This looks awesome and "correct" 😎 some questions:...
- is that 10" or 8" woofer?
- How did you bend the metal poles ? 🙂
As for finding out pipe resonance, normally I just lift the speakers a bit and poke the microphone. I'ts very revealing.
Ha! you never told me that. But this is exactly one 'issue' I could not fix with my Pluto clone!
I measured the pipe resonance via impedance and FR and could see them 210hz...420hz... etc. from memory. Tried this and that with damping but in the end they were still 'audible'. I even tried notching *every* peak at one stage 😀
Here is my effort at omni speakers.
I enjoy them very much but wish I had some measurement equipment to show me changes in x'overs etc. Perhaps Santa will be nice to me this year..... 😀
As it is I use a single cap of 6.8 ufd to crossover 1st order around 7 kHz - and that's it.
There is more info over at this old thread - but it is a very long thread: http://www.diyaudio.com/forums/multi-way/100390-ohm-acoustics-walsh-f-speaker-remakes-6.html
I enjoy them very much but wish I had some measurement equipment to show me changes in x'overs etc. Perhaps Santa will be nice to me this year..... 😀
As it is I use a single cap of 6.8 ufd to crossover 1st order around 7 kHz - and that's it.
There is more info over at this old thread - but it is a very long thread: http://www.diyaudio.com/forums/multi-way/100390-ohm-acoustics-walsh-f-speaker-remakes-6.html
Attachments
Last edited:
It's a 10". Same Peerless unit as the Pluto sub. Poles were bent in a special way 😎. Meaning I had a shop do it.
Poke the mic? What's that?
The fullranger was a Ciare HX201 with massive treble rise, so that direct sound plus treble reflection gave a fairly good frequency response.
I hate dipole airiness, so I prefered it in a closed or almost closed setup.
Only for bass. Not in a range where wavelength compared to the size of the driver basket becomes significant.
Even Pluto has it. It is just a matter of audibility. Maybe here the alu membrane has a slight advantage over less rigid materials. But at this stage the statement is speculative.
The graph is highly misleading because just a single measurement was used to generate the sonogram. You can use any single measurement to generate such a graph. It even doesn't need to be flat if normalization is used.
For such a concept it would be more interesting to see a vertical sonogram. Unsmoothed, free field, mic at about 2m distance and not at 1m.
Last edited:
errr... I think you underestimated the design prowess of Herr 2Pi 😀
He was the one who taught me sonograms. Many sonograms was produced thereafter. 😎
Yes, because
a) the infinite number of horizontal points is replaced with just a single data point (measurements should prove or disprove a hypothesis, here it's just an illustration of the hypothesis)
b) normalization masks any resonance
c) the mic is too close, i.e. not in the far field
c) With such a small driver the mic is in the far field at 1m distance
a) I verified that the response did not change horizontally. So it does not matter if I use 1, 2, 3 or much more measurements. The time that was saved to take snap shots every 5° or 10° or whatever degrees is spent well elsewhere.
b) agreed on normalization. That's why I also show other diagrams and mention that it is normalized.
Compare the vertical measurements. There I went the full 9 yards because the response does change.
a) I verified that the response did not change horizontally. So it does not matter if I use 1, 2, 3 or much more measurements. The time that was saved to take snap shots every 5° or 10° or whatever degrees is spent well elsewhere.
b) agreed on normalization. That's why I also show other diagrams and mention that it is normalized.
Compare the vertical measurements. There I went the full 9 yards because the response does change.
Yes? There's a diffraction device involved. Did you confirm with measurements?
How did you verify if not by measurements??
Could you also posts a sonogram with vertical data?
Yes, if I remember correctly I went up to 1.7m.
I did measure it ! I probed one side with one of the little poles. Then I compared the graphs by substracting them from each other. The result was a flat line. Only the calculation/the picture is based on one measurement because it would have made no difference.
I have no reason to cheat here. I am not afraid of the work. One intention of the horizontal sonogram was and is to show that nice pictures cannot describe how a speaker really sounds regardless of how perfect they are.
If you can read that better than the polar plots I can do that (next week).
Thanks for the additional information.
Well, one can read that graph one way or the other. I obviously chose the wrong one 🙂
Would be nice.
Well, one can read that graph one way or the other. I obviously chose the wrong one 🙂
Would be nice.
There you go. Again, 0° is on listening axis. + is up and - down.
This is normalized. The exact same data that is published on my site as polar plots.
Attachments
- Status
- Not open for further replies.