Impedance chart, need explanation

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I don't know how to interpret an impedance chart for a subwoofer in a bass-reflex enclosure. Does it give any information ?

I know that for free-air, the peak in impedance indicates the resonance frequency of the speaker.

Before measuring the impedance, I made a nearfield frequency measurement and I could accurately tell the resonance frequency of the box (Fb), which is 52 Hz.

I have attached the impedance curve. Why does it have 2 peaks ? 52 Hz corresponds to the lowest point between the 2 peaks. Is this correct ?
 

Attachments

  • impedance chart.jpg
    impedance chart.jpg
    271.4 KB · Views: 745
Hi,

The minimum impedance between the
peaks indicates the port tuning frequency.

Vented has two peaks, sealed just one.
Related to 2nd order and 4th order.

Interpretation of the two peaks is not so simple.
Generally overdamped the lower peak is lower
and underdamped the lower peak is higher.

Without running off to check I think classically
maximally flat they should be equal, but I don't
like maximally flat at all, I like overdamped
that maximizes F-6dB to F-10dB.
Maximally flat optimises F-3dB.

I like room gain friendly alignments.

rgds, sreten.

Note that 52Hz tuning is speaker, not subwoofer territory.
All vented subs IMO need tuning < 41Hz. Low E on a bass.
 
Last edited:
Electrically, the vented box speaker looks like a series LCR filter in parallel with a parallel LCR filter. The parallel LCR is the speaker, and the series is the box/port. A parallel notch filter (speaker) causes an impedance peak. The series notch of the box/port then splits this single peak into two peaks.

If you can measure that same box sealed and vented with no damping material, you can determine many things, mostly from the maxima and minima. The best simple reference for this is D'Appolito's Testing Loudspeakers.

The heights of the peaks are strongly dependent on damping material and flow resistance in the port. The lower peak is affected mainly by the the port - more port flow resistance (or more velocity) means a lower peak, for this reason this peak will also change with drive level. The upper peak is affected mainly by box damping, more damping means a lower peak. Leakage losses from driver leaks or poor construction will reduce both peaks and raise the minima between. The minima is the port tuning frequency in the absence of driver inductance. The minima can be quite broad, and its exact frequency (and that of the zero phase point) can be shifted away from the actual port tuning frequency by driver inductance. For these two reasons, close miking the driver and finding the minimum is the most accurate way to determine tuning frequency.

Proper box volume and tuning frequencies are very close to (Small-Margolis):
Vb=Vas*20*Qts^3.3
Fb=Fs*0.42*Qts^-0.96
or (Keele):
Vb=Vas*15*Qts^2.87
Fb=Fs*0.42*Qts^-0.9

You can lower tuning from this and get a drooping response, or raise it and get a peak.
 
Last edited:
Now that you mention it, the tuning frequency seems a bit high.
It's an 8" car woofer (with 6.5" effective diameter) in a box made
by the manufacturer. I don't know why they chose the Fb so high.

Hi,

Try lining the port with porous foam about 3/8" thick.
Last time I did it I used a cheap washing up sponge.

Cut the sponge into a rectangle port length x 3xdiameter,
Fold it into a tube and insert into the port. This will tidy
up any tendency to boom and go deeper, tighter. The
foam also helps with chuffing by smoothing airflow.

rgds, sreten.
 
Member
Joined 2008
Paid Member
At low frequencies, the impedance curve precisely reflects the motion of your cone. That's what it tells you. That's wonderful.

In the case of using a resonant box to counter-act the natural motion of the driver, it (viewed in isolation) tells you not much about sound output in the range around the bumps. Odd to try to make accurate sound when the cone motion is poorly related to the sound output, eh.

Ben
 
At low frequencies, the impedance curve precisely reflects the motion of your cone. That's what it tells you. That's wonderful.

What on earth are you talking about now? The impedance curve doesn't look even vaguely similar to the displacement curve. Here's the impedance and displacement of a random ported box.

