Seas 27TBFC/G impedance plot mismatch

[GrahamS]

Hello everyone, I am a long time htguide reader but this is my first post. My builds so far include PE Encore bookshelf speakers and some Natalie P towers (which I love).

I recently purchased two Seas 27TBFC/G tweeters for a BAMTM build. I tested both of them with my WT-3 and found about a 1 ohm variance in the impedance plot. This impedance difference was fairly consistent throughout the frequency range. Is it normal for two drivers to have this much difference in their impedance measurements? If not, which driver is defective? If I can still use these drivers, should I increase the tweeter series resistor (R6) value from 7 to 8 ohms for the lower impedance driver in order to level the high frequency response? Attached below is the impedance plot graph. Any help is appreciated.

[augerpro]

Been breathing solder fumes all day so correct me if I'm wrong, but isn't it more like 1/4 ohm difference?

[jkrutke]

Been breathing solder fumes all day so correct me if I'm wrong, but isn't it more like 1/4 ohm difference?

Yup, looks like 1/4 ohm to me also. Practically nothing to worry about, well within what I would expect for consistency. One will probably be a mere fraction of a dB different than the other.

Here's one of my recent 27TBFCG impedance plots.

[GrahamS]

Yes, you are correct it does read about only a .25 ohm difference.ops: Most of the other drivers that I have measured in the past were spot on. Correct me if I am wrong, but I think I overreacted.

Thanks for your help

[pedroskova]

Stupid question...what would cause the double peak / shoulder in the impedance? It's just visible in John's curve as an asymmetry in the peak.

...a chamber reflection...resonance?

[GrahamS]

The reason for the asymmetric impedance peak is beyond me. When I tested my drivers they were still cold out of the shipping box. I know that it is unrelated, but when I ran an impedance sweep on my Natalie P towers there was a wrinkle in the curve around 160 Hz. Software confirmed that I should have expected a resonance around that frequency due to the tower height. I followed jkrutke's recommendation and added 8” of dampening material to the bottom of the enclosure. The solution worked. The impedance plot smoothed out and the speaker has a more even response around that frequency.

Thank you Jkrutke and Augerpro for your help. I appreciate the contributions that you two and other senior htguide members make with your designs and knowledge. After I make these BAMTM's for my brother, I plan to move on to the ZDT 3.5 or the ER18NRX Zaph design. How much MDF dust do I want to clean up this winter...?

[Dave Bullet]

Silly question... did you simply swap one 27tbfc/g with the other and remeasure? - or did you have an extra length or different gauge hookup wire between the two tweeters.... The constant difference in impedance across frequency could be additional series resistance.

[jimangie1973]

Stupid question...what would cause the double peak / shoulder in the impedance? It's just visible in John's curve as an asymmetry in the peak.

...a chamber reflection...resonance?

Probably so.

I have a pair of Milleniums that I opened up to replace the voice coils. I noticed the damping felt pad was incorrectly placed in one of the units. I glued it in the correct position and noticed a second peaking above the resonance frequency. The frequency response also showed more energy around the second impedance peaking area relative to prior. I removed the glue and simply adjusted the pad position. The second impedance peak went away.

[jkrutke]

Stupid question...what would cause the double peak / shoulder in the impedance? It's just visible in John's curve as an asymmetry in the peak.

...a chamber reflection...resonance?

It's a restriction between the front chamber and the rear chamber, usually from too small of a pole piece hole. The front chamber (behind the dome) has it's own characteristic resonance which is higher in frequency. The rear chamber and the front chamber add up to another volume which also has a characteristic lower frequency.

Ferrofluid improves on this by damping the two peaks into almost one. (looks like the impedance plot I posted has a bit more ferrofluid) This shows why I prefer the 27TB with the F over the one without - the twin peaks are pretty bad in the 27TBCG. A few select other chambered tweeters have superior pole piece design, giving us a single impedance peak without ferrofluid. The Vifa XT25, the RS28 and any of the higher end Scan Speaks for example. The Seas 27 series tweeters are among the best in the world but they do have their faults and a restrictive pole piece one of them.

[dlr]

It's a restriction between the front chamber and the rear chamber, usually from too small of a pole piece hole. The front chamber (behind the dome) has it's own characteristic resonance which is higher in frequency. The rear chamber and the front chamber add up to another volume which also has a characteristic lower frequency.

I suspect it's not a form of double resonance interaction. The old Morel MDT-30 had this type of impedance and had similar construction. It had very poor damping material in the vent with a standard large diameter felt ring on the top, nothing to restrict the opening. I've seen this same characteristic in other tweeters as well. It also came with the damping foam poorly placed. The comment MDT-32 in the next graph comments are wrong, it was the old MDT-30. Note how the red curve shows changing "double peaks", but little change outside of that area. I believe that it's actually a shift in the damping range effectiveness.



