Tweeter resonant frequency missing

[Dave Bullet]

Trying to use Speaker Workshop to measure impedance.

I have been through all the calibration steps and have succesfully managed to measure passive components within close tolerances.

I have no problem measuring free air impedance of woofers. AS expected, I see the Fs of the driver and rising impedance with frequency.

However, when I measure tweeters - there is no resonant frequency at all. The impedance is flat until the natural rise toward the top end of the frequency range.

I am using a cable jig with an 8.2ohm resistor.

Attached is the tweeter free air impedance measurement.

Any help much appreciated

Thanks,
David.

 

[Dave Bullet]

PS: I have measured 2 different tweeters and woofers. Attached below is the Wharfedale woofer

No setup changes were made between measurements (other than switching drivers over).

 

[Keruskerfuerst]

What type of tweeter are you using?

 

[Dave Bullet]

Sorry - should have said.

The image posted is a 3/4" metal dome tweeter (Wharfedale proprietary). The other tweeter I tested is a Vifa D19-TD-05 fabric (plastic?) 3/4" dome. The Vifa plot is below (similar symptoms).

Both tweeters were removed from working speakers so nothing wrong with the tweeters.

I have tried different sample sizes, rates, repeat counts to no avail.

Thanks,
David.

 

[richie00boy]

Do the tweeters have ferrofluid?

 

[Dave Bullet]

I'm not sure. It looks like I can remove the faceplate on the Wharfedale tweeter but am reluctant to do that.

Good thinking... I thought "maybe this is a case of ferrofluid flattening out the resonant peak", but surely even the most damped / fluid filled tweeters have some sort of peak? (even if only a few dB above nominal?)

They both seem too flat - even for ferrofluid tweeters.

From memory, I don't think the Wharfedale's do. I couldn't find the D19 spec in the Tymphany archive pages. will google it now.

Cheers,
David.

 

[soongsc]

Did you do the interchannel latency calibration?

 

[Dave Bullet]

Hi soongsc,

Yes I did. Found -14usec flattened the phase rise toward the upper end as shown in the attached image (zoomed in)

I also found a 500msec latency got the MLS pulse starting consistently around 400msec.

Cheers,
David.

 

[soongsc]

Hi soongsc,

Yes I did. Found -14usec flattened the phase rise toward the upper end as shown in the attached image (zoomed in)

I also found a 500msec latency got the MLS pulse starting consistently around 400msec.

Cheers,
David.

Is the greem line prior to interchannel latency?
If it is, then it seems the tweeter impedance must be right. I also have a tweeter that has a flat impedance and flat phase as well.

 

[Keruskerfuerst]

Possible reasons for "flat" impedance curve:
1. Ferrofluid
2. copper short circuit rings.

 

[soongsc]

Verly light flexible diaphragmed drivers usually have flat impedances.

 

[Dave Bullet]

Hi Soongsc,

Is the greem line prior to interchannel latency?

I don't understand... the green line (phase) in the measurement.calibration graphic crossses over about 500Hz. I can try and tweak the interchannel delay a little more to get the phase down to zero degrees at 20KHz to see if that helps.

Thanks for your replies. It's possible both tweeters have a flat impedance, but I'm still suspicious! I've got a pair of Peerless tweeters and new Seas 27TBFC/G to test (I know those have some sort of resonant frequency - even though they are ferrofluid filled).

I'll post the results of those tomorrow.

Cheers,
David.

 

[augerpro]

REcently on another forum I seen some one who didn't get a peak for some Accuton woofers. Turns out the VC was fused and wouldn't move.

 

[soongsc]

This is what my measurement.calib looked like. What I did was asjust the interchannel latency so that there is a reagion where you adjust the delay, and the curve doesn't move, until the curve starts drooping.

This won't make the resosnace peak show up, but will correct the impedance in the high frequencies. Verly light flexible diaphragmed drivers usually have flat impedances. A typical example is the MANGER driver.

 

[Dave Bullet]

Thanks augerpro and soongsc,

The tweeters are both operational so I know its not a voice coil / motor failure issue.

Soongsc,

It looks like my interchannel isn't as finely tuned as yours. I sse you are only a fraction of a degree out at the top end.

I will correct my phase to be as close to flat as possible like you suggest. At least it will make high frequency (>10KHz) measurements more accurate.

I'll try those other tweeters with a more finely tuned interchannel delay and see how I go.

Thanks for all your help. Reassuring to know at least there's nothing fundamentally wrong with my setup. I'm the kind of person that gets each stage right before progressing, otherwise errors compound errors and you don't know where it started - wasting more time troubleshooting.

I think the most important virtue is patience when it comes to anything new! (and 2nd.... not being afraid to ask the dumb questions!) 🙂

Cheers,
David.

 

[5th element]

It could also be that the vifa D19 IS showing the bump at resonance. The wiggle in the curve at around 2500 hz could be the result of a highly damped resonance due to ferrofluid, although it does look a 'bit' odd in its shape. What you really need to do is get your hands on a tweeter we know has a highly pronounced low resonance, such as a non ferro seas etc. Of course your woofer measurements are turning out fine.

 

[Dave Bullet]

Thanks Matt,

I too thought the 2.5KHz hump might be due to Fs. After all it is probably a bottom of the line Vifa 3/4" dome - so wouldn't be surprised with an Fs > 1500Hz. Due to the weird shape, I began to wonder if my setup was becoming increasingly inaccurate at guessing Fs the higher in frequency.

I will do an added mass VAS measurement on one of the woofers just to confirm it raises Fs appropriately.

I'll try my 27TBFC/G (although with Ferrofluid) - being a new tweeter and having Seas and other (ie. Zaph) impedance data should confirm my setup.

I'm now wondering how the impedance profile of old tweeters (ferrofluid or not) might change over time. After all, I know the Wharfedale tweeters are at least 16 years old.

Cheers,
David.

 

[planet10]

I have seen ferrofluid tweeters with no dicernable resonance bump. I've also seen some vinatge non-ferrofluid cone tweeters with no bump (usually aperiodic).

The one D19 i have a measure of (at least 16 years old) does some a bump -- not much of one mind you -- and it is in the 2-2.5 k region

dave

 

[planet10]

and here is another (earlier) measure of the similar SEAS 19mm dome.

dave

 

[Keruskerfuerst]

Driver units with a flexible shaft dome have not resonance frequency in common sense.