Can Purifi's SPK5 be improved with a passive radiator?

[Rick Sykora]

In several instances on ASR, it has been stated that the Purifi's new 6 inch woofer is best with a passive radiator. They may answer this soon as I understand SPK6 will use its passive radiator (pr). Rick Craig used the SB 5x8 pr with the Purifi woofer in his Purezza. He also used it in his Integrity. Since I built the Integrity, I have the same SB pr. Now that I have a SPK5, it was not all that hard to create a new back baffle with the SB pr instead of the S-shaped port. It was a bit challenging due to the pr height, but moved the speakon towards the corner to fit. Also, note that I had removed the SPK5 brace (as cabinet is already solid, did not find any indication that altered results at middling SPLs used here.)

While have been building subwoofers for about 3 decades, had never designed one with a pr. Since I use Bassbox Pro and it can model pr alignments, thought I would start with modeling the SPK5 with the SB pr. The vented SPK5 has an f3 of just below 50 64 Hz, so was pleasantly surprised to find the pr model f3 was lowered to 40 48 Hz. Along with no major indication of the port resonance exhibited by the vented SPK5 model, the initial checks looked promising. Here are my baseline SPK5 measurements with REW...

SPK5 frequency response (1 meter, 75db, 3ms gate)

Ignoring the low bass (gated out), this aligns well with other public frequency response measures of the SPK5.

SPK5 woofer nearfield (0m, output level maintained, no gating, 1/24 smoothing)

Using 105 dB line as reference, 3 dB down is 50 Hz See later post, but the updated f3 is 64 Hz. So, seems to match well with Bassbox modeling...


SPK5 port (0m, output level maintained, no gating, 1/24 smoothing)

Note the prominent port resonances at 400 Hz, 800 Hz and 1.2 kHz (all projected in the Bassbox model). While not shown, even without gating, there is no obvious sign of the 400 Hz port resonance affecting the on-axis response. However, given that the port is on back baffle and room modes likely dominate, this is not a definitive measure. Will have to wait for Amir to remeasure to see what Klippel NFS finds.

 

[McFly]

What is the port resonance's (400hz) phase in relation to the woofer? Still a full 180degree?

 

[Rick Sykora]

What is the port resonance's (400hz) phase in relation to the woofer? Still a full 180degree?

In relation to the woofer, the port output is in phase.

 

[Matias]

In my ignorant opinion, since response does not change much, but port noise is eliminated by a passive radiator, why not? Price maybe, but other than that, I only see benefits.

 

[Rick Sykora]

In my ignorant opinion, since response does not change much, but port noise is eliminated by a passive radiator, why not? Price maybe, but other than that, I only see benefits.

Cost can be a factor. Notably, the Purifi PR is not inexpensive. Will get into this more after I share the results after mods. With a correctly designed rear-facing port, I have not heard any major noise issues. So am not sure I would mod just to eliminate the port resonances, but I was attracted to the potential bass extension of the pr...

 

[Rick Sykora]

Here are my first pass results with the SB passive radiator...

SPK5 frequency response (1 meter, 75db, 3ms gate)

...same as vented, so will save the server storage

SPK5 woofer nearfield (0m, output level maintained, no gating, 1/24 smoothing)

Despite the bass extension I expected from modeling, did not materialize. Almost the same as with the port, but with the 5th order rolloff. TBH, did not achieve the modeled tuning, but tried adding some more damping and had little effect. On the other hand, the pr resonances are quite different...

SPK5 pr nearfield (0m, output level maintained, no gating, 1/24 smoothing)

Maybe a bit more interesting compared directly with the port...

So we see the pr (red trace) has some resonance around 200 Hz and 700 Hz, but knocked down the higher Q ones at 400 Hz and 1.2 kHz. Circling back to the question about whether I would use the pr, seems I have some more work to do. TBH, could not hear any difference in bass, but like eliminating that 1.2 kHz resonance from the vent. So, for now am going with the pr, but still plan to work on it.

For the SPK5 owner, the Purifi pr would likely be a better fit. Going to get some and try. For others, it is a small footprint and may have some other advantages too. While waiting for those to arrive, want to see if I can get that pr bass extension and maybe try the active crossover too!

 

[Rick Sykora]

Amir's latest measurements made me revisit the bass performance of the SPK5 vented design. Upon review of the Bassbox model, I found I had entered a port diameter of 3 inches. The real port is closer to 2 inches, this increased the projected to 64 Hz.

p.s. Also found the pr modeled box had 2 passive radiators. Corrected the f3 to 48 Hz based on 1 pr in design. Now with bass extension comparable to the vented SPK5, only advantage was reduction in port resonance. Explains my measurements now too!

 

[tktran303]

Here are my first pass results with the SB passive radiator...

SPK5 frequency response (1 meter, 75db, 3ms gate)

...same as vented, so will save the server storage

Hi Rick,

I wonder what’s going on here.

The dip in woofer response should correspond very closely to the impedance dip, so it should be at 40Hz +/- 1Hz.

