jackgiff
Well-Known Member
Why did you raise the crossover point to 1.8 KHz?
Zilch's AK Design Collaborative - Econowave Speaker
- Thread starter Zilch
- Start date
Why did you raise the crossover point to 1.8 KHz?
Not by preference; I'd sure be happier if it were down around 1.5 kHz. I put up the EconoWave PCD Challenge over on the Tech Talk forum:
http://techtalk.parts-express.com/showthread.php?t=215536
It's a juggling act, much as Wayne describes: woofer with heavy resonance in the XO region, acoustic center offsets, wide C/C spacing, and necessity to precisely aim the forward axis due to consequential narrow spacing of the vertical nulls.
I've posted my FRD and ZMA files over there. The eWave XO will model in PCD with a few tricks. There's no constant directivity HF model, so use a small tweeter piston diameter, like 1", perhaps....
Edit: Now with a 3/4" roundover of the woofer opening:
Attachments
Last edited:
juniorJBL
SUB 1500 :)
What did that do for the Measurements before and after? Did it help the forward lobe?
Last edited:
I haven't rerun the vertical polar map, but it's unlikely to have altered that. You can see the on-axis effect in the plot above -- about 1/2 whit worth of difference, and mostly in the high frequencies. Until proven otherwise, I calling it an "aesthetic" variable.
[Maybe 1/8 whit at 400 - 500 Hz....
]Edit:
It did move the forward axis up a couple tenths of a degree, and raised the XO frequency there slightly:
Attachments
Last edited:
juniorJBL
SUB 1500 :)
I haven't rerun the vertical polar map, but it's unlikely to have altered that. You can see the on-axis effect in the plot above -- about 1/2 whit worth of difference, and mostly in the high frequencies. Until proven otherwise, I calling it an "aesthetic" variable.
[Maybe 1/8 whit at 400 - 500 Hz....
Aesthetic sounds good too!!:yes:
(pun intended)
Russellc
Addicted Member
Not by preference; I'd sure be happier if it were down around 1.5 kHz. I put up the EconoWave PCD Challenge over on the Tech Talk forum:
http://techtalk.parts-express.com/showthread.php?t=215536
It's a juggling act, much as Wayne describes: woofer with heavy resonance in the XO region, acoustic center offsets, wide C/C spacing, and necessity to precisely aim the forward axis due to consequential narrow spacing of the vertical nulls.
I've posted my FRD and ZMA files over there. The eWave XO will model in PCD with a few tricks. There's no constant directivity HF model, so use a small tweeter piston diameter, like 1", perhaps....
Edit: Now with a 3/4" roundover of the woofer opening:
Definately looks more finished, whether it sounds better or not!:yes:
Russellc
4343
Super Member
It does indeed. I'd be tempted to run a roundover bit (might need a smaller radius) all around the protruding front edges. Wonder if that would be a measurable diff?
Is it just me, or is there a whole whit there at 7K?
We're going to do more, like move the baffle incrementally forward, now that we have a stable baseline. I am anticipating some surprises; there is no dearth of blather about baffle edge diffraction, but precious little data.
Yup, at least one whit there, but elevated methane levels might account for it....
Last edited:
4343
Super Member
Cerwin Vega Waves
Downwind from Berkeley Farms, are you?
Here's a pair of little E-Waves I'm almost done with.
Still need to bolt the baffle in, the speaker carpet is all that holds it in right now. Of course, the original baffle with the same CV's and a piezo tweeter mounted between never did get bolted in... The 220Ti weighs a lot more than the piezo though, and I finally put pole sockets on the bottom, so It needs bolting now. Been playing them all day, sound is much better that it was! Next up is adding some fiberglass stuffing.
I am running the 2 4 ohm CV's in series, but in a 2.5 mode with the mids filtered out of the bottom "Helper" woofer.
...Yup, at least one whit there, but increased methane levels might account for it....
Downwind from Berkeley Farms, are you?
Here's a pair of little E-Waves I'm almost done with.
Still need to bolt the baffle in, the speaker carpet is all that holds it in right now. Of course, the original baffle with the same CV's and a piezo tweeter mounted between never did get bolted in... The 220Ti weighs a lot more than the piezo though, and I finally put pole sockets on the bottom, so It needs bolting now. Been playing them all day, sound is much better that it was! Next up is adding some fiberglass stuffing.
