InDIYana build - SuperNova Minimus

[Jake]

Coming along realike nice wogg. The duratex will look nice with the walnut all shined up.

[wogg]

Finished and back together, leads still out the front for crossover prototyping.

[Billet]

Looks nice!

[wogg]

Looks nice!

Thanks! I feel like I'm cheating a bit compared to the other builds, so I'm counting on nostalgia factor.

[Billet]

My boxes are made this year but in the same general nostalgic theme. Maybe we could try to park them on the same table at InDIYana?

[KEtheredge87]

Thanks! I feel like I'm cheating a bit compared to the other builds, so I'm counting on nostalgia factor.

Nothing wrong with nostalgia factor! I played that card HARD last year with my SuperBees ;) I've never heard the original Nova 15's, but I am very excited to hear your upgraded rendition. InDIYana is going to be great this year!

[Kevin K.]

I like the old school look, nice work!

[wogg]

Thanks KEtheredge87 and Kevin K. ! Really looking forward to this one, I've only ever been to MWAF so this should be a cool contrasting experience.

[wogg]

Used some clippy test leads to prototype the crossover together as simulated, and looky there, they are pretty much the same. The overlay is the actual ARTA measurement.



Here's the system distortion measurement, under 1% everywhere.


And the measured reverse null. This doesn't seem that deep, but it's there.


This is only 1 speaker so far, now I need to assemble its mate and hope it's the same. I did blast some music through the 1 out in the middle of the room like it was. I pretty much hate mono music though, especially when it's not actually mono but just the L channel of a stereo track. That made it hard to have actual impressions, and I did find the bass weak. I do tend to like bass boosted systems though, so I'm reserving my opinion on that part until they're both together. The port isn't glued at the moment, so I can adjust tuning. May need to go higher to punch up 50Hz, but that will require sawing off some of the elbow. After some proper listening...

[jhollander]

A 1.2 mH inductor on the woofer looks light. Might want to take a near field and blend in

[wogg]

A 1.2 mH inductor on the woofer looks light. Might want to take a near field and blend in

Yeah, I'm going to have to figure that out. That will tell me if it's flat and I just like bass or if I under did the baffle step due to my measurements ending at 300Hz. The downside of using the iMM6 like I am is that I can't use a mic stand, so I have to basically hold the computer in place by hand to take a close mic sweep. Not exactly ideal, but should get me in the ball park.

Also...

What's a good tool to blend both the close mic woofer measurement and a port measurement? I played with Blender in Excel to do that, but wasn't sure about the port part. The woofer blend is intuitive, but the port is not. I think I have to adjust the SPL manually using the radiating diameter formula, then add it on the tab that is typically the baffle signature.

[KEtheredge87]

What's a good tool to blend both the close mic woofer measurement and a port measurement? I played with Blender in Excel to do that, but wasn't sure about the port part. The woofer blend is intuitive, but the port is not. I think I have to adjust the SPL manually using the radiating diameter formula, then add it on the tab that is typically the baffle signature.

Hey Wogg, it sounds like you're on the right track with the blender tool. I know Jeff Bagby's white-paper mentions summations of port response with the nearfield response, but I do not recall an actual example. Within my own brain, I wonder how things would be different if you have front firing ports vs. rear firing ports? If Jeff B. ever gets a wild hair to instruct the masses further... perhaps an example of this would help!

[jhollander]

The baffle response tends to flatten out before roll off so without the port the blended response will give you an idea of the reduction required for the baffle step. Without blending you could target -6 dB from the peak woofer response near 1K and be close.

[wogg]

This is how you start gathering a bin of extra crossover parts...

So I did the close mic measurement and port combination. Reading through Jeff's white paper again, he uses PCD to sum the woofer and port, then Blender to add that to the far field. I use WinPCD, but the process is similar. I just added the port and cone close mic sweeps as a woofer and tweeter and added them without any crossover to mess with them, using the woofer ZMA for both.



Then the sum was saved out, and pulled into blender to mate with the far field. That was processed for min phase, as was the tweeter far field and the system sum far field. Then it was back into WinPCD for checking the crossover. With all files processed for min phase, and the far field woofer + tweeter not including the close mic part, I found the Z offset alignment not as clean as it was before. I did my best to get it close and went from there.

Long story short, I was about 3 dB short of full baffle step. That would be great for near wall or bookshelf placement, but not for stands in the middle of a room. So... I tweaked the crossover and bought more parts.

Here's the new crossover with an overlay representing the prior version without the close mic, you can see the 3 extra dB of baffle step applied.

[wogg]

For some improved listening testing, I pulled one of my studio monitors off the stand and plopped the L speaker in it's place. Without the baffle step my Dynaudio definitely had a bit stronger bass, but the rest of the spectrum was very comparable between them. I didn't sense any odd imaging due to any major response variations between L and R. It was really pleasant to listen to, and I burned some time just playing music while tweaking the crossover.

Very nice... coming together!