Author Topic: EEVblog 1514 - Sonos: The Juicero of Wifi Speakers? (TEARDOWN)  (Read 2584 times)

0 Members and 1 Guest are viewing this topic.

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37742
  • Country: au
    • EEVblog
EEVblog 1514 - Sonos: The Juicero of Wifi Speakers? (TEARDOWN)
« on: November 14, 2022, 07:49:41 am »
Dumpster teardown of a Sonos Play 5 Gen 1 Wifi speaker system.
Is it the Juicero of wifi speakers?

 

Offline SiliconWizard

  • Super Contributor
  • ***
  • Posts: 14488
  • Country: fr
Re: EEVblog 1514 - Sonos: The Juicero of Wifi Speakers? (TEARDOWN)
« Reply #1 on: November 14, 2022, 10:39:34 pm »
Fascinating internal "trumpets". :-DD

What I'd be curious about would be to see independent testing of this - actual performance (frequency response @ 1m including distortion figures), and not just what is claimed by the vendor.

 

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37742
  • Country: au
    • EEVblog
Re: EEVblog 1514 - Sonos: The Juicero of Wifi Speakers? (TEARDOWN)
« Reply #2 on: November 14, 2022, 11:02:41 pm »
What I'd be curious about would be to see independent testing of this - actual performance (frequency response @ 1m including distortion figures), and not just what is claimed by the vendor.

"Performance" is subjective. They can just say any result is deliberate.
Dr Dre made $3BN by selling Beats, which is nothing more than his signature algorithm "sound". Put it on measurement gear and you'd probably say it's crap.
 

Offline SiliconWizard

  • Super Contributor
  • ***
  • Posts: 14488
  • Country: fr
Re: EEVblog 1514 - Sonos: The Juicero of Wifi Speakers? (TEARDOWN)
« Reply #3 on: November 15, 2022, 02:27:27 am »
What I'd be curious about would be to see independent testing of this - actual performance (frequency response @ 1m including distortion figures), and not just what is claimed by the vendor.

"Performance" is subjective.

With audio and to some degree, that is true. Hence the heavy debates constantly occuring about audio, and hence, at its extreme, the audiophool movement.
Which is why I mentioned objective measurements.

They can just say any result is deliberate.

Well, as long as it remotely makes sense, sure. Heck, if for instance it had bad distortion figures, but this "matched" the typical kind of distortion you get with tube amplifiers, why not. There are amateurs of this kind of distortion. But they are not typical buyers for this kind of gear... and the typical distortion you're bound to get with a less-than-ideal class-D (assuming it could be a mediocre implementation, in spite of the apparent "overengineering") is uh, not that pleasing.

But I think they are showing the frequency response in their manual, and probably some distortion figures? So one could at least check that those are not plain lies.

Apart from the amplifier, and despite the trumpets, I'm a bit wary of any stereo speakers inside the same enclosure. It's very hard to get anything close to what you can get with separate enclosures.

Dr Dre made $3BN by selling Beats, which is nothing more than his signature algorithm "sound". Put it on measurement gear and you'd probably say it's crap.

It may very well be indeed. Or not, that would have to be checked. The only thing here is: if whatever figures officially claimed by the vendor do not match reality, then those are just lies. Now of course, any abstract statement of "quality" doesn't count. It's very much exactly as with audiophool gear. If they claim their device has, say, -120 dB THD+N over 10 Hz to 20 kHz, while the reality is -90 dB, they are fricking lying. But if they claim their device sounds "warm with pristine highs", then sure, you can hardly beat that with any measurement.

I think we'll agree that popularity is not a sign of quality.
 

Offline coppercone2

  • Super Contributor
  • ***
  • Posts: 9466
  • Country: us
  • $
Re: EEVblog 1514 - Sonos: The Juicero of Wifi Speakers? (TEARDOWN)
« Reply #4 on: November 15, 2022, 04:30:56 am »
I think that we will be seeing more threaded inserts and stuff because of automation, I think the price is going down because of things like automatic screw guns etc, which is nice, because working on tons of press fit crap is annoying. There is probobly some new cheap thread insert installer machine in the plastic factory in China. That capacitor bank looks annoying too, they should put 4 x more screw and a part over that.

Check out the setups they have now for factory screw installation, you are talking mini-bolt guns that have feeds like chainguns.. I think I even saw one with a automatic screw changer so you can just automatically send the correct screw and bit into the driver. Or just basic ones that work like coin sorters and drop you off a screw at head height (foot pedal control?).. all those actuators in Shenzen are being used for something

Would not suprise me if there are situations where the mold complexity is more expensive then some kind of insert-machine and screw gun. Then they have freedom to change whatever at a lower cost since its like putting parts on a PCB not re-machining complex stuff. The amount of back and forth between mechanical/electrical goes down. Robot arm + heat cartridge vs machining a complex mold that has to match the PCB within like 0.01 inches to be 'good'. How much do people love a cosmetic redesign without the engineering effort? (that does factor into bottom line now)



And there is probobly a army of mini CNC lathes making screws cheap cheap cheap now.

I don't think screws will be a good sign of quality for too long, specially with developments in CNC grinders and abrasive science to make more stable abrasives (cubic boron nitride)

What is gonna determine quality is how that stands the test of time, if modularity tricks were used.