An externally hosted image should be here but it was not working when we last tested it.


In the case of using a resonant box to counter-act the natural motion of the driver, it (viewed in isolation) tells you not much about sound output in the range around the bumps. Odd to try to make accurate sound when the cone motion is poorly related to the sound output, eh.

Ben

A resonant box does not counteract the natural motion of the driver. The resonance provides a little bit of boost a frequencies where it's needed.

The sounds resonant boxes make are plenty accurate when properly designed, when tuned below 40 hz transient response and group delay are largely inconsequential. What you hear is frequency response and if frequency response if fine it will sound fine.

Even with sealed boxes the displacement curve doesn't reflect the impedance curve OR the cone motion so basically everything you posted is wrong and based on a misconception of how sealed and ported boxes work.
 
Last edited:
What on earth are you talking about now? The impedance curve doesn't look even vaguely similar to the displacement curve.
Even with sealed boxes the displacement curve doesn't reflect the impedance curve OR the cone motion so basically everything you posted is wrong and based on a misconception of how sealed and ported boxes work.

Impedance reflects velocity. You can look at it as velocity over force. The points where there is high impedance you get a lot of velocity for little force.

bentoronto's post was vague and misleading, yours seems like a straw man argument.
 
Impedance reflects velocity. You can look at it as velocity over force. The points where there is high impedance you get a lot of velocity for little force.

,,, yours seems like a straw man argument.

That may be the case but let's look very carefully at Ben's exact wording.

At low frequencies, the impedance curve precisely reflects the motion of your cone.

Notice that Ben very specifically said cone motion, not velocity. And here's the problem with that.

Can you draw me the cone motion (displacement graph) for this impedance curve.

2jdi8m0.png


It could be an enclosure with a sealed chamber (like a front loaded horn) or it could be an enclosure without a sealed chamber (like a tapped horn, tl, etc).

What about this one?

29fyuk8.png


If it's a flh the cone motion will look like the light trace and if it's a tapped horn the cone motion will look like the dark trace. (This displacement example is a random flh from my files that I ran as a flh, then I deleted the rear chamber and ran the sim as a TH, BOTH WITH THE SAME VOLTAGE LEVEL.)

2hyxlee.png


So yeah, if the example impedance is a flh it was have MASSIVELY different cone motion than a tapped horn, and it's pretty hard to tell what type of design it is with some designs from the impedance curve alone.

Those two example impedance curves look pretty similar (in shape, not in frequency of resonances or in amplitude) but they are dramatically different types of enclosures. And if you have a weird design it would be impossible to tell what kind of enclosure is was from the impedance graph alone. So if you can't reliably tell if the design has a sealed chamber or not based on the impedance curve alone you can't even guess what the cone motion (displacement graph) would look like. And never mind the smaller differences in trying to determine EXACTLY what the displacement graph would look like for ANY design based on impedance alone.

For example I seriously doubt you could draw the cone motion (displacement graph) for a simple sealed box impedance curve. Sure, you know in general what it's supposed to look like but I seriously doubt you could draw it accurately for any given voltage level.

MOVING ON...

bentoronto's post was vague and misleading ...

Ben's post was not vague at all, he said what he meant. And it's not misleading, it's just plain wrong.

If he was talking about the relationship between velocity and impedance, that would apply at all frequencies, not just at low frequencies as he stated.

And let's be realistic here, Ben didn't post to add to the discussion, he posted to derail it, as his number one mission on this forum is to speak about "the evils of resonances", as he implied in his second paragraph.

Going a bit further, Ben doesn't understand how enclosures work. A couple of examples - he believes there are only 3 types of enclosure suitable for a subwoofer that uses a moving coil driver - OB, sealed and front loaded horn because in his opinion these enclosure types don't have resonances - despite the fact that an undersized flh (like his beloved Klipschhorn) is riddled with resonances. And for OB he uses an antique woofer with maybe 2 mm xmax to fill in the bass BELOW THE KLIPSCHHORN.