I am running the 2 4 ohm CV's in series, but in a 2.5 mode with the mids filtered out of the bottom "Helper" woofer.
Attachments
Last edited:
I pulled the speaker stands from my econowaves with 15" JBL 2226H in a 4507 clone cabinet. This brought the 2226H from 18" above bare floor to about 5" above floor. As expected, I have stronger bass now. The downside is it's not as tight and voices are a bit more muddied now.
Answers to Some Questions:
SO, what happens to the forward axis and crossover frequency as we adjust the L-pad up and down?
I optimized the crossover for flattest balance between woofer and waveguide on the forward axis with the L-pad set to its midpoint, Red.
Advancing it to the 3:00 position raised the HF output ~1.25 dB, Green.
I then found the equivalent setting, 11:00, lowering it the same amount, Blue.
At each position, I adjusted the mic height up or down to find the maximal inverse null and determine the crossover frequency and forward axis.
XO Frequency:
-1.25 dB = 1.848 kHz
0 dB = 1.817 kHz
+1.25 dB = 1.785 kHz
I.e., +/- 31.5 Hz, or 1.73%
What's happening? See §4.3 here:
http://sound.westhost.com/bi-amp2.htm
We are walking up and down the lowpass slope as we raise and lower the highpass amplitude.
Forward Axis:
Relative to the midpoint between the drivers,
-1.25 dB = +1.95°
0 dB = +2.28°
+1.25 dB = 2.60°
I.e., +/- 0.325° or ~20' (minutes)
I'm calling it "stable and predictable."
SO, what happens to the forward axis and crossover frequency as we adjust the L-pad up and down?
I optimized the crossover for flattest balance between woofer and waveguide on the forward axis with the L-pad set to its midpoint, Red.
Advancing it to the 3:00 position raised the HF output ~1.25 dB, Green.
I then found the equivalent setting, 11:00, lowering it the same amount, Blue.
At each position, I adjusted the mic height up or down to find the maximal inverse null and determine the crossover frequency and forward axis.
XO Frequency:
-1.25 dB = 1.848 kHz
0 dB = 1.817 kHz
+1.25 dB = 1.785 kHz
I.e., +/- 31.5 Hz, or 1.73%
What's happening? See §4.3 here:
http://sound.westhost.com/bi-amp2.htm
We are walking up and down the lowpass slope as we raise and lower the highpass amplitude.
Forward Axis:
Relative to the midpoint between the drivers,
-1.25 dB = +1.95°
0 dB = +2.28°
+1.25 dB = 2.60°
I.e., +/- 0.325° or ~20' (minutes)
I'm calling it "stable and predictable."
Attachments
Last edited:
skywave-rider
Slumming in Akihabara
Designing a system with fixed L-Pad to match driver sensitivities, getting the attenuation right on paper is the way to go if you really want to find the acoustic center through measurement. I'm just assuming this. So when using a variable L-Pad, we should calculate the value of attenuation and set the L-Pad to that value -- that is if splitting hairs is the method -- and let the measurements guide us to the acoustic center.
Your comments, please.
Edit:I should have realized that "on paper" sensitivities aren't the whole story, driver/waveguide combo must be measured acoustically for output.
Last edited:
skywave-rider
Slumming in Akihabara
4343
Super Member
Yep.:thmbsp: I used a bypassed electrolytic, 130uF as a trial, sounded fine, so it's still in there. Thought about biasing it, but probably won't 'cause I only have 4 of the caps, and need another pair for the 2 10" boxes that are waiting for better weather, and for waveguides.
Oh, and the XO is the 4-ohm version, since there is only the one woof doing mids, where the cap looks like a short circuit. Had to crank the lows at 80Hs up 12 db due to the woofs being 8 ohms in series at the low end, but they sound pretty good in a corner, outdoors they will need a sub, I'm hoping the pair of 2226's I just got will make my home-built TCB really rumble! Going to hook up the SB-5 next and see what that can do for them.
Next up is a set of D-130 floor monitor's that are going to have the 2402's replaced by the 6X6 WG... The hole in the mids was actually helpful with live mics close by, but otherwise they sound like "boom-TSSSH" (Note lower case 'Boom", they are 130's after all!)
200!
Last edited:
Yes, and unless you're modeling the L-pad as well, include one at the desired nominal setting when measuring the impedance and frequency response of the HF driver. In PCD, we're a resistor short of being able to put one in the model with eWave HF comp topology.
