of the woofers output. May be smaller inductor will work better in this configuration. I really don't want to stress the tweeter that much.
The efficiency of the system should be in to 91db with power handling limits only gated by the tweeter capabilities.
There are some line array qualities that are starting to come through. Life like effortless sound. Large vertical image. I only have one assembled so far. More work ahead.
Also - anothre member seemed to tame the resonances with dampening the frames with rope caulk, others with peal n seal. You may try this and re-measure. Electrical notch may not be needed.
There's nothing wrong with reinforcing the frame and using some additional dampening. The major issue does not come from the frame but from the cone and that needs to be notched out. Anyone with some experience measuring drivers will recognize a narrow ridge propagating a few milliseconds
The update is this.
I've designed a notch filter. It's a parallel bottomless notch consisting from .35mh coil and 2mfd capacitor. The difference in the sound is significant to say the list. Anyone who already build systems based on these Sonys, should try notching out 6k resonance.
I've attached a picture of the different notch filters effects on frequency response. This is not the final version. Because notch can affect the tail of the woofers roll off, finalized version will be designed after on axis response is leveled out. I am now running a second order on both, tweeter and the woofer array.
Array is starting to sound really nice. Needs more base and I'll be playing with a U-pole idea aka adding the wings. Friend of mine came by and listened and then I told him what the drivers are his jaw dropped.
Very cool that you're running with this again. You inspired me to get off my **** and get mine going! Still waiting on some measurement gear and my cnc buddy, but hopefully I won't be too far behind you!