Bass Reflex Enclosure Tuning Using Impedance Plot

tekuhn

tekuhn

Active Member
I have a strange situation that I'm hoping someone more experienced can explain to me. I am installing a Celestion TF1225 driver in a ported 2.6 cu.ft. enclosure and having a hard time getting the ports tuned. The drivers advertised Fs is 63Hz and mine measured 60Hz in free air. Using WinISD, I have decided to tune the box to 57Hz for the desired alignment. When I made the ports the dimensions specified by the calculator, the tuning dip showed 41 Hz instead of the expected 57Hz. As I shorten and widen the ports to raise the tuning, I can never raise it to the desired tuning. When I tune the box to make the two impedance spikes equal, it shows the tuning to be about 45 Hz. My understanding is that when the two spikes are equal that the box resonant frequency Fb and the driver resonant frequency Fs are equal. Shouldn't the tuning dip show about 60Hz when the spikes are equal? I have tried removing all the foam to make sure I don't have it over-stuffed, and it doesn't significantly change anything. I am checking impedance using REW and a homemade test box which calibrates perfectly and I trust the results. Any ideas what's going on?

TF1225_Impedance.jpg
 
Last edited:
Register to hide this ad
As far as I know, the only way to get Fs and Fb to be equal is if the box size is "perfect".
And 60hz and 57hz are close enough that the difference doesn't matter..

But I have to ask, why would you want that?
Looks like a recipe for a pronounced [vaguely] 60hz "thump" to me since the two resonances involved will re-enforce each other.

I'm no expert [not even close], but I do wonder if what you consider "optimal", given the box size, isn't actually the case.
 
I'm not sure what you mean about the box size being perfect. I ought to be able to tune the enclosure to any frequency I want (regardless of volume), and if I tune the box to the Fs of the driver, wouldn't the peaks be equal and the low dip in between them be at Fs? My goal is not for the Fb and Fs to be equal, I'm shooting for Fb to be 57Hz. I'm not sure if you modeled it, but as the TF1225 is a pro audio driver and it does not offer very low extension but is high efficiency. Tuning the box for 57Hz will offer a slight bump up at 100Hz, but the best F3 of about 52Hz and matched with a sub. My point about the peaks being equal is that it should indicate that the box is tuned to about 60 Hz, yet the graph shows that it is tuned to 45 and I don't understand why.

shot.jpg
 
All that matters to determine the Fb tuning is the minimum which
should be very close to the zero crossing in the phase. The height
of the peaks depends on the Qms of the driver, box absorbtion and
leak losses, and vent losses. I don't worry much about the peaks.
 
Last edited:
Tuning Fb to Fs falls out of Butterworth alignments both B4 and B6
alignments call for a tuning ratio, h = 1 which means that Fb = Fs.

On the other hand you should tune to get the response that you want.
A bigger box provides more output at Fb, box and port losses attenuate
the output at Fb. Most guess at typical losses with a Q = 7 and then a
B4 alignment requires a Vb = Vas, and a B6 Vb = .5 Vas. Reducing the
box size by 1/2 attenuates the output at Fb by 6 dB and the B6 alignment
gets the system back to -3dB at Fb through the use of a peaking (6dB)
high pass filter.

I like low tunings because the excursion limited power handling is
high at Fb. 40 Hz and lower is good without a sub and with an
appropriate driver. I see that you're using a sub so no need to tune
low.

Peaking at 100 Hz will help reduce the amount of baffle step that is required,
is that what you had in mind?
 
I am apparently not making my question or concern clear. I am not asking what is the ideal alignment for my application. I am expressing concern that my impedance plot does not make sense for a driver with a free-air Fs of 61Hz. Here is the information that I am referencing from http://www.mh-audio.nl/calculators/impedancecurve.html.

"The Bass Reflex alignment is through the impedance minimum between the two impedance peaks defined and therefore accurately determined only with a impedance measurement.


Depending on how the 2 peaks compare to each other, in terms of height,
we can consider 3 particular cases:
  1. The two peaks match in height.
    This means that the resonant frequency of the driver in free air (fs) matches the resonant frequency of the box (fb).
  2. The first peak is higher than the second means that fb > fs.
  3. The second peak is higher than the first peak means that fb < fs.

