Purifi SPK5 Speaker Review (Prototype)

[sgoldwin]

Just to be clear, this particular speaker is not built by Purifi. It is a Purifi design, but built by hand by little ol’ me then disassembled and reassembled by @amirm after the speaker leads became disconnected during shipping. In other words, until we know different I would blame my ineptitude rather than Purifi’s design.

 

[dukanvadet]

It looks like the box has come apart on the upper left seam on the picture of the back?

 

[MZKM]

Waveguides tend to reduce distortion by loading the driver, thereby increasing its efficiency and hence reducing the voice coil displacement required to reach a given SPL.

Having said that, the kinds of shallow, wide-beamwidth, conical waveguides often used in home audio don't tend to provide very significant loading, and hence the distortion benefits are not generally very great.

But many 1” tweeters even at 1W have a lot of distortion <2kHz; so I still would imagine even if the loading is enough to increase the SPL to match it’s upper range, that it’ll have a decent amount of distortion.

However, if you have a tweeter with high base sensitivity, then I can see that happening. Many tweeters have >90dB sensitivity.

 

[andreasmaaan]

But many 1” tweeters even at 1W have a lot of distortion <2kHz; so I still would imagine even if the loading is enough to increase the SPL to match it’s upper range, that it’ll have a decent amount of distortion.

However, if you have a tweeter with high base sensitivity, then I can see that happening. Many tweeters have >90dB sensitivity.

Ah, I didn't realise when you said "What about distortion?" that you were asking a rhetorical question

I.e. I thought you were asking whether waveguides increased or decreased tweeter distortion. Hence my general answer...

 

[ROOSKIE]

Very strange choice in tweeter.
Disappointing. Maybe go with SB, Satori or heck even Dayton RS.
That Peerless Corundum 1.25" tweet could be good.
Not impressed with that woofer.
It is out of my price range anyway so now I can officially not be tempted.
It will be said again, expensive drivers can't carry a design. (& all those expensive xover components are not gonna do it either)
They are fun but really, like star athleats and film stars more trouble than not - yet they do sell the tickets. Ah marketing, marketing.
Oh well.
Thanks for the extra fun review, looking forward to more unusual reviews like this.
Purfi might wanna hire D.M.

 

[sgoldwin]

Question: How come the crossover board has 2 speakon connectors, but only 1 speakon connector is provided at the back of the cabinet for the connection?

All the speakons are 4 pole carrying separate channels for the woofer and tweeter. The crossover has an input speakon feeding the crossover network from the amplifier and an output speakon which connects to the speakon on the back of the speaker.

 

[ROOSKIE]

However, if you have a tweeter with high base sensitivity, then I can see that happening. Many tweeters have >90dB sensitivity.

Yah, that sensitivity doesnt mean they dont distort at the same SPL.
Some less sensitive designs are very robust and low in distortion at high volume and some sensitive ones cant handle it very well.
Just depends ever case is unique.

 

[Vasr]

Just to be clear, this particular speaker is not built by Purifi.

The crossover as well? Not cheap Chinese caps but expensive Jensens.

 

[sgoldwin]

The crossover as well? Not cheap Chinese caps but expensive Jensens.

The crossover is the SPK5 Crossover Kit ordered from Purifi and assembled by me.

 

[pjug]

I did not appreciate the black silkscreen/solder mask one bit. It completely covered the traces making it very hard for me to figure out what went where.

If you use a camera and near IR filter chances are that the solder mask will be transparent in the IR. I have some PCBs with black solder mask and a camera sees right though the black. Too late now I guess but maybe helpful for the future.

 

[Vini darko]

Seems like to many compromises. Erin measured a speaker (don't remember the name) recently with this woofer and it did really well.
Edit: here it is

 

[AnalogSteph]

Ignoring the bass for a moment, I am not happy to see the rising distortion in lower treble. Anyone want to guess as to the source?

Very likely the tweeter hitting its excursion limits.

Here is the impedance graph:

Our port issue/resonance is clearly apparent. Can't figure out the messiness around 30 to 40 Hz though. Anyone has a guess for that?

Either a seal issue or something with the port wobbling around. In any case there's something really amiss. The entire 10-100 Hz region looks quite different from the impedance plot in the design document, with peaks much lower than they should be. Smells like bad seal to me.

BTW, I don't see how the massive port is supposed to be supported. If it's just floating in thin air this doesn't seem like it would stand up well to drop testing.

