• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Rebuild of Purifi SPKx speakers

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
2,418
Likes
4,460
Location
Stow, Ohio USA
After helping fix the review speaker, I was thanked by receiving parts for a pair. As some of you may know, one of the woofers went bad and so I ordered two new ones. The front baffle tweeter cutouts were leaky and needed to be redone. This essentially meant removing and replacing the font baffle. Then there was the SPK5's twisty vent and external crossover. After some woodworking and modeling (sorry no pics), got to here...

IMG_0605.jpeg

The original plywood internal brace in one of the speakers was so poorly cut, I removed both and replaced with MDF. As you can see, decided to mount the crossover internally. This was done after deciding to forgo both the long SPK5 vent and/or a passive radiator. This is why this is now an SPKx rather than a SPK5. The SPK4 used a straight vent and my modeling suggested this was fine.

Here is a look at the new front baffle. It is thicker than the original by half an inch and used the extra thickness to bevel it to fit. The bevel was somewhat undersized and attached with wood glue and the other half of the joint was attached using polyurethane glue for added damping. Note my addition of cleats to the underside of the brace. This reinforces the brace as it is otherwise held in place by a thin layer of wood glue.

IMG_0606.jpeg


My paint job is multiple coats of Rustoleum satin black spray paint. It is a work in progress. Expect driver mounting and other work will scratch it up and will touch-up and polish when I am done...
 
Last edited:
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
2,418
Likes
4,460
Location
Stow, Ohio USA
Getting back to a straight vent meant flipping the back panel upside down. Like so...

IMG_0609.jpeg


This has a ramification with the crossover mounted internally as the speak-on wiring is tight against the large inductor. This will be resolved by adding some binding posts and reducing the wiring on the speak-on. More to come.

Here is the new port...
IMG_0608.jpeg


Am still playing with the tuning. The SPK4 vent was about twice as long but had nasty port resonance and a much higher f3. The f3 now is around 55 Hz but is tighter bass overall. Will share more (including measurements) next...
 
Last edited:
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
2,418
Likes
4,460
Location
Stow, Ohio USA
First will show the port modeling from Bassbox Pro. Here is the current shorter tube....

1658487525681.png


Here is the longer tube...

1658487756322.png


Visually this seems subtle but can result in a fairly significant difference in measurements. Sure one might argue the audibility, but the longer one also results in a higher f3 of about 63 Hz.

Let's look at nearfield measurement. Red trace is port and purple one is woofer.

1658495707131.png


So there is an unknown resonance around 240 Hz and you can see the one predicted by Bassbox at just over 1 kHz. Bear in mind the Bassbox model does not include the impact from the crossover.

Let's compare the shorter port and the longer port and see what we find...
1658497141540.png


Red trace is shorter and blue is the longer one. So here we see some tradeoffs. The longer port has less output but we can also see the uglier port resonance around 750 Hz.
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
2,418
Likes
4,460
Location
Stow, Ohio USA
Just for review, here is the port (aka the snake) from the SPK5…

image.jpg


The snake is a little thicker, but about 16 inches long. Not too surprisingly, has a port resonance around 400 Hz. As have mentioned before, Bassbox Pro predicts this resonance and is confirmed in Amir’r review. Along with the port resonance, it struggles to fit inside the cabinet with the crossover mounted internally. IMO, it was a step in the wrong direction for Purifi.

So, there are at least 2 questions as I ponder the tradeoff I want to make for the port. One is bass accuracy and the other is the audibility of the port resonance. Will look more closely at these next…
 
Last edited:
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
2,418
Likes
4,460
Location
Stow, Ohio USA
After comparing nearfields, modeling in VCAD and doing some listening, looks as though a 4 inch port is swinging a bit too much the other way. While Bassbox readily shows the port resonance, its graph resolutions are very limiting. The shorter port was introducing more of a boost in the lower bass than I like. There were multiple hints of this in the measurements, but after listening to some of my bass test tracks, heard more bass boom than I like…

The bass boom was a bit more noticeable in some tracks than others, but my test rig is in the middle of the room and pretty sure it would get worse closer to the walls. There is another factor as the crossover was designed the long snake port too. With the snake though, Amir complained the bass was weak. So, maybe the SPK4 had it right, am going to keep trying and also see if I can better tame the 240 Hz resonance too.