I did those measurements because we've been speculating about how much the L-pad is influencing the outcome, and I've previously documented measurable differences in the response when one is in place vs. not. I can certify that the location of the forward axis via inverse null is an extremely sensitive measurement; we're talking incremental changes in mic height of less than 1/4" here when finding depths on the order of -40 to -50 dB, i.e., the diameter of the mic itself. I can go in there this morning and find it is not precisely where I left it last night.
Bottom line of this is that while variables such as cabinet edge diffraction, woofer opening roundover, and L-pad adjustments do in detail make measurable differences, seen in the context of the inherent unit-to-unit variability of the compression drivers themselves, they are clearly higher order, and do not warrant getting all exercised about; there are elements of far greater significance in play here.
Yes, 1.8 kHz is higher than I'd like to be playing DC300, but I can tell you unequivocally in trading that off against getting the phase and forward axis properly aligned, in terms of the audible result, the latter is far more significant. Are there other driver and alignment combinations that'll get to a better place? Very likely, and therein lies the fun of it all -- discovering them.... :thmbsp:
Last edited:
Did you build from the plans I posted several years ago on LHF?
That's going to work, too, I bet, and constant directivity will make a major difference.
The other vintage combo I'd like to try which Gordon and I have discussed here is D123s in pairs, run in 0.5 mode. I have four of them in the dungeon somewhere that I haven't seen in 20 years, probably.... :tongue:
Last edited:
16 ohm econowave
Hello econowavers,
This project has got me thinking, and I'd like to give it a go. I have a couple of questions though, regarding what I'm thinking of trying. I'm looking at implementing this as a 16 ohm speaker, as I have (or will have, it's in bits currently, but has worked) a valve amp that only has a 16 ohm output. This amp also will only do about 4W/ch so I'm hoping to hit 99-100db/W/m (or higher) if possible.
As far as the 16 ohm thing goes, I have read through the series of posts that start here: Link (go to post 700, the resolution to this problem is over the next few pages). The main points are here: Link - posts 726 and 735. The L-Pad is changed to 16 ohms as well.
I believe this is for an 8 ohm woofer, 16 ohm selenium CD, and gives an overall speaker impedance of 16 ohms. Have I got this right?
Woofers:
I have some old 60's vintage Australian Rola Hi-Fi 12UX drivers that were reconed to have 8 ohm voice coils. One is blown (not by me), the other has a cone rub sometimes and will need work at some point, but is ok for testing. These are rated for 30 watts, an Fs of 30Hz, and have enormous alnico magnets, I reckon their eficiency to be around 100db - give or take. I'm not sure if they will be worth using for this project, I'll test with the one that works and see. If they turn out to not sound any good, can anyone think of a high-efficiency 12" woofer that would be appropriate? Inexpensive is good, but it doesn't have to be bargain basement. Bass is good too, power handling is not important (I've got 4 watts here).
Cheers
Hello econowavers,
This project has got me thinking, and I'd like to give it a go. I have a couple of questions though, regarding what I'm thinking of trying. I'm looking at implementing this as a 16 ohm speaker, as I have (or will have, it's in bits currently, but has worked) a valve amp that only has a 16 ohm output. This amp also will only do about 4W/ch so I'm hoping to hit 99-100db/W/m (or higher) if possible.
As far as the 16 ohm thing goes, I have read through the series of posts that start here: Link (go to post 700, the resolution to this problem is over the next few pages). The main points are here: Link - posts 726 and 735. The L-Pad is changed to 16 ohms as well.
I believe this is for an 8 ohm woofer, 16 ohm selenium CD, and gives an overall speaker impedance of 16 ohms. Have I got this right?
Woofers:
I have some old 60's vintage Australian Rola Hi-Fi 12UX drivers that were reconed to have 8 ohm voice coils. One is blown (not by me), the other has a cone rub sometimes and will need work at some point, but is ok for testing. These are rated for 30 watts, an Fs of 30Hz, and have enormous alnico magnets, I reckon their eficiency to be around 100db - give or take. I'm not sure if they will be worth using for this project, I'll test with the one that works and see. If they turn out to not sound any good, can anyone think of a high-efficiency 12" woofer that would be appropriate? Inexpensive is good, but it doesn't have to be bargain basement. Bass is good too, power handling is not important (I've got 4 watts here).
Cheers