I think we are all in agreement with this? yes?
I have used impedance data from DATSv3 to design passive crossovers, and the resulting systems have matched simulation very closely, so I can infer that this little device, though simple, is sufficient. I also infer that the REW + resistor method does essentially the same thing... Yes?
j.
Not quite sure what you're looking for. What property/effect/process do you want to express by the value you are looking for?
I am open to ideas... Is there a way to quantify the severity of the resonance based on the shape of the impedance curve? In the example I showed, I estimate the impedance curve rises 0.1 Ohm relative to the nominal value of 4.1 Ohm at 640 Hz. So, 0.1/4.1 = 2.4% ... Is that enough to calculate some value of energy, or some relative assessment value?
What resonances are we talking about?
Cabinet resonances are easy, they totally comply to f = c / (2*s) = 344/(2*s)
s here is the distance between walls in meters.
c = speed of sound 344 m/s
f = resonance (standing wave) in Hz
This formula is without a so called end-correction.
In very simply words, this correction is needed to match the theoretical result with practice because how non-linear acoustics works (not entirely correct, but it will do).
But it will give you perfectly fine ballpark figures.
These will also line up with burst-decay (= waterfall in periods) plots with a nearfield measurements without any damping material inside the cabinet.
For a ported systems, the same formula can be used (although the max and min values are at different positions) to also estimate the port resonance.
There are also resonances from the surround, typically around 800Hz-2500Hz,
This always results in a dip in the response with a corresponding blip in the impedance.
With bigger woofers (say roughly 10 inch and above) you can sometimes also see a resonance in the suspension/spider.
Which is typically lower in frequency, around 200-400Hz or so.
By far the best way to determine what resonance belongs to what, is to just simply measure the woofers impedance in free air first.
Or look at 3rd party measurements.
If you want to do it well, measure the nearfield response as well + distortion.
Compare this to the measurements inside the cabinet. (again incl distortion etc etc)
Any issues that didn't show up in the free air measurements are not part of the
Last edited:
For example,
I should have been more clear. This impedance burble corresponded to a small cabinet structural resonance on the rear wall... NOT an acoustical "standing wave" resonance.
If there is a way to discount an impedance discontinuity (what I call a burble), based on the relative size of the discontinuity, without going to all the trouble of investigating the cabinet, I would be very interested in discussing it.
Last edited:
In general, my question would be different.
Before measuring, think about (or even just write down), what resonances we could expect and at what frequencies.
With logic deduction we can often already figure out what is left after all of that.
Even more so when we have information about free air measurements as well.
I personally find it always hard to answer such questions from people exactly because of this.
I don't know the exact system, dimensions and drivers.
So to some extend it can be a very big and open question if that makes any sense?
Especially when people are not 100% familiar with the specific drivers.
The context for the impedance plot I showed above is here in this link. The driver is a Purifi PTT6.5M04.
https://www.diyaudio.com/community/threads/purifi-waveguide-project.394174/post-7349489
My main interest is looking at the impedance plot from an electrical signal processing standpoint, or an electrical power standpoint, and extracting information about the 640 Hz resonance based on its shape. Can I get the Q of the resonance? (maybe)... Can I estimate how much acoustical radiation will come from the cabinet wall? (probably not)... But I am interested in anyone's thoughts.
https://www.diyaudio.com/community/threads/purifi-waveguide-project.394174/post-7349489
My main interest is looking at the impedance plot from an electrical signal processing standpoint, or an electrical power standpoint, and extracting information about the 640 Hz resonance based on its shape. Can I get the Q of the resonance? (maybe)... Can I estimate how much acoustical radiation will come from the cabinet wall? (probably not)... But I am interested in anyone's thoughts.
So you agree.
The problem comes whe you measure the wrong thing, or, you don't know what effect is causing the issue that needs measuring or even worse, think it's one kind of distortion when in fact it's another. This can really send you down the garden path.
So if we follow the step-by-step-guide
1 - measure loudspeaker in free air, incl impedance, distortion and burst decay (waterfall in periods)
2 - predict what resonances are to be expected inside the cabinet
3 - measure the same things with the loudspeaker inside an empty cabinet
4 - measure again with damping material
If the little resonance already showed up at (1), don't look further it's a driver problem.
If the resonance seems to be in ballpark at (2), use (3) to double confirm, especially waterfall/burst decay is leading here.
Even more so when this resonance gets a lot less significant at (4)
Additional there can be a step 2a with ported system as well as a step 5.
edit: Sorry, I was misreading the without any measurements part.
Well, in that case, not really, unless you know the speaker very well or the cabinet dimensions.
Dimensions can often be estimated, so you do have a global idea.
Last edited:
Note the word "pitfall" in my previous post.
There are lots of tangents (garden paths) to venture down. Nobody said there wasn't.
What is the point of your stupid post??????
Dave.
Me too!
Yes, something that doesn’t involve taking a Nearfield measurement of all panels.
Or getting a stethoscope.
This escalated when you misunderstood the intention in #164... We all do that from time to time, don't we?
However the measurement issue behind it is one worth going into detail for wouldn't you agree?
..and don't call me Shirley.
As far as I experienced it's difficult to know whether a resonance creating the impedance anomaly is kept inside the enclosure or will radiate outside.
With your example: a resonating enclosure back is much less problematic than a resonating baffle or a "standing wave" exiting through the cone.
I didn't misunderstand anything.
And you made an accusation, so I responded to it.
Dave.
I think the number one measurement is wallet mass. I’ve found it makes a huge difference in the quality of my projects. There is a definite fall off in audible improvement if the wallet gets too light. Heaven forbid the wallet mass goes negative! This may cause a sudden reversal in the cash flow continuum and whole stereo components may disappear.
Good questions. What I usually do is (1) identify the source / process that causes the burble and (2) try to do something against it (if possible) and use my ears to assess the improvement. This usually gives me a feeling for the importance of the burble and if / how much it needs to be taken care of. For example, if the burble is due to a standing wave in the cabinet, I add damping material (different amounts, at different locations in the box). Then I take impedance measurements to see how the burble changes, and I listen to the speaker to hear how the sound changes. There's usually a sweet spot (or a compromise) between reducing the burble and overdamping the cabinet.
@DcibeL
very kindly pointed out to me that one can view both charts in one, using an impedance plot with a logarithmic Y scale- this allows you to see peak, but also all the little wobbles.
@hifijim
since we’re having a grand time discussing/debating impedance data, I will soon post a few (?10) impedance plots, we can all play a “spot diagnosis” game.
No frequency response, no distortion measurements included. (?needed)
I then call upon my upon my DIYA brains trust… and we see if we can collectively solve them.
Last edited:
I scanned your posts about the midrange cabinet, and find it difficult to understand why it is the rear panel given its thickness. Also the frequency is a bit low i guess being very thick and braced.
Also looking at the shape of the burble i cannot see a panel resonance, more a mechanical connectivity issue.
I will track down some of mine measurements to show simple resonances and their shape.
What you can do is use the speaker as input and tap somehow the cabinet.
When listening with the stethoscoop was there a particular spot where it was the most noisy?