Well, as an end user it actually gives us important information about the problem show in the graphs I just posted earlier.
So a change in the height of the impedance peak, will say something about how linear the BL is.
A change in Fs will tell us something about the change in Cms.
A smaller change in both parameters will always translate in better performance.
Which can either be cross-corelated with the distortion measurements.
To quote Klippel:
But most importantly will help picking the right driver for a certain cabinet.
That last part can't be seen from just distortion graphs.
edit:
Fyi, note that Klippel here uses "Re = electrical voice coil resistance at DC"
So the confusion circle is complete again.
Last edited:
With wool we have a similar non-linear behavior.
So wool acts differently at different SPL's
The excursion totally depends per driver, depending on size and motor structure.
If you just want to measure the amount of wool and the increase of volume, this can be easily be done by measuring Qms, Qes, Qts and Fs. You don't even need to go into full TS parameters for that.
The influence of the amount of wool in given volume and driver like you mention was basically what we did in late 70-ties and my self last time in 2005.
It is not only the amount but also the structure of the wool and the fit in the cabinet, a bit of a balancing act to get to a Qts of ~ 0.7 and a damping of the cabinet resonances.
The wool structure is important, F.i. carding the wool gave audibly a very poor cabinet resonance damping, leaving holes in the filling also but to a lesser extent, combining with f.i. BAF was the worst.
Today one can measure a lot more in a fraction of the time needed then. So that is my aim next to listening of course.
In 2005 i did use ArtaLimp for testing which is small signal (100mV or so). Now 19 years later i am surprised that the Fs and the Qts of the woofers and midrange in my system basically has not changed.
When the testcabinet is made for the Purifi PTT8.0X08-NAB02 i want to do more comprehensive measurements while stepwise increasing the amount of wool, not only small signal, also large(r) signal .
When done i do the same for the PTT4.0M04-NAC04. The challenge here is the small cabinet volume, thus also small amount of wool.
Interesting times ;-)
It is not only the amount but also the structure of the wool and the fit in the cabinet, a bit of a balancing act to get to a Qts of ~ 0.7 and a damping of the cabinet resonances.
The wool structure is important, F.i. carding the wool gave audibly a very poor cabinet resonance damping, leaving holes in the filling also but to a lesser extent, combining with f.i. BAF was the worst.
Today one can measure a lot more in a fraction of the time needed then. So that is my aim next to listening of course.
In 2005 i did use ArtaLimp for testing which is small signal (100mV or so). Now 19 years later i am surprised that the Fs and the Qts of the woofers and midrange in my system basically has not changed.
When the testcabinet is made for the Purifi PTT8.0X08-NAB02 i want to do more comprehensive measurements while stepwise increasing the amount of wool, not only small signal, also large(r) signal .
When done i do the same for the PTT4.0M04-NAC04. The challenge here is the small cabinet volume, thus also small amount of wool.
Interesting times ;-)
Luckily we still have HifiCompass to help us out here!💙💛
Let’s donate if we are able and appreciate the work at Yevgeniy is doing!
I agree it should be up to transducer manufacturer to supply Kms(x), BL(x) and Le(x) graphs. If not by themselves, then send it to some line who has a Klippel KA3 eg. @bikinpunk or Warkwyn etc.
End user/designer starts off with TSP/small signal models to design prototype box but needs to verify and refine at both low and high voltage testing.
The variation in Kms, BL and Le change in free air for sure, but what about once driver is in box? As voltage drive level changes so do these 3 parameters affecting box tuning to some degree. But for small drive units, at higher drive levels SPL compression effects seems to occur much more so than cabinet tuning differences.
We haven’t even got to talking about the room but that’s another matter.
In practice I test at .283 V, 2.83V, 8.95 and 28.V (0.1/1/10/100W) depending on driver (lower voltage on high sensitivity tweeter or midranges and high voltages on big woofers/subwoofers) and take distortion measurements and that tells me a lot about what I’m doing (or not doing) with the integrity of my enclosure as well as well as driver limitations.
Last edited:
The whole manufacturer and who measures what argument 100% beats me? How is that any different than distortion measurements?
I am just saying that if you measure impedance anyway, you might as well measure it on multiple voltages.
Right because it WILL give you more information to build speakers!!!
I am just saying that if you measure impedance anyway, you might as well measure it on multiple voltages.
Right because it WILL give you more information to build speakers!!!
I’m not opposed to it, I just think it is a lot of work for a little gain.
Is there a way to do fast impedance sweeps at different voltages quickly?
(like the way one does with distortion measurements at different voltages?)
I use the 9 yo DATs V2 software on my 16 year old WT3 for impedance sweeps; so I may be out of date. Or not.
Is there a way to do fast impedance sweeps at different voltages quickly?
(like the way one does with distortion measurements at different voltages?)
I use the 9 yo DATs V2 software on my 16 year old WT3 for impedance sweeps; so I may be out of date. Or not.
Last edited:
This woofer is on my radar. Did you measure the parameters or use them from the FAitalPro datasheet?
Important is that it only can be done with a constant voltage source or maybe with a small reference resistor.
The series reference resistor method doesn't work.
It all depends how many voltage steps you want to make obviously.