In VCAD, I have also started to integrate more of the modeling. This gives a better look with the speaker as a system as can combine cabinet tuning with baffle step and crossover impacts. Am trying to avoid modding the crossover if possible. Here is a look at the model so far…
1658678625624.png
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
2,418
Likes
4,460
Location
Stow, Ohio USA
Due to Covid we rescheduled one of our 2021 Christmas parties to this July. So have been busy with preparing for the party. Just before the party, I had installed the 6 inch port and started to review the speaker build. The impedance curve showed a small resonance around 224 Hz. As I had taken down my speaker measurement rig for the party, I just managed to get some of it set back up. Along with the resonance, I found some cabinet leaks! After I fixed those, I got the speaker on test stand and tried listening for the resonance. What I heard was very subtle.

Since Sunday I have been working to alleviate the resonance. It appears to be related to the removable back baffle. I added some additional gasket tape and it has almost entirely eliminated the resonance. At this point, I wanted to do some listening and liked what I heard. The bass was much more solid than the original SPK5 tuning. Will share some measurements in the next post to show the changes visually. :)
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
2,418
Likes
4,460
Location
Stow, Ohio USA
First a look at the change in bass frequency response due to the new tuning....

1659526358635.png

The red trace is the original SPK5 snake port and the green is the new 6 inch straight port. The new tuning should resolve Amir's complaint of lack of bass as it is now flatter down to mid 40s and about 4 dB higher at 40 Hz. The new tuning also reduced the port resonance. Neglecting the pipe resonance, could see why the lower box tuning and shallower rolloff of the SPK5 might be desirable for an application where the speaker would be positioned closer to the back wall.
 

fpitas

Major Contributor
Forum Donor
Joined
Jul 7, 2022
Messages
1,937
Likes
2,173
Location
Northern Virginia, USA
Do you think the angled (vs sweeping curve) corners on the snake port are responsible for the rolloff?
 

voodooless

Master Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
5,697
Likes
9,379
Location
Netherlands
First a look at the change in bass frequency response due to the new tuning....

View attachment 222070
The red trace is the original SPK5 snake port and the green is the new 6 inch straight port. The new tuning should resolve Amir's complaint of lack of bass as it is now flatter down to mid 40s and about 4 dB higher at 40 Hz. The new tuning also reduced the port resonance. Neglecting the pipe resonance, could see why the lower box tuning and shallower rolloff of the SPK5 might be desirable for an application where the speaker would be positioned closer to the back wall.
I guess this is always a trade-off with the use case in mind. If you were to EQ the speaker, the lower tuning will give it more extension and possibly a bit better power handling down low. The f10 difference is very significant. Port resonances can be fixed in other ways as well if needed.
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
2,418
Likes
4,460
Location
Stow, Ohio USA
Here is a look at the (still not final) impedance curve...

1659527822100.png

The box tuning is just over 40 Hz and you can see evidence of the resonance issue I mentioned by the slight jag earlier right around 240 Hz.
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
6,615
Likes
6,710
Location
Brussels, Belgium
I guess this is always a trade-off with the use case in mind. If you were to EQ the speaker, the lower tuning will give it more extension and possibly a bit better power handling down low. The f10 difference is very significant. Port resonances can be fixed in other ways as well if needed.
EQ won’t increase max output. Furthermore, the responses meet at 35Hz so I doubt whether the F10 is 33Hz or 27Hz would make a difference. We’re already past the point of meaningful content.
 

voodooless

Master Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
5,697
Likes
9,379
Location
Netherlands
EQ won’t increase max output.
I never said it would. The lower tuning though may give it just a bit extra down low, depending on use case that may be relevant. You’ll trade-in a bit above the tuning usually, but that’s just a few dB, and the extra extension may be of relevance. Like I said, it will depend on the intended use case and setup. Personally I would probably go for something in between these two options.
Furthermore, the responses meet at 35Hz so I doubt whether the F10 is 33Hz or 27Hz would make a difference.
At 30 Hz the difference is already 5 dB. In-room it will definitely make a difference.
We’re already past the point of meaningful content.
That depends on the use case, but in many cases, that may very well be.
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
2,418
Likes
4,460
Location
Stow, Ohio USA
Do you think the angled (vs sweeping curve) corners on the snake port are responsible for the rolloff?
As this is a modeling rather than a measurement, the rolloff is purely due to the difference in port length.
 