Were you to remove the damping material in the vent completely you'd see a single, very high magnitude peak I believe, a bit higher in frequency. My guess is it would be around 1KHz based on what is seen in the two curves of that Seas unit. The damping material looks to be most effective from about 600-1700Hz or so. It's damping the single resonance peak, essentially dragging (damping) that range down with some small shift outside this area.

If someone is willing to remove the vent damping and re-measure we could see what the vent damping alone is doing. If it is a restriction, a double-peak will be seen at it's maximum with nothing in the vent.

Dave

[Quwiksilver]

It's a restriction between the front chamber...

John,

Your experience testing and building over the years seems to have imparted quite a bit of knowledge in driver design. Have you ever considered designing your own drivers? Can we ever hope for some high value, stellar performing drivers made by someone with "Zaph Licensed" technology?

Scott

[dmalphur]

I think you figured out his retirement plan!

[jkrutke]

Were you to remove the damping material in the vent completely you'd see a single, very high magnitude peak I believe, a bit higher in frequency. My guess is it would be around 1KHz based on what is seen in the two curves of that Seas unit. The damping material looks to be most effective from about 600-1700Hz or so. It's damping the single resonance peak, essentially dragging (damping) that range down with some small shift outside this area.

If someone is willing to remove the vent damping and re-measure we could see what the vent damping alone is doing. If it is a restriction, a double-peak will be seen at it's maximum with nothing in the vent.

I've got a couple of those. Heres the last one I did, the Seas DXT. See attached. Blue is the stock impedance with a light wad of damping in the pole piece. Purple is with the damping removed. Red is with the rear chamber sealed off.

Indeed, the double peak with no damping (purple) is pretty sharp. The center resonance frequencies based on the volumes don't move, but where the smaller peak shows up depends on the destructive summing. In the blue stock plot, the damped 550Hz peak has less destructive summing with the 1000Hz front chamber resonance causing a little more of a plateau shape.

Now if the pole piece were free flowing enough, the 1000Hz resonance wouldn't be there in any of these plots. Who knows if the designers hit what they thought was happy medium. Take away enough steel from the pole piece and you lose BL, so they might have seen a little secondary peaking as a trade off.

The best tweeter rear chamber I've ever seen was made by B&W. It was from their flagship system. Most tweeters have a small hole leading to a larger rear chamber. Small hole + large rear chamber = restriction through the hole. In the B&W tweeter, the hole was the rear chamber, just really long, tapered and well damped. There was no transition to a large volume.

[jkrutke]

John,

Your experience testing and building over the years seems to have imparted quite a bit of knowledge in driver design. Have you ever considered designing your own drivers? Can we ever hope for some high value, stellar performing drivers made by someone with "Zaph Licensed" technology?

Scott

Nah, my "driver design skills" are nothing really special, just a few observations I've made over the years. Nothing more than the rest of us freaks around here.

[dlr]

Indeed, the double peak with no damping (purple) is pretty sharp. The center resonance frequencies based on the volumes don't move, but where the smaller peak shows up depends on the destructive summing. In the blue stock plot, the damped 550Hz peak has less destructive summing with the 1000Hz front chamber resonance causing a little more of a plateau shape.

Now if the pole piece were free flowing enough, the 1000Hz resonance wouldn't be there in any of these plots. Who knows if the designers hit what they thought was happy medium. Take away enough steel from the pole piece and you lose BL, so they might have seen a little secondary peaking as a trade off.

I still don't see it as a dual resonant chamber. One of my earliest experiments was with a D27TG-45 back in May of 2000. I've got a page at my site in Tweeter Tweaks/Adding a Rear Chamber to a Tweeter on this. The graph below is one of a number there:



This curve is for an empty chamber. I measured without a chamber altogether as well and it is not much different, this is on that page as are the corresponding SPL curves. The chamber shifts both impedances a very tiny bit when there is no damping in the chamber. Note the second peak for this tweeter at about 2K. This is a pipe resonance. It also appears to some degree if the original, or any, stuffing is placed in the chamber, but is not pressed against the chamber side vent opening. If not, the pipe resonance will be only partially damped. I've measured this and found that some stuffings in chambers are not optimal or placed badly as well, allowing the pipe resonance to establish.

At the time I did this work I had not yet come upon the vent stuffing issue. The Vifas I started with had none at all, I was unaware that it is used at times. But note that with a properly stuffed chamber, the pipe resonance is totally damped:



Now here is the chamber with a different wool felt stuffing (standard F11, not Lamb's wool) I experimented with that reduced the Fs from 700Hz to 550Hz:



This was when I found that if there is a gap between the vent opening and the felt that the pipe resonance will appear.

I believe that the second peak in most, if not all tweeters is some form of incomplete damping of the pole-piece pipe resonance with the exception of those with a cover over the vent opening as in the SS 9500 and some others that do restrict the opening. If there were some kind of restriction issue, the chambers would be almost ineffective. You'd be surprised how little of an opening is required.

Dave

[jkrutke]

I believe that the second peak in most, if not all tweeters is some form of incomplete damping of the pole-piece pipe resonance with the exception of those with a cover over the vent opening as in the SS 9500 and some others that do restrict the opening. If there were some kind of restriction issue, the chambers would be almost ineffective. You'd be surprised how little of an opening is required.

Damping is nothing more than flattening peaks. Yes, you can add enough damping to make a peak go away. No argument there.

Let's look at the impedance curve of a different tweeter to make a better point. This is the low end Hi-Vi TN25. It's a neo slug magnet with a very small hole leading to a large rear chamber. Since the hole through the slug is so small, the restriction to the rear chamber is very high. As such, the upper resonance based on the volume behind the dome is the higher one. No amount of damping is going to make this tweeter have a single peak any lower than 1500Hz. The best you could do is densely pack the rear chamber until it effectively has no volume and the lower peak goes away. This tweeter has no organ pipe resonances or rear chamber reflections, things that I agree do have to be resolved with damping. It's just got a 1/4" hole that is a serious restriction leading to the rear chamber.

[dlr]

Damping is nothing more than flattening peaks. Yes, you can add enough damping to make a peak go away. No argument there.

Let's look at the impedance curve of a different tweeter to make a better point. This is the low end Hi-Vi TN25. It's a neo slug magnet with a very small hole leading to a large rear chamber. Since the hole through the slug is so small, the restriction to the rear chamber is very high. As such, the upper resonance based on the volume behind the dome is the higher one. No amount of damping is going to make this tweeter have a single peak any lower than 1500Hz. The best you could do is densely pack the rear chamber until it effectively has no volume and the lower peak goes away. This tweeter has no organ pipe resonances or rear chamber reflections, things that I agree do have to be resolved with damping. It's just got a 1/4" hole that is a serious restriction leading to the rear chamber.

You can add damping that skews the shape as well. I had thought pipe resonances were the culprit for skewed responses.

Dave

Edit: After looking over some other measurements, I think I have to agree. I haven't done much work with neo tweeters, but it would explain the poor response of the MDT-44 that I have. That's the only neo I experimented with and had little success. It hadn't occurred to me to be a dual resonance. Measurements are nothing like the supplied curves, impedance and SPL. It's quite a disappointment.

I do think that improvements could be made, but it would require a dome-shaped, very effective material, like that in older non-vented designs, but then the vent would be even less effective. The MDT-44 has a large felt ring with an opening that matches the vent opening. There isn't a lot empty space, so I'm still a bit puzzled by the Fc of the resonance. But then the tweeter's T/S parameters enter into it as well.

I was recently looking at a picture of an opened Seas T27CF002. It appeared to have a felt disc rather than a ring on the vent opening. It's neo, but it must be the hexadym magnet arrangement and a larger opening (guessing here) that makes it work.

[Jonasz]

So what is the bad thing about a double peak compared to a single? How does a "double peak" tweeter sound compared to a more well designed single one?

[dlr]

So what is the bad thing about a double peak compared to a single? How does a "double peak" tweeter sound compared to a more well designed single one?

The problem is that any resonance, save the one at Fs, tends to be destructive and causes frequency response anomolies. The MDT-44 I mentioned has a really bad dip in the on-axis response related to the higher impedance peak that is significant. It necessitates either a very high crossover Fc or a trap that can require very large components. If the overall transfer function gives it a smooth response, which can be done, then the sound should be fine. It can limit the usefulness of the driver, the case I ran into with the MDT-44. That one is a real disappointment.

Dave

[Jonasz]

The problem is that any resonance, save the one at Fs, tends to be destructive and causes frequency response anomolies. The MDT-44 I mentioned has a really bad dip in the on-axis response related to the higher impedance peak that is significant. It necessitates either a very high crossover Fc or a trap that can require very large components. If the overall transfer function gives it a smooth response, which can be done, then the sound should be fine. It can limit the usefulness of the driver, the case I ran into with the MDT-44. That one is a real disappointment.

Dave

Thanks Dave for a good explanation.

[fjhuerta]

... No amount of damping is going to make this tweeter have a single peak any lower than 1500Hz. The best you could do is densely pack the rear chamber until it effectively has no volume and the lower peak goes away. This tweeter has no organ pipe resonances or rear chamber reflections, things that I agree do have to be resolved with damping. It's just got a 1/4" hole that is a serious restriction leading to the rear chamber.

https://www.htguide.com/forum/showthread.php?t=25578

Been there, done that. I wish I had kept those tweeters and see if they got any better on an HD sweep.