Depending on how reactive the speaker system is, the phase angle will take different values.
The phase angle will have 0 degrees at the resonant frequency and at the point where the impedance starts to rise due to voice coil inductance."
 
Using the information above, when the impedance peaks are equal - which they basically are in the provided plot - shouldn't the low impedance point between the two peaks be occurring at 61Hz? And if so, what might be causing it to occur at 45Hz? There are no air leaks.
 
I'd like to know where whoever the author is came up with that theory as I
don't believe that it is correct. I've studied the classic papers in the AES for
over 30 years and have never seen that claim.

I'll repeat to determine box tuning Fb all that matters is the minimum.

The acoustical impedance seen by the driver cone looking into the box is a
series resonant acoustical circuit that "looks" like a short at resonance. The
cone should appear as if it were blocked so that it cannot move and this is
why you see a minimum at Fb. Ideally for a very high Q box the impedance
at Fb = voice coil resistance (Rvc) since it is not moving. Losses in the box
raise the depth of the acoustical short and then it is higher than Rvc.
 
I've sent an email using the Contact link from the website. Perhaps they will shed some light on it for me.

I removed the CD horn from the cabinet so there's a huge hole (wide and short) which should raise the tuning substantially, I would think. I ran the impedance test and the box tuning shows to be about 80Hz. Again, I would have expected that to raise the tuning in a major way, but doesn't seem to have. I am not looking at this with any experience, mind you, just using the port length calculator, it looks like changing the port length even a few tenths of an inch will make a big change, but it's not working that way.
 
1: I find that most on line calculators are mostly a crude guide line as there unknowns the equations assume, and few items in the real world both match the online calculators and achieve respectable performance.
2: All information I've got from Pete was spot on. He knows what he's talking about, provided you can understand it. Usually takes me several tries.
 
Thank you all for the help - I'll post again when I have more information. I just went back to the sellers website to look at their published specifications for the TF1225, and they now have something totally different from before. I need to compare to the manufacturers specs. It's possible that this driver is not well suited to a 2.7 cu.ft. vented box at all. :mad:
 
  • Resonant Frequency (Fs)55.6 Hz
  • DC Resistance (Re)5.08 ohms
  • Voice Coil Inductance (Le)0.86 mH
  • Mechanical Q (Qms)3.11
  • Electromagnetic Q (Qes)0.42
  • Total Q (Qts)0.37
  • Compliance Equivalent Volume (Vas)2.38 ft.³
  • Mechanical Compliance of Suspension (Cms)0.17 mm/N
  • BL Product (BL)14.32 Tm
  • Diaphragm Mass Inc. Airload (Mms)48.52g
  • Moving Mass Of Diaphragm (Mmd)39.60
  • Maximum Linear Excursion (Xmax)2.5 mm
  • Surface Area of Cone (Sd)531.3 cm²
Specs of your driver.
Looking at your Qts and Vas values, your box seems a "bit large", but it should not be an issue, other than physical size being a bit big.


As to why your box tuning Fb does not change with altering the port dimensions, ....I have never experienced that, nor can guess as to why.
Perhaps recheck the dimensions you are using for you box and so on, but Measured Impedance IS THE BOX TUNING Frequency, (Fb) no matter what a calculator says.

Not to sound confrontational I hope, but measured impedance is the absolute tuning frequency above all else.

Absolutely Pete knows what he is talking about.
 
Last edited:
I'm very appreciative of Pete's, and everyone else's help. I apologize if I have come across as not. I know and agree that the low impedance dip between the peaks indicates the tuned frequency and it should line up with the phase crossing at 0. For some reason, when I make my ports the diameter and length that all calculators seem to agree with for my size box, Fb measures WAY lower than expected. I have tried shortening and increasing the diameter of the ports to raise the tuning, I have been unable to get anywhere close to the 55-57 Hz where I would like to be. Even with the port tubes removed, which leaves me with ports the length of the front baffle thickness, it is still tuned too low. As I described above, I finally removed my cd horn to create a large gaping hole (port), and the Fb finally got up to 80Hz as indicated by the dip.