Last edited:
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
2,418
Likes
4,460
Location
Stow, Ohio USA
I guess this is always a trade-off with the use case in mind. If you were to EQ the speaker, the lower tuning will give it more extension and possibly a bit better power handling down low. The f10 difference is very significant. Port resonances can be fixed in other ways as well if needed.
Agree, but I tend to use away from walls and only eq for room if at all. The new tuning is very comparable to SPK4 so gather Purifi considered comparable tradeoffs.

I will explore the audibility of the port resonance a bit later, but due to other project demands, do not plan to spend much more time on these speakers.
 

alex-z

Addicted to Fun and Learning
Joined
Feb 19, 2021
Messages
632
Likes
1,144
Location
Canada
Could you close mic the port and try stuffing it with polyfill? In theory 6" of low density material should cut down on the unwanted higher frequency resonances without compromising the 40Hz port output. Although the density difference might change the port tune.
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
2,418
Likes
4,460
Location
Stow, Ohio USA
Just tried some nearfield measures and did not like what I saw. However, am working with only some polyfill behind the woofer. The Purifi plans call for copious quantities (300 grams) of cotton damping. I have some from the original speakers, so added 100 grams to the back chambers. 300 grams would mean some really dense packing!

Have just about eliminated the 240 Hz resonance. Appears that was mainly the large PCB for the crossover. I usually mount it with generous application if butyl rubber, but felt some flex along the long edges. Also added more to the large ribbon inductor as it was looser than I like. Back to the speaker stand with an adjustable port to take some more measurements…
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
2,418
Likes
4,460
Location
Stow, Ohio USA
...and here is after more butyl rubber and adding 100 grams of cotton damping to upper chamber:

1659574272989.png


The resonance blip is less pronounced, and the overall traces are smoother too. This may be partly due to some added soldering. As I listened for the resonance, had jiggled a couple of caps and found they were not making solid contact with the board.
 
Last edited:
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
2,418
Likes
4,460
Location
Stow, Ohio USA
Decided that having the other speaker available for comparison would be useful, so have been putting it together today. Did a quick impedance check without all the damping and gasketing in place. The same 240 Hz resonance is still present and tracks the first one closely considering the less completed state. I suspect the resonance is related to the woofer/cabinet interaction but was not present in the original SPK5 implementation. The main difference is I moved from bolts to screws to mount the woofer. It is still very obvious in nearfield measurements so have so more work to do as yet!
 
Last edited:
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
2,418
Likes
4,460
Location
Stow, Ohio USA
Gasketing is in place on the second speaker for the most part and am using it in my Understanding Speaker Impedance thread. The joint seals are much better but have not sealed the port in place. This allows me to readily change the vent length or seal the vent entirely. Once I have the same size port in both boxes and comparable damping, will post comparison measurements.
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
2,418
Likes
4,460
Location
Stow, Ohio USA
Just when I thought I was done with this resonance issue, I realized I still had work to do (as shown above). Note that while negligible on impedance measure, was clearly prominent on port nearfield. Although the front baffle was thicker than the original, I swapped the bolts for screws. I now have a resonance around 240 Hz which was clearly not a pipe resonance. This meant some other part of the speaker. I tried various internal damping and securing crossover parts without major improvement. In another thread, had tried putting a rubber ball between the brace and the woofer magnet and saw some change. So started experimenting with other materials like wood and butyl and got noticeable results with butyl rubber.

Here is a look through the port opening on the back baffle..,

1661383105296.png

The blue is a baseline measure. The green is with about golf ball wad of butyl rubber between the woofer magnet and the brace. This had a greater effect than a wood shim I had tried. The red trace is when I added the 2x6 inch port back into the baffle. You can see why this sort of activity has tradeoffs. While the butyl rubber helped reduce the 240 Hz resonance, it increased around 600 Hz and around 1500 Hz. With port in place, you can see some of the higher frequency resonances are much worse. So the butyl rubber is taking some of the woofer chassis energy and distributing it into the cabinet. Am going to listen to the result but does not appear to be the improvement I sought.
 
Last edited:
Top Bottom