Boom box build (another one, but first for me)

[wogg]

I got some final measurements done today to see how this thing ended up. The measurement setup was:

• 1M distance, on axis with the R side speaker being the only one driven. Sourced with the aux cable instead of Bluetooth, which allowed me to pan and shut down the other speaker for a better measurement.
• Far field measurement was gated, valid above about 270Hz
• Close mic on cone and PR, SPL adjusted for diameter difference and summed.
• Full response blended using the Excel tool at 300Hz.
• Displayed using the WinGraph program just updated

Key takeaways: The contour filter was pretty successful, this thing is pretty flat on axis save that nasty breakup peak at 6kHz. Notably, that was followed by a 12kHz spike in the measurement that didn't make it out of the blender for some reason. Unavoidable side effect of running ND90's full range. The bass response would be better tuned higher for sure, but this thing actually makes quite usable bass down below 40Hz. I had it sitting on the desk next to me while sweeping with an online generator and was amazed at the amount of output at 37Hz. As it stands, I'd call it at 80Hz for ~F3 and just above 40 for F10. It's behaving more like a low Q sealed box due to the too low PR tuning.

Put this thing near a wall or corner for the party and you're all set.

[wogg]

Oooo those are nice! And cheap too. Thanks!

[Millstonemike]

Outstanding. But you need a manufacturer label. That can be custom made here.

[a4eaudio]

Those drivers and passive radiators look great in that chamfered baffle! Well done!

[wogg]

Full assembly glamor shots! Regular work precluded me from making new confirmation squiglies though, so this will do for now.

Fancy shot


Straight on front, so you can see the matching handles on both sides.


And the back side, with the charging port, pairing switch, charge LED, and a little slot for some venting.

[wogg]

Filters assembled and mounted. Screwed to the top with a liberal blast of hot glue to help secure the hefty inductors. Added a little denim to the back panel, likely not doing much but if it smooths the upper mids at all it's a plus.

The kiddo is blasting a wide variety through it. Sounds quite nice, very full and plenty loud.

Tomorrow I'll do some glamor shots of the completed project and some verification measurements.

[wogg]

White Duratex going on while I wait for the final parts.

[wogg]

I'm using a 19V brick for this thing, it'll get loud and sound awful if cranked but no physical damage.

Got some measurements this morning. Measured about 24" away with both channels driven. I used the KAB amp for measurements connected through bluetooth, to mirror exactly how it will be used.

Here's the distortion, gated at 5ms and measured straight on in the center of the box, so both speakers are 15 degrees off axis.

Not bad at around and under 1% for the whole range, this was at a good 85+dB output. The high frequency roll off is due to the off axis response, those ND90s beam pretty hard.

I came up with a pretty aggressive contour filter. This is applied to an on axis measurement on the right side and used the actual R side impedance data. I blended with close mic of the driver only at 300Hz, didn't add the PR data. The measurement was taken with both drivers blasting, so there is interference between R and L in play. The depression between 700 and 800Hz lines up with the wavelength between the drivers, so I'm not compensating that as it will change across the axis.



Here's the schematic of the filter.


That gives the filter transfer function a 6dB smiley centered just above 2kHz.


I did apply the filter to the center off axis FRD data as well, using the same ZMA data for a rough approximation of what will happen dead center. The high frequency roll off is of course unavoidable but the response was generally smoothed out as well. This should help considerably.

[Millstonemike]

"and my daughter (11) wants her own speaker."

An EBS alignment might be better ... kids love their bass. But the 11-year old is soon to be a tween and then a teenager. Hence heavier demands on a boombox

On batteries (12 V) at max volume, the ND90s reach close to double their Xmax below 40 Hz. They may handle that - within their Xlimit. Something to test. It could be a problem plugged into a 24 V PS - too easy to overpower the ND90s. . Maybe they'd survive an old laptop supply at 18-19 V. Something to test ...

Box is looking good ...

[wogg]

Here we go, cones and guts loaded up and sealed. Test leads out the back for measurements.

Turns out the tuning is too low. Modeling was off with the Dayton parameters, and I didn't bother measuring the actual drivers. I just figured I'd shoot for 60Hz. I removed the nuts to make the PRs as light as possible, tuning is just above 50Hz on one side and just below on the other. That may loosen up with some break in, but I can't remove more weight so it'll end up a little bit of an EBS alignment.

I'll take acoustic measurements tomorrow for the contour filter. As it is, the midrange is too shouty.

[wogg]

Same process and materials from when I did my gloss black lacquer MicroSats a while back. Wet sand 1000 grit, though I was really light this time to avoid going through the poly. Followed by the rubbing compound, then the polish.

I'm not quite settled, there's a little milkiness when the bright shop lights hit it that I don't like. But... I'm not going for perfect piano gloss, so given the reality of what I'm doing this may work just fine.

[wogg]

It's almost time to do some sand and polish! I just have to finish off this regular work day.

I did smear the MDF edge with Titebond last night to seal it up. That should further minimize the potential risk of swelling up my top panel while trying to wet sand. Fingers crossed.

[wogg]

Ugg... I hate waiting. I'm going to give it a full 48 hours before I mess with that finish at all, so all I can do is stare at it and find flaws.

Wet sanding will be tricky with the unprotected MDF top panel glued on there, I don't need that part sponging up and getting all mushy. So the actual wet sanding will be minimal, but I will use rubbing compound / polish on the whole thing. The surface won't be mirror flat of course, but I think I can even out the imperfections enough that they don't bother me.

I can't wait to get this thing sealed up and make noise. I need to tune the PR's and do a impedance / SPL sweep to create a contour filter yet.

[tomzarbo]

After you get it as flat as you want surface wise, it might just be easiest to put one last very thin coat of rub-on poly to get your even sheen back. Minwax makes decent rub on in pints, dries fast too.
Looking great!

TomZ

[davidroberts]

That is why I always use lacquer. Never had the patience for poly.

If you do any sanding you might want to give it some time, poly takes a while to fully cure.

Build is looking good. Can't wait to hear your opinion on the results.