The frequency response and the phase response of a minimum phase sound source can be derived from each other using the Hilbert transform. There are some tools available to do this. I just wrote one in Excel that works great, and allows you to do the typical manipulations that are required to get wide-band amplitude and phase data for a driver. This includes scaling and combining responses obtained via different methods, and then extracting the minimum phase from the resulting "blended" amplitude response. It's called the "FRD response Blender" and I should be posting the initial release soon on my web site.
-Charlie
Hi All
I do measurement with arta lip but the vas make no sense, the BL is also much higher then the original factory measurement.
see here.
measured myself.
Thiele-Small parameters:
Fs = 28.84 Hz
Re = 6.33 ohms[dc]
Le = 93.12 uH
L2 = 612.12 uH
R2 = 4.40 ohms
Qt = 0.38
Qes = 0.44
Qms = 3.00
Mms = 39.36 grams
Rms = 2.376598 kg/s
Cms = 0.000773 m/N
Vas = 143.18 liters
Sd= 363.05 cm^2
Bl = 10.149454 Tm
ETA = 0.75 %
Lp(2.83V/1m) = 91.89 dB
Added Mass Method:
Added mass = 50.00 grams
Diameter= 21.50 cm
and factory
Rated power 110 Watt
Maximum power 140 Watt
Nominal impedance Z 8 Ohm
Frequency response fc - 7000 Hz
Mean sound pressure level 91 dB (1 W/1 m)
Excursion limit 22 mm
Resonance frequency fs 19 Hz
Magnetic induction 1 Tesla
Magnetic flux 1250 µ Weber
Height of front pole-plate 8 mm
Voice coil diameter 5 cm
Height of winding 1,45 cm
Cutout diameter 23,3 cm
Net weight 2,5 kg
D.C. resistance Rdc 5,7 Ohm
Mechanical Q factor Qms 3,19
Electrical Q factor Qes 0,38
Total Q factor Qts 0,34
Equivalent volume Vas 230 l
Effective piston area Sd 330 cm2
Dynamically moved mass Mms 42 g
Force factor Bxl 8,7 T · m
Inductance of the voice coil L 1,2 mH
Why so much differents.
rae there more software who I can test for TH easurement.
Why do LIMP clicks so much with stepped sine, it don,t work wel.
regards
I do measurement with arta lip but the vas make no sense, the BL is also much higher then the original factory measurement.
see here.
measured myself.
Thiele-Small parameters:
Fs = 28.84 Hz
Re = 6.33 ohms[dc]
Le = 93.12 uH
L2 = 612.12 uH
R2 = 4.40 ohms
Qt = 0.38
Qes = 0.44
Qms = 3.00
Mms = 39.36 grams
Rms = 2.376598 kg/s
Cms = 0.000773 m/N
Vas = 143.18 liters
Sd= 363.05 cm^2
Bl = 10.149454 Tm
ETA = 0.75 %
Lp(2.83V/1m) = 91.89 dB
Added Mass Method:
Added mass = 50.00 grams
Diameter= 21.50 cm
and factory
Rated power 110 Watt
Maximum power 140 Watt
Nominal impedance Z 8 Ohm
Frequency response fc - 7000 Hz
Mean sound pressure level 91 dB (1 W/1 m)
Excursion limit 22 mm
Resonance frequency fs 19 Hz
Magnetic induction 1 Tesla
Magnetic flux 1250 µ Weber
Height of front pole-plate 8 mm
Voice coil diameter 5 cm
Height of winding 1,45 cm
Cutout diameter 23,3 cm
Net weight 2,5 kg
D.C. resistance Rdc 5,7 Ohm
Mechanical Q factor Qms 3,19
Electrical Q factor Qes 0,38
Total Q factor Qts 0,34
Equivalent volume Vas 230 l
Effective piston area Sd 330 cm2
Dynamically moved mass Mms 42 g
Force factor Bxl 8,7 T · m
Inductance of the voice coil L 1,2 mH
Why so much differents.
rae there more software who I can test for TH easurement.
Why do LIMP clicks so much with stepped sine, it don,t work wel.
regards
Stage/recording mics are generally have lower noise and better dynamic range, but may not be particularly flat. However, most publicise a general frequency response you can use to make up a calibration file.
Yup.
kees, your resonant frequency is very different to the published spec, this will also be affecting the VAS. If the resonant freq comes down the VAS should go up. The overall Q is not that different.
Have you broken in the drivers? have a read of this app note from SB accoustics about measuring T/S params http://www.sbacoustics.com/index.php/download_file/-/view/191/
I had the same problem with ARTA with the clicks and pops. Try REW
Tony.
Have you broken in the drivers? have a read of this app note from SB accoustics about measuring T/S params http://www.sbacoustics.com/index.php/download_file/-/view/191/
I had the same problem with ARTA with the clicks and pops. Try REW
Tony.
Hi Tony
These drivers are old, from 1997 so I think she will be ready with broken in if she are so old.
I do use a headphone output of a X-fi soundcard now, I am going to build a amplifier of 10 a 20 watts meaybe then it go the better way. Or limp is now accurate.
I was go for REW but I have to register on a site there with a lot of questions.
thanks for your help.
kees
These drivers are old, from 1997 so I think she will be ready with broken in if she are so old.
I do use a headphone output of a X-fi soundcard now, I am going to build a amplifier of 10 a 20 watts meaybe then it go the better way. Or limp is now accurate.
I was go for REW but I have to register on a site there with a lot of questions.
thanks for your help.
kees
LIMP is very accurate for T/S measurement, there is only really one thing that needs to be calibrated to make sure its working properly.
A few points here -
First of all, your Re figure is significantly higher - 6.3 ohms compared to a claimed 5.7 ohms. This is going to throw your sensitivity and BL figures out a bit.
Have you confirmed with a multimeter whether the driver is actually 5.7 ohms or really is 6.3 ohms as claimed by LIMP ? Remember to subtract out the resistance of the multimeter leads when measuring this.
Assuming that LIMP is wrong and the multimeter and manufacturers figure for Re is correct you need to calibrate LIMP.
In Setup->Measurement make sure the value for Reference Resistor is the exact value of the series resistor you're using in your test harness - measure it with a multimeter and remember to correct for any resistance of the multimeter leads. I use 33 ohms on my test harness.
Next you need to perform a calibration, this allows for gain errors in the left and right channels of your sound card input. Connect your impedance measuring harness but leave the speaker disconnected, then perform the calibration under Record->Calibrate. This is described in section 4.7 in the manual.
Now that its calibrated what you want to do to verify it is to perform an impedance sweep of a resistor of known value somewhere in the 4 to 20 ohm range, compare the exact figure you get to the reading from a multimeter and they should now agree. Repeat your test with the speaker.
Secondly, your value for Fs being a lot higher than the manufacturers claimed figure is not only accurate, but is commonplace. Nearly all manufacturers claim lower Fs figures than reality, some by a large amount. For example my Visaton W300S has a claimed Fs of 22Hz, but measured is 28Hz, measured by multiple methods including manually sweeping it with an audio generator, resistor and voltmeter.
There is nothing that limp could do wrong to read Fs incorrectly as it is simply looking for the peak in the impedance curve. Even if the calibration above was wrong the Fs figure will still be correct.
The reason why manufacturers figures are often so much lower is because Fs is not a static figure but varies with excursion, temperature, age, and other factors.
Fs is dictated by moving mass, which stays constant, and suspension compliance, which is anything but constant, and varies dramatically with the above factors.
The mass/compliance of a woofer forms a harmonic oscillator, however because the suspension compliance is non-linear with excursion the resonant frequency changes with excursion. (The suspension would have to be perfectly linear for the resonance frequency to be independent of excursion)
Typically this manifests as Fs being high at very low power/excursion levels where T/S parameters are usually measured, and dropping quite considerably as excursion is increased. Depending on the suspension material the shift can be modest or dramatic.
On another driver I have - a Coral Flat 8 II, (8" full range driver with cloth surround) the Fs drops all the way from 48Hz to 37.5Hz as excursion increases!
It seems that many manufacturers measure and quote Fs at a relatively high excursion level while most DIY'ers measure it at very low excursions, thus the apparent contradiction. It is still debated whether the low excursion or high excursion T/S parameters are more valid or not when it comes to a speakers performance in the final cabinet.
A number of other figures have knock on effects from compliance being stiffer than the manufacturers claims - Fs will be higher, Vas will be lower, often much lower, Qms may be noticeably different.
Another thing I notice about your measurements is a significantly different Sd figure, a figure which you enter yourself either in the form of Sd or diameter. Depending on whether you used delta-mass or delta-compliance as the measurement method, this difference in Sd will also cause errors. Sd should be calculated by measuring the diameter of the cone including half the surround width on each side.
Regarding the clicking noise when using stepped tones, to fix that simply go into Setup->Generator and reduce the output level to at least -2dB. If its set to 0dB its not uncommon for clipping to occur depending on your sound card setup. Mine clicks badly on 0dB but does not click at all on -2dB.
Last edited:
I did again a measurement, I measure indeed 5.7 ohm, but the leads are 0.4 ohm, and the measurement was 6.1 ohm so this is 6.1 - 0.4 = 5.7.
the QMS is higher as the speaker factory value, and so on, but I think now
everything is close, I go make a amplifier of 10 watts and use low value 10 ohm resistor I think
it is better, I make a very simple amp with opamp and mosfets or darlingtons.
regards
kees
Thiele-Small parameters:
Fs = 34.53 Hz
Re = 5.70 ohms[dc]
Le = 215.76 uH
L2 = 989.08 uH
R2 = 15.79 ohms
Qt = 0.50
Qes = 0.58
Qms = 3.56
Mms = 31.23 grams
Rms = 1.904538 kg/s
Cms = 0.000680 m/N
Vas = 104.09 liters
Sd= 330.06 cm^2
Bl = 8.140839 Tm
ETA = 0.71 %
Lp(2.83V/1m) = 92.07 dB
Added Mass Method:
Added mass = 24.17 grams
Diameter= 20.50 cm
the QMS is higher as the speaker factory value, and so on, but I think now
everything is close, I go make a amplifier of 10 watts and use low value 10 ohm resistor I think
it is better, I make a very simple amp with opamp and mosfets or darlingtons.
regards
kees
Thiele-Small parameters:
Fs = 34.53 Hz
Re = 5.70 ohms[dc]
Le = 215.76 uH
L2 = 989.08 uH
R2 = 15.79 ohms
Qt = 0.50
Qes = 0.58
Qms = 3.56
Mms = 31.23 grams
Rms = 1.904538 kg/s
Cms = 0.000680 m/N
Vas = 104.09 liters
Sd= 330.06 cm^2
Bl = 8.140839 Tm
ETA = 0.71 %
Lp(2.83V/1m) = 92.07 dB
Added Mass Method:
Added mass = 24.17 grams
Diameter= 20.50 cm
REW?
What is REW I am not familiar. I need to measure frequency response of my speakers. I have a behringer mic/dsp with rta but I was told I need gated pink noise. I didn't realize that arta won't work with my laptop.
Although clunky, REW is top notch. Send a donation once you get the hang of itBetter than Holm and Arta IMO.
What is REW I am not familiar. I need to measure frequency response of my speakers. I have a behringer mic/dsp with rta but I was told I need gated pink noise. I didn't realize that arta won't work with my laptop.
most cards are full duplex for as far I now.
so you are able to use it, if I am right..
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.