View attachment 41View attachment 42
View attachment 43View attachment 44View attachment 45
And these are distortion responses of 2 samples.
It was measured at nearfield of PR.
I know you said that the distortion is not caused by PR,
it maybe just leaked through PR?
but why they are different? Do we have to assume that level of difference is normal for Sointuva WG? Both of them were measured before disassembly.
And, Lars Risbo said that around 400Hz distortion is caused by magnet-frame resonance.
But why they aren’t showed in many of driver measurement of Purifi’s?
And it should be clear what number of order of harmonics should be occured by resonance and the level too.
In Erin's review, 3rd HD is dominant. But the 2 sample showed that 2nd HD is dominant. (also 3rd exist.)
Would you please let us know about more specific information of Purifi's magnet - frame resonance?
Well, as Lars points out in his comments above, this is an inherent problem of all speaker drivers to a lesser or greater degree. The magnet/motor and frame create a mass/spring oscillator.
The mounting of the driver directly affect the way the mass / spring behaves and how it damps the oscillation.
Its emphasised in the Purifi because of its very high MMS and strong motor. It also sticks out because its a generally very low distortion driver.
You will note the reduced level, but more spread out response of sample B. This is due to a more compliant interface between the driver frame and the cabinet. Sample A is less compliant and stiffer, hence higher "Q" response. They both left here very similar, but something obviously changed with sample A. With what I have witnessed of the reviewers behaviour and his "fiddling/tweaking/modifications" , I am not inclined to believe him that nothing was touched.
Ultimately you get to a situation like in the ASR directiva. If its a very tight stiff interface the response gets extremely pronounced and sharp. High Q.
You can also see it in Amirs test of the SPK5
You can see below how he misinterprets this as a port resonance when performing near field measurements
So, yes the vibration created by the resonance is going to couple into the cabinet and be visible through ports or passive radiators.
As to why this doesnt show in the Purifi driver data, I simply suspect its due to the type of baffle they tested it on and how it was mounted. I dont think this is misleading, as Im am confident you will find many/most drivers show additional resonance when actually mounted in real world speaker cabinets as opposed to their test data.
I have personally discussed all of this with Lars, Bruno and Claus. Its not a huge problem when as a designer you understand the mechanisms, you can mitigate it down to very low levels, and we have.
Lets get real about this. The distortion the reviewer found is only at about
0.6% !!!!!!! Thats inaudible in music, and even with a sine tone difficult.
We dont hide the issue, its shown in the measurements on the product page We show -46db / 0.5%
However we agreed with Purifi a better way to mitigate this issue. It can be improved further, but Im not going to discuss how
I am sure you can understand why.