And another glaring example is this - "... a driver with a resonance at 15 Hz ought to show a clear whomp-up at 15 Hz in a near-infinite baffle ..." When it was explained to him that only high q drivers will have a frequency response bump at or near resonance in IB, this was his comment. "Can we get an authoritative judgment from a recognized or credentialed expert about that one way or the other?"

I could go on with examples all day. I've tried several times to discuss this stuff with Ben but he won't talk to me. He clearly has no understanding of acoustic science and continually posts the same misinformation year after year.

So if you really want to believe Ben was talking about the relationship between velocity and impedance that's fine, but he wasn't. He was literally saying the cone motion follows the impedance curve. And from his latest post it looks like he's prepared to argue with you on that point. Besides, as I pointed out, if that is what he meant, it would not apply to only low frequencies as he stated.
 
Last edited:
just a guy wrote, "The sounds resonant boxes make are plenty accurate when properly designed, when tuned below 40 hz transient response and group delay are largely inconsequential. What you hear is frequency response and if frequency response if fine it will sound fine."

Really cannot agree at all. Because a port or passive radiator is a "sympathetic resonator" being driven by the active driver, that output must be delayed by one half cycle (more or less depending on tuning of the components) to align with the active driver motion. This means tuned at 40Hz the port is half a cycle behind the active driver in time that being 12.5mS. Tuning at 33.3Hz puts the delay at 15mS which is dangerously close to echo. Also consider the stored energy and delay of the port system, the port is still producing output after the active driver has stopped moving. Below 40Hz means muddy bass. Rumbling noises instead of distinct notes. Further, unless the port is at least a third the area of the active driver, power compression of the port will occur at relatively low SPLs. Typically have found ported system to not be able to produce a signal at the tuned frequency over 90dB while the active driver can play much louder. This is why larger ports and their cousins passive radiators gained popularity. A solution to power compression always experienced when using small ports.

Tossing out the "group delay" term in this discussion raises a red flag to me personally because this is a term experience has shown is used by people who do not know what a group delay is or how it applies to any filter theory. It is often the case this is caused by using a faulty model which makes a woofer in a box into a high pass filter with the low cutoff frequency as the low knee of the electrical equivalent of a high pass filter. A speaker is largely a mechanical device in reality. From an actual motion analysis, a woofer is a low pass 2nd order device with the end of the low pass band marked by the resonant frequency. A woofer exhibits (more or less) linear motion below the resonance frequency with motion falling off at -12dB/Octave above the resonant frequency. With this in mind the "group delay" occurs below the resonant frequency having little effect on anything. Just a guy has been lead astray by confusing faulty models as reality as so many have.

In the end small ports are a complete waste of time as power compression sets in long before any increase in useful SPL occurs. Large ports and passive radiators will play louder with less power compression concerns however, these always suffer from the delay necessary to make the coupled resonator work properly. 40Hz and above seems satisfactory for passive and large ports. Below that the delay starts to become quite apparent and muddy sound really sets in. To make that worse, use a large woofer, like an 18", which necessarily has a very floppy cone due to its' stupid large size and mud is the result. I never use a woofer over 10" for many reasons: one, it is not nearly as floppy; two, the overall frequency response in the piston region is much broader than a big floppy driver; three, several small cabinets are much easier to build rigid and light weight than a big stupid cabinet for a big stupid driver. Recall four 10" have the same cone area as a single 18". Which do you believe would have the better transient response, a 10" or 18". The one with the wider bandwidth of course. Use a 10" and then apply electronic equalization and do not go down the path of storing a bunch of energy in a coupled resonator system if you want good bass along with easy cabinets to build and move.

Big stupid woofers in big, heavy stupid cabinets are for people who want a lot of big and stupid around.
 