BTW, the T/S parameters you posted above are significantly different from what they advertised when I bought my drivers about 18 months ago. I just sent an email to PE earlier tonight asking about it. At the time of purchase their advertised "Optimal box size using BassBox was 3.08 cu.ft. for an F3 of 49 Hz. Now they are recommending 1.4 Cu.Ft. for an F3 of 70.55 Hz. I would have never chosen this driver if I knew it was this much of a mismatch to my cabinets and my desired performance requirements.

But, regardless of how well these drivers are suited to my cabinets, the box should still tune to the desired frequency determined by the box volume and the port area - and nothing else, right?
 
Here is a copy of a spec sheet from Celestion for the TF1225 at the time I bought mine. As you can see, these specs are much more suited for a box the size of mine:

TF1225.jpg
 
I've just run some simulations and they predict exactly as you say - with the two peaks at the same height the resonant frequency should be the same as the free air resonance. I have no idea why you're getting strange readings. Are you sure your test set-up is calibrated correctly?

edit - having thought about it, miscalibration wouldn't offer an explanation. Even if the impedance values were wrong the system would still correctly identify when the two peaks were the same height.
 
Last edited:
rothwellaudio said:
I've just run some simulations and they predict exactly as you say - with the two peaks at the same height the resonant frequency should be the same as the free air resonance. I have no idea why you're getting strange readings. Are you sure your test set-up is calibrated correctly?

edit - having thought about it, miscalibration wouldn't offer an explanation. Even if the impedance values were wrong the system would still correctly identify when the two peaks were the same height.
Click to expand...

Thank you for checking that. I have tested the calibration of my DIY test box multiple times using non-inductive resistors of various values and it always produces a perfectly flat line indicating the exact value of the resistor. Yesterday they released a new version of REW that includes additional calibration steps to increase accuracy of the impedance testing. I have loaded it, but not retested yet. I honestly don't expect it to change.

The author of the article I posted above responded back and I don't understand what he said in the second sentence. It does sound like he's saying the same thing that 45rpmspinner said:

"The lowest impedance of the impedance curve is the resonant frequency ( fb ). If the impedance peaks are the same the reflex port is perfect tuned with the box volume. The resonant frequency ( fb ) should be always lower than the resonant frequency ( fs ) of the woofer. I’m sure there is no mis alignment with your loudspeaker."
 
Actually, rather than suspect the accuracy of the impedance values of my testing, maybe I should be suspect of the frequency values. REW uses the stereo output of the PC sound card for testing. It uses one channel as the reference and the other for the testing. If there was some latency I could see the graph being skewed along the frequency axis. I think I'll slow down the sweep and see if it changes the results.
 
45rpmspinner said:
Looks like a recipe for a pronounced [vaguely] 60hz "thump" to me since the two resonances involved will re-enforce each other.
Click to expand...

tekuhn said:
The author of the article I posted above responded back and I don't understand what he said in the second sentence. It does sound like he's saying the same thing that 45rpmspinner said.
Click to expand...
I don't think so. If the box resonant frequency and the free air resonant frequency are the same, there is no reinforcement. In fact, the opposite is true. The resonance of the box/port damps the motion of the cone rather than exaggerating it, hence the impedance drops to a minimum at that point.
Anyway, I'm still not sure why you're getting the results you are. :dunno:
 
If in fact the published speaker specs are incorrect then this goes a long way to why you are having tuning issues. The fact that you have to open a large port in order to get a significant change in tuning. Could point to either your speaker specs are off or you box is to big or both.

Best to get good specs and go from there.
 
The box tuning has little to do with the driver specs.
Would you post a picture or provide dimensions of the box with woofer and port placement?
The basic model assumes a lumped model where the box is a simple compliance
and the port is a mass. A long enclosure will behave more like an MLTL with the tuning
being off.
There is an end correction when calculating a port tube so be sure to use a "good"
calculator.
You could block the port and then with a piece of wood block off the waveguide opening
until you get the desired tuning, then use that cross sectional area for the final
cut out. You could use a slot without a duct, shape shouldn't matter much.

Which port tube calculator did you use and what were your input values?

A low tuning should sound good actually, have you tried listening to it?
 
You must log in or register to reply here.
Back
Top Bottom