 

[tktran303]

Circling back to the question about whether I would use the pr, seems I have some more work to do. TBH, could not hear any difference in bass, but like eliminating that 1.2 kHz resonance from the vent. So, for now am going with the pr, but still plan to work on it.

Can you take an ground plane measurement to see the differences?

 

[Rick Sykora]

Hi Rick,

I wonder what’s going on here.

The dip in woofer response should correspond very closely to the impedance dip, so it should be at 40Hz +/- 1Hz.

you might have missed that the build did not achieve the targeted box tuning that was modeled for the pr. Am investigating, but have had some other distractions.

 

[Rick Sykora]

Can you take an ground plane measurement to see the differences?

Not inside, and outside is 13 degrees and ground is snow covered.

 

[Rick Sykora]

As I mentioned, this is my first time doing a passive radiator design (with Bassbox). So, after not getting the f3 result I expected, I revisited the model again. Bassbox had flagged some of the params for the SB pr. Removing the flagged params, the recalculated f3 was 72 Hz. Much closer to my measured results (but majorly disappointing for the hoped bass extension). At least it helped with the port resonance and is much easier to work with than the twisty port tube!

The woofer tuning is considerably lower in the model than measured (48 vs 56 Hz). So, something is still off. Bear in mind, there is no crossover involved at this point and it will certainly play a future role. Bassbox allows you to model the crossover impact, but am not using as yet. Anyway, improving the woofer tuning does not appear it will get the desired bass extension. Will be good to try the Purifi pr as expect it is a better match.

 

[Rick Sykora]

So, just like a port, found Bassbox will recalculate the PR params if you ask it to suggest a box. While makes sense for port, (which you can lock its params), it means you can readily lose the PR params if not careful. Anyway, re-modeled both the SB and Purifi PRs. In both cases, to readily get a lower f3, the box volume needs to be smaller than the SPK5 is...

This seems counterintuitive, so am compiling some measurements for the SPK5/SB combo and will share once I can take corresponding measurements.

 

[Rick Sykora]

As am trying to reconcile Bassbox modeling with the real world results, considered the simplest mods that would make the greatest impact. Using the SPK5 with the SB pr as a baseline, I tried some variations and took some measurements. Here are the test cases:

1. Sealed off the back baffle (known to leak) with wide masking tape. This yielded a slight improvement, so kept for subsequent tests...
2. Reduce the cabinet volume by ~30% keeping cotton-based damping. This had less impact to tuning than I expected.
3. Restored original volume and removed damping material. This did not alter other tuning resonances, but the box tuning increased significantly (34 Hz to 40 Hz)

Not really getting the result I had hoped to achieve, am circling back to FR measurements today...

 

[Rick Sykora]

Well this took a while as my nearfield measurements were not conclusive. They were not as clean as my prior ones either. Have to get some other work done so will settle for reducing the 400 hz resonance for now. Will pick back up when the Purifi passive radiators arrive.

 

[jannek]

Maybe this build is interesting for you: https://hificompass.com/en/projects/2-way-systems/puri-bliss-bewg

It is a nice comparison of the possibilities of BR and PR with the Purifi driver.

 

[Rick Sykora]

Maybe this build is interesting for you: https://hificompass.com/en/projects/2-way-systems/puri-bliss-bewg

It is a nice comparison of the possibilities of BR and PR with the Purifi driver.

Yes, saw this before but had skipped over as do not have the milling machinery to do the waveguide. However, I had missed the part about the BR vs PR. Even after he created a pretty fancy port, it still had a port resonance and he abandoned in favor of the passive radiator.

Have been reviewing the simpler SPK4 port too. It is almost 10 inches and has a port resonance. If you shorten the port, you can still get some decent bass. As I mentioned before, a bit of bass eq works well since there is some excursion to spare.

 

[Rick Sykora]

Cut a new SPK5 back baffle in order to mount a Purifi PR and try compare it to the spec SPK5. Found Purifi recommends 2 PRs. Originally thought this was just for power handling, but seeing much better bass extension too. Confirmed this in Bassbox and VirtuixCAD. Will share more pics as I get more comfortable with VC.

To test 2 PRs, will need a different cabinet as will require side mounting. so have order a few Denovo cabinets to use for this purpose.

 

[Rick Sykora]

You may have noticed I have been struggling with modeling the PRs in Bassbox. While I get the approach that was taken, the PR interface seems incomplete. As I started developing more cabinet alignments, it became clear there was something else amiss.

After comparing some of the different alignments, I noticed the box volume for my original SPK5 pr model was off. The volume was based on using the external dimensions as internal ones. This meant about 25 liters rather than a 15 liters. So, this resulted in some expectation bias on my part when doing the measurements. My vented SPK5 modeling did not share this error, so any findings I stated for it are fine.

I regret if this error caused any major concerns, but am human too! I have fixed and so will adjust this thread to reflect and move forward...

 

[TabCam]

Would two side PR be advantageous for getting a kind of cardiod bass/midrange or am I now too optimistic and is at the back the best location and why?