 

[amirm]

Either a seal issue or something with the port wobbling around. In any case there's something really amiss. The entire 10-100 Hz region looks quite different from the impedance plot in the design document, with peaks much lower than they should be. Smells like bad seal to me.

This very well could be. I have told @sgoldwin to wait before sending me another unit as to collect all thoughts in this regard.

 

[Rick Sykora]

Very likely the tweeter hitting its excursion limits.

Either a seal issue or something with the port wobbling around. In any case there's something really amiss. The entire 10-100 Hz region looks quite different from the impedance plot in the design document, with peaks much lower than they should be. Smells like bad seal to me.

BTW, I don't see how the massive port is supposed to be supported. If it's just floating in thin air this doesn't seem like it would stand up well to drop testing.

I will second (or third), this is exactly the sort of bass impedance results you get when the enclosure has major leak(s).

 

[Ericglo]

Very strange choice in tweeter.
Disappointing. Maybe go with SB, Satori or heck even Dayton RS.
That Peerless Corundum 1.25" tweet could be good.
Not impressed with that woofer.
It is out of my price range anyway so now I can officially not be tempted.
It will be said again, expensive drivers can't carry a design. (& all those expensive xover components are not gonna do it either)
They are fun but really, like star athleats and film stars more trouble than not - yet they do sell the tickets. Ah marketing, marketing.
Oh well.
Thanks for the extra fun review, looking forward to more unusual reviews like this.
Purfi might wanna hire D.M.

-
As Vini noted above, Rick Craig has already done a design.

Erin tested the Purifi and it looks like going to 3.5khz is to high for a crossover.
https://www.erinsaudiocorner.com/driveunits/purifi_ptt65w04/

 

[mitchco]

@amirm very interesting. Different from the Purifi SPK4 Demo Kit I reviewed a while back. Here are some REW measurements at 30cm in my room with no gating and REW's default 500ms window:

That's 1/24 octave smoothing. Excellent response down to 32 Hz with low (relative) distortion compared to other speakers I have measured. Take the distortion measurements with a grain of salt as I don't know the distortion of the measurement mic I used, nor the mic preamp and I forget what SPL level I used (i suspect 83 dB SPL). But good enough to get a characteristic and to compare, relatively speaking, to the KEF LS50 in comparison (same conditions):

Note that the LS50 is down by 20 dB SPL at 33 Hz as compared to the Purifi.

The LS50's overall distortion, relatively speaking, higher than the Purifi's overall and especially at 33 Hz.

While I am not a fan of the AMT21CM2.1, it had a similar sonic profile to the LS50's except the bass response went down to 32 Hz, with a port tuning of 35 Hz (I think they may have lowered it to 32 Hz on my box, but I would have to dig through my correspondence with Purifi). The XO was at 2.5 kHz. The bass response was amazingly loud and clear - smoked the LS50. It also did a pretty dang good job of pressuring my room. So much so, I gathered the family in to listen. Everyone very surprised to feel the bass from such a small speaker.

I wonder what happened with this SPK5 pair?

PS. Here is a link to the SPK4 Demo Kit application note and a link to my review of Purifi PTT6.5 Woofer and 1ET400A Amplifier Technology Review.

 

[amirm]

I will second (or third), this is exactly the sort of bass impedance results you get when the enclosure has major leak(s).

What is an easy fix for this? Put in a gasket around the rabbit before screwing in the rear panel?

 

[amirm]

@amirm very interesting. Different from the Purifi SPK4 Demo Kit I reviewed a while back. Here are some REW measurements at 30cm in my room with no gating and REW's default 500ms window:

Thanks. There are some similarities though. You also show a dip around crossover. And same peaking of the tweeter.

 

[Vini darko]

What is an easy fix for this? Put in a gasket around the rabbit before screwing in the rear panel?

If you're allowed to make mess. Whitetack fixes speaker leaks like a champ.

 

[mitchco]

Thanks. There are some similarities though. You also show a dip around crossover. And same peaking of the tweeter.

Agreed, not my fav tweeter, but did not show any distortion spikes up high...

While on the topic, I wish the heck that speaker manufactures just stop with the stupid dip between 1 and 2 kHz and step rise in response above 2 kHz. I know it is a popular voicing technique, but it is not "accurate sound" This voicing destroys the balance on mixes and pushes the vocals too far forward in the mix, aside from giving it that "presence" sound. If that would stop, then I can stop eq'ing speakers above Schroeder ;-)