But at a minimum it takes only one more measurement (at maximum/highest voltage) , since a low voltage impedance measurement is already being done.
For the piston part (low frequency) of a loudspeaker design this information is not of little gain, it's essential data to predict how good (or bad) it will sound.
The technical data are taken from the datasheet. The measured Fs and Qts you can see in the impedance measurements
There is a good read on this on Linkwitz's website.
I don't know on the top of my head anymore unfortunately, so I am gonna be very lame by saying use your favorite search engine, lol
All i could find and already knew was this:
https://www.linkwitzlab.com/Stereophile interviews Siegfried Linkwitz.htm
It was the non-linear part of your remark i am most interested in.
https://www.linkwitzlab.com/Stereophile interviews Siegfried Linkwitz.htm
It was the non-linear part of your remark i am most interested in.
@JanRSmit Yeah, that's not the one I am thinking of. I really have to search for it, sorry.
It has been a while, but in a very short and rough summary, it has to do with how soft damping materials work on a thermodynamic level. Which creates some non-linearities on how it reacts to sound pressure waves.
Probably has something to do with the heating effect because of the absorption.
Worst case I have to dig into my old thermodynamic physics books, but those are somewhere stored far away in a box
This has been a well discussed subject here on Diyaudio, at least in the past (> 10 years ago).
(so maybe that is a good start)
So sometimes I fall into the big mistake that certain things have become common knowledge by now, lol
Sorry.
fyi, if I remember well, this effect is mostly an issue at lower frequencies, when there is sufficient cone excursion.
Maybe that part was a bit unclear.
It has been a while, but in a very short and rough summary, it has to do with how soft damping materials work on a thermodynamic level. Which creates some non-linearities on how it reacts to sound pressure waves.
Probably has something to do with the heating effect because of the absorption.
Worst case I have to dig into my old thermodynamic physics books, but those are somewhere stored far away in a box
This has been a well discussed subject here on Diyaudio, at least in the past (> 10 years ago).
(so maybe that is a good start)
So sometimes I fall into the big mistake that certain things have become common knowledge by now, lol
Sorry.
fyi, if I remember well, this effect is mostly an issue at lower frequencies, when there is sufficient cone excursion.
Maybe that part was a bit unclear.
Last edited:
In another overview page Linkwitz mentions natural wool not being frequency independent etcetera. But no study reporting found at all.
In the seventies a series of tests were done, and when i joined redone.
long haired "bailey" was then popular in transmission line boxes. In sealed boxes it was very poor. Mixing with BAF also very poor.
The wool we used and i still use is selected on a particular capability. Being that the core wool fibre( thorougly cleaned) reacted on sound pressure variation, thus absorbing/releasing energy. Worked well in the low frequenties.
If memory server me well it was wool coming from New Zealand.
The other aspect is short haired and very dense pressed in to blocks of several hunderd kilo's. What we did is pluck the wool in small plucks of wool of different sizes.
Due to these plucks one gets a filling with sort of random varying fill density.
This also helps in reducing/damping cabinet resonances, as it does prevent a straight "path" needed for a resonance.
All in all very audible.
And yes too much wool does not sound good ;-)
But as i did not find any study applicable, i want to do some extended measurements with the Purifi 8" woofer. The woofer i did test in 2005 is the scanspeak ss8555-01 with its moving mass increased. Not in the same quality league as the Purifi woofer, nor do i have my full set of notes of that excercise anymore, other than some loose note sheets which by accident were not thrown away.
In the seventies a series of tests were done, and when i joined redone.
long haired "bailey" was then popular in transmission line boxes. In sealed boxes it was very poor. Mixing with BAF also very poor.
The wool we used and i still use is selected on a particular capability. Being that the core wool fibre( thorougly cleaned) reacted on sound pressure variation, thus absorbing/releasing energy. Worked well in the low frequenties.
If memory server me well it was wool coming from New Zealand.
The other aspect is short haired and very dense pressed in to blocks of several hunderd kilo's. What we did is pluck the wool in small plucks of wool of different sizes.
Due to these plucks one gets a filling with sort of random varying fill density.
This also helps in reducing/damping cabinet resonances, as it does prevent a straight "path" needed for a resonance.
All in all very audible.
And yes too much wool does not sound good ;-)
But as i did not find any study applicable, i want to do some extended measurements with the Purifi 8" woofer. The woofer i did test in 2005 is the scanspeak ss8555-01 with its moving mass increased. Not in the same quality league as the Purifi woofer, nor do i have my full set of notes of that excercise anymore, other than some loose note sheets which by accident were not thrown away.
The measurements of the Scan-Speak 18W/8741T00 woofer have been added:
https://hificompass.com/en/speakers/measurements/scan-speak/scan-speak-18wu/8741t00
https://hificompass.com/en/speakers/measurements/scan-speak/scan-speak-18wu/8741t00
You often see that with a low loss surround - it doesn't dampen surround resonances completely. This is sometimes less obvious when listening (the small dip in frequency response doesn't have a big influence).
But here you also have a widened & irregular frequency response off axis ... not happy with that.
But here you also have a widened & irregular frequency response off axis ... not happy with that.