Really cannot agree at all. Because a port or passive radiator is a "sympathetic resonator" being driven by the active driver, that output must be delayed by one half cycle (more or less depending on tuning of the components) to align with the active driver motion. This means tuned at 40Hz the port is half a cycle behind the active driver in time that being 12.5mS. Tuning at 33.3Hz puts the delay at 15mS which is dangerously close to echo. Also consider the stored energy and delay of the port system, the port is still producing output after the active driver has stopped moving. Below 40Hz means muddy bass. Rumbling noises instead of distinct notes.

That's fine, you don't have to agree but I'll show here why I'm correct in saying below 40 hz transient response and group delay are largely inconsequential.

First thing I would like you to do is show what the average decay time is in a regular room. Modal ringing can (and often does) completely swamp out the decay time of a resonant enclosure when tuned below 40 hz. And modal ringing can occur well up into the midbass frequencies.

So what about outside? When used outside the resonant delay will be a bit more audible but the extra spl in an outside environment is more than worth the tradeoff.

Now note that I said it's LARGELY inconsequential BELOW 40 hz. What that means is that the port will blur the sound a bit across time but it's not really a big deal. And when I said 40 hz what I meant was that no one should ever consider tuning anything higher than 40 hz, 40 hz would be the absolute limit if you want decent sound quality.

By the time you get down to 20 hz tuning and lower the resonant delay is almost completely inconsequential, as these frequencies are more tactile than audible and the delay isn't readily apparent or important.

Further, unless the port is at least a third the area of the active driver, power compression of the port will occur at relatively low SPLs. Typically have found ported system to not be able to produce a signal at the tuned frequency over 90dB while the active driver can play much louder. This is why larger ports and their cousins passive radiators gained popularity. A solution to power compression always experienced when using small ports.

Note that I said "properly designed". That would include proper port sizing with 10 m/s velocity or less at xmax. This type of port will show very little port compression, if any.

And not to nitpick, but power compression and port compression are two very different things. Ports do not experience power compression ever. Power compression is a thermal compression that the driver alone experiences.

Tossing out the "group delay" term in this discussion raises a red flag to me personally because this is a term experience has shown is used by people who do not know what a group delay is or how it applies to any filter theory. It is often the case this is caused by using a faulty model which makes a woofer in a box into a high pass filter with the low cutoff frequency as the low knee of the electrical equivalent of a high pass filter. A speaker is largely a mechanical device in reality. From an actual motion analysis, a woofer is a low pass 2nd order device with the end of the low pass band marked by the resonant frequency. A woofer exhibits (more or less) linear motion below the resonance frequency with motion falling off at -12dB/Octave above the resonant frequency. With this in mind the "group delay" occurs below the resonant frequency having little effect on anything. Just a guy has been lead astray by confusing faulty models as reality as so many have.

Your entire post raises red flags for me but that's inconsequential. I have no been led astray and group delay, transient response, ringing, etc, are all terms that are on topic. The use of a technical term in a technical discussion shouldn't be raising any flags or implying things that are way beyond anything that was said.

In the end small ports are a complete waste of time as power compression sets in long before any increase in useful SPL occurs. Large ports and passive radiators will play louder with less power compression concerns however, these always suffer from the delay necessary to make the coupled resonator work properly. 40Hz and above seems satisfactory for passive and large ports. Below that the delay starts to become quite apparent and muddy sound really sets in. To make that worse, use a large woofer, like an 18", which necessarily has a very floppy cone due to its' stupid large size and mud is the result. I never use a woofer over 10" for many reasons: one, it is not nearly as floppy; two, the overall frequency response in the piston region is much broader than a big floppy driver; three, several small cabinets are much easier to build rigid and light weight than a big stupid cabinet for a big stupid driver. Recall four 10" have the same cone area as a single 18". Which do you believe would have the better transient response, a 10" or 18". The one with the wider bandwidth of course. Use a 10" and then apply electronic equalization and do not go down the path of storing a bunch of energy in a coupled resonator system if you want good bass along with easy cabinets to build and move.

None of this rings remotely true.

Again, I said "properly designed" not undersized ports. And ports don't suffer power compression, they suffer port compression.

In a regular room with high decay times all through the modal region and up into the midbass, it's unlikely that the sound above tuning would be clear but you can easily point out the resonant delay at tuning.

Large cones are not necessarily "floppy" or produce "muddy" sound. They are also not stupid.

Here's a large driver with a large cone. The cone is very strong. As you can see from the data sheet it's breakup isn't happening until after 1khz, so your superior transient response based on wider bandwidth is also incorrect. This driver can play cleanly well into the midrange, if the usual crossover point for a subwoofer is around 80 hz, how much bandwidth do you think the driver needs?

Big stupid woofers in big, heavy stupid cabinets are for people who want a lot of big and stupid around.

Yeah, not only wrong but completely untechnical. We can have a serious conversation if you want but you'll have to step it up a few notches.

The point here is that if you use the same driver and design a sealed box and a ported box with the SAME frequency response and listen to them in an average room with modal issues, in a blind test you won't be able to determine which one (sealed or ported) you are listening to with any statistical accuracy as they will both sound very much the same.

This is because what you hear is mostly frequency response. Resonant decay times in properly designed enclosures are large inconsequential below 40 hz.
 
The last part of my last post was screwed up because I had to run to work. A large portion that was supposed to be quoted was not and I didn't add the link. Here is the last half of the last post presented properly.

In the end small ports are a complete waste of time as power compression sets in long before any increase in useful SPL occurs. Large ports and passive radiators will play louder with less power compression concerns however, these always suffer from the delay necessary to make the coupled resonator work properly. 40Hz and above seems satisfactory for passive and large ports. Below that the delay starts to become quite apparent and muddy sound really sets in. To make that worse, use a large woofer, like an 18", which necessarily has a very floppy cone due to its' stupid large size and mud is the result. I never use a woofer over 10" for many reasons: one, it is not nearly as floppy; two, the overall frequency response in the piston region is much broader than a big floppy driver; three, several small cabinets are much easier to build rigid and light weight than a big stupid cabinet for a big stupid driver. Recall four 10" have the same cone area as a single 18". Which do you believe would have the better transient response, a 10" or 18". The one with the wider bandwidth of course. Use a 10" and then apply electronic equalization and do not go down the path of storing a bunch of energy in a coupled resonator system if you want good bass along with easy cabinets to build and move.

None of this rings remotely true.

Again, I said "properly designed" not undersized ports. And ports don't suffer power compression, they suffer port compression.

In a regular room with high decay times all through the modal region and up into the midbass, it's unlikely that the sound above tuning would be clear but you can easily point out the resonant delay at tuning.

Large cones are not necessarily "floppy" or produce "muddy" sound. They are also not stupid.

Here's a large driver with a large cone.
http://www.parts-express.com/pedocs/specs/294-689-bc-speakers-21sw152-4-specifications.pdf
The cone is very strong. As you can see from the data sheet it's breakup isn't happening until after 1khz, so your superior transient response based on wider bandwidth is also incorrect. This driver can play cleanly well into the midrange, if the usual crossover point for a subwoofer is around 80 hz, how much bandwidth do you think the driver needs?

Big stupid woofers in big, heavy stupid cabinets are for people who want a lot of big and stupid around.

Yeah, not only wrong but completely untechnical. We can have a serious conversation if you want but you'll have to step it up a few notches.

The point here is that if you use the same driver and design a sealed box and a ported box with the SAME frequency response and listen to them in an average room with modal issues, in a blind test you won't be able to determine which one (sealed or ported) you are listening to with any statistical accuracy as they will both sound very much the same.

This is because what you hear is mostly frequency response. Resonant decay times in properly designed enclosures are large inconsequential below 40 hz.
 
bentoronto,

Your posts here on this forum align to a specific, easily recognizable pattern commonly practiced on many technical/informational forums, most frequently by paid networking actors. So I must ask you straight up...

Are you here playing a disruptive/disinfo role as a gov/corporate/marketing company troll?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.