« Last Edit: November 15, 2022, 04:48:18 am by coppercone2 »
 
The following users thanked this post: EEVblog

Offline coppercone2

  • Super Contributor
  • ***
  • Posts: 9466
  • Country: us
  • $
Re: EEVblog 1514 - Sonos: The Juicero of Wifi Speakers? (TEARDOWN)
« Reply #5 on: November 15, 2022, 04:42:03 am »




and those are primitive compared to what I saw once on youtube (house in some generic 'mail your parts and we assemble' place in china.
« Last Edit: November 15, 2022, 04:44:15 am by coppercone2 »
 

Offline Dundarave

  • Regular Contributor
  • *
  • Posts: 152
  • Country: ca
Re: EEVblog 1514 - Sonos: The Juicero of Wifi Speakers? (TEARDOWN)
« Reply #6 on: November 15, 2022, 05:13:15 am »
I’ve had an identical pair of these for over 10 years now, and they still impress the hell out of me with the sound quality, power, and ease of wirelessly connecting Apple Music, internet radio, etc. 

As Dave demonstrates, the build quality is very high, and the fact that they’re still doing daily what I consider an excellent job after 10 years puts them in a far different class than a “Juicero”.  Not having to have a separate amp, visible speaker wires all over the place, and hacks of all kinds to connect to cloud music have made me a repeat buyer x 6 of the smaller versions (for other rooms at home).  I think these things are great!
 

Offline bob808

  • Frequent Contributor
  • **
  • Posts: 281
  • Country: 00
Re: EEVblog 1514 - Sonos: The Juicero of Wifi Speakers? (TEARDOWN)
« Reply #7 on: November 15, 2022, 07:29:58 am »
All that and I still noticed a couple of NE5532. I guess that says a lot about the mighty NE5532.
 

Offline 2N3055

  • Super Contributor
  • ***
  • Posts: 6663
  • Country: hr
Re: EEVblog 1514 - Sonos: The Juicero of Wifi Speakers? (TEARDOWN)
« Reply #8 on: November 15, 2022, 08:00:43 am »
"Trumpets" are not trumpets. Venting and bass reflex tubes need to have gradual transition to have less turbulence at the edges. It lowers low frequency distortions.. There are many bass reflex vents on speakers that have same shape. Except you don't see inner part sticking in the air usually...
These do look funny, but it is normal thing to do.
 

Offline 2N3055

  • Super Contributor
  • ***
  • Posts: 6663
  • Country: hr
Re: EEVblog 1514 - Sonos: The Juicero of Wifi Speakers? (TEARDOWN)
« Reply #9 on: November 15, 2022, 08:19:41 am »
All that and I still noticed a couple of NE5532. I guess that says a lot about the mighty NE5532.
5532 was designed specifically for pro audio. And done right. There is this misconception in the world (fueled by digital craze) that everything (including mechanical engineering and analog electronics) will follow Moore's proposition (because that is what it is, a proposition, not a law) and be "better" every day...
 

Offline bob808

  • Frequent Contributor
  • **
  • Posts: 281
  • Country: 00
Re: EEVblog 1514 - Sonos: The Juicero of Wifi Speakers? (TEARDOWN)
« Reply #10 on: November 15, 2022, 08:58:06 am »
There might be certain applications which might call for better spec opamps but usually NE5532 is good enough for audio. In this case most likely used as a buffer for the two inputs, whatever comes from the wifi and from the AUX connection on the back.
 

Offline 2N3055

  • Super Contributor
  • ***
  • Posts: 6663
  • Country: hr
Re: EEVblog 1514 - Sonos: The Juicero of Wifi Speakers? (TEARDOWN)
« Reply #11 on: November 15, 2022, 09:56:59 am »
There might be certain applications which might call for better spec opamps but usually NE5532 is good enough for audio. In this case most likely used as a buffer for the two inputs, whatever comes from the wifi and from the AUX connection on the back.
There are actually not many opamps that are better for audio...  At 50x the cost.
5532 is bipolar, so for high impedance inputs jfet based opamps will have less current noise..
And that is basically it.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16621
  • Country: us
  • DavidH
Re: EEVblog 1514 - Sonos: The Juicero of Wifi Speakers? (TEARDOWN)
« Reply #12 on: November 18, 2022, 06:30:42 pm »
Dave, since the enclosure is part of the bass speaker suspension, there is a lot of vibration and stress applied to the boards and parts which explains why it is mechanically overbuilt.  Anything less would rattle and fall apart.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16621
  • Country: us
  • DavidH
Re: EEVblog 1514 - Sonos: The Juicero of Wifi Speakers? (TEARDOWN)
« Reply #13 on: November 18, 2022, 06:35:04 pm »
There might be certain applications which might call for better spec opamps but usually NE5532 is good enough for audio. In this case most likely used as a buffer for the two inputs, whatever comes from the wifi and from the AUX connection on the back.

There are actually not many opamps that are better for audio...  At 50x the cost.
5532 is bipolar, so for high impedance inputs jfet based opamps will have less current noise..
And that is basically it.

JFETs have less current noise but suffer from non-linear input impedance resulting in significant distortion when driven from high input impedance, even 10 kilohms.  There are some specialty JFET input parts now which address this.

There are all kinds of good audio operational amplifier now, including the 5532.
 
The following users thanked this post: 2N3055

Offline thm_w

  • Super Contributor
  • ***
  • Posts: 6389
  • Country: ca
  • Non-expert
Re: EEVblog 1514 - Sonos: The Juicero of Wifi Speakers? (TEARDOWN)
« Reply #14 on: November 18, 2022, 09:47:23 pm »
Dave, since the enclosure is part of the bass speaker suspension, there is a lot of vibration and stress applied to the boards and parts which explains why it is mechanically overbuilt.  Anything less would rattle and fall apart.

Its a bit of a stretch to say anything less would fall apart, although they probably designed for the worst case scenario (max volume) with safety factor added.
Some of their later units are not built like this. 78 fasteners is crazy.

https://www.ifixit.com/Teardown/Sonos+Play:3+Teardown/12475
https://fccid.io/SBVRM010/Internal-Photos/Internal-Photos-Additional-2723871
Profile -> Modify profile -> Look and Layout ->  Don't show users' signatures
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf