New subwoofer design, Hi-Vi D10G

[Dr.EM]

I recently thought about how I could put to use some parts I have around. Namely, 2 x Hi-Vi D10G drivers and a 160W into 4 ohm 'plate' amplifier (ESP P101). I unfortunately doubt this will ever get built, but I thought it may be of interest to the forum!

The basic design is both drivers loaded into a 90L sealed enclosure. This is too small for these drivers and also they won't play to subwoofer frequencies in this configuration, so the obvious choice is to use a Linkwitz Transform circuit to correct the high Q and extend the low frequency. I'd ideally have the -3dB point switchable to different settings to allow easier setup in different rooms but shown is what I'd consider a quite appropriate setup for medium size rooms. With -10dB at 18.5Hz I'd expect to hear it quite flat to 20Hz with room gain added in:



Here is the max SPL trace for these drivers in the 90L box:



And the cone excursion with 160W input power:



And the group delay, inclusive of the LT circuit:




This is designed to be an accurate music subwoofer, I wouldn't expect it to have the power sub 30Hz for much HT use, but it depends just how much room gain you get. The design lends itself to corner placement too so this should help get as much out of it as possible, and onto that design!

The design uses 2 drivers, so for me it was obvious I'd want to implement both vibration cancelling