I don’t think the impedance plot reveals cabinet wall resonances very easily. More so the modal behavior of the enclosed air volume and its impact on the loudspeaker cone.
Impedane curve in free air vs. in box are needed to distinguish between spider/frame resonsnce form box/driver opening resonances and standing waves
The magnitude of imp. peak in Jim's example look insignificant to me, but I'm not an expert.
Low freq large impedance peaks tell about motor/driver or box tuning resonances.
---
Hörnli shows interesting variation! I haven't seen such before. Is it really possible that room modes make impedance to vary so much?
The magnitude of imp. peak in Jim's example look insignificant to me, but I'm not an expert.
Low freq large impedance peaks tell about motor/driver or box tuning resonances.
---
Hörnli shows interesting variation! I haven't seen such before. Is it really possible that room modes make impedance to vary so much?
Cabinet wall resonances are way to small to be detected that way.
Or with any of these other methods tbh.
Maybe with a near-field mic measurement, but the sound leakage coming directly from the drivers is enormous.
The purpose of the game kinda beats me?
To me it sounds like guessing the issues afterwards without the correct measurements?
Maybe it's possible if you know the drivers well, but doing it the right way is just much easier and finding the problems better.
But it does, i am searching for results where the rear cabinet panel was not stiff/stiffly mounted.
And if mem serves me it showed a similar pattern.
Why would this be a measurement issue?The differences could be explained as differences in the room loading or rather unloading the cone at very specific frequencies likely corresponding with room modes at those particular frequencies.
The different wiggles near 100 Hz are not surprising, given different position in the room, and hence different room modes coupling back into the driver. What is unusual is that impedance is rather high (25...30 Ohm) and almost perfectly flat (within +/- 1.5 Ohm up to 2 kHz). Why is that?
Nah, I don't think so.
Based on the setup it looks like he's using a series resistor of probably 60 ohms or so. Unless there are pads on both inputs of his sound-card, that speaker is not making very much sound during this test.
Not really enough to excite rooms modes......and certainly not enough to develop and act on the driver to contaminate the measurement.
The traces are wobbling around quite a bit more than expected in the low range. I suppose something environmental could be affecting it. Maybe the speaker was sitting on a window sill and there was a locomotive going by?? 🙂
Also, the phase plot is not tracking the magnitude correctly.
If seeing something like that, the first thing I'd do is substitute a resistor for the DUT and verify I had a perfectly flat magnitude trace and a flat phase plot on zero degrees.
Dave.
You’re right about the phase plot. The impedance minimum is quite high too and the general shape plus the irregularities do point at measurement errors. But still: room modes are level independent and can cause these tiny peaks, I think.
@markbakk
Yeah, it's not out of the realm of possibility.
But remember this is not an acoustic measurement.....except for the speaker itself working as a microphone and coupling back into the test scheme.
Relatively speaking though, that's not a very good microphone. And, it's quite busy producing sound during this entire test. 🙂
There is also the variable of stuffing inside the box that might have moved (or be moving around during) between tests.
I think I better get dressed in my nit-picking armor now waiting for some of the other members to give me grief. 🙂
Dave.
Yeah, it's not out of the realm of possibility.
But remember this is not an acoustic measurement.....except for the speaker itself working as a microphone and coupling back into the test scheme.
Relatively speaking though, that's not a very good microphone. And, it's quite busy producing sound during this entire test. 🙂
There is also the variable of stuffing inside the box that might have moved (or be moving around during) between tests.
I think I better get dressed in my nit-picking armor now waiting for some of the other members to give me grief. 🙂
Dave.
I don't see what would be wrong with the phase plot. Impedance is almost flat, and so it makes sense that phase is almost flat, too.
I don't see what would be confusing about the plot scale. It simply shows the impedance data ranging between 25...30 Ohm.
I’d expect a phase null at the impedance peak (corrected for inductance of course) but here it reaches zero way above the frequency of the peak.
A stepped sine sweep filters everything except the measurement frequency and the harmonics you look at. A MLS signal needs some FFT transformation and depending on size of the MLS you get more or less information at low frequencies. It reacts different to noise!
With the Audio Precision System a continues sweep reacts differently (more sensitive) to low frequency noise as a stepped sweep. This can make problems when you need calibration and use 1/4" mics with pretty high self noise at low frequencies for e.g. and will also influence your THD measurements.
Sorry to sidetrack the impedance topic, but I just watched this how-to video for time alignment using a usb mic and thought it was interesting. I did not know you could do this. Could be useful for active systems.
Video is pretty long but worth it IMO.
Video is pretty long but worth it IMO.
@Samps
I mentioned it in this post
but it didn't gain much traction. Thanks for bringing it up again.
Take 2:
For people who like to read, rather than watch a 1 hour video:
https://youtubetranscript.com/?v=El-kwZ5_nnU
For those who are time poor, I asked ChatGPT to summarize the aforementioned transcript to tell us what it's about:
This is what the LLM had to say-
Of course the LLM could be off by a little (or a lot) and not know that it doesn't know.
Have you have a chance to review the video? Can we all dump our dual channel measurement setups ofneBay now before everyone knows about this trick? If not, why not?
I mentioned it in this post
but it didn't gain much traction. Thanks for bringing it up again.
Take 2:
For people who like to read, rather than watch a 1 hour video:
https://youtubetranscript.com/?v=El-kwZ5_nnU
For those who are time poor, I asked ChatGPT to summarize the aforementioned transcript to tell us what it's about:
This is what the LLM had to say-
Of course the LLM could be off by a little (or a lot) and not know that it doesn't know.
Have you have a chance to review the video? Can we all dump our dual channel measurement setups ofneBay now before everyone knows about this trick? If not, why not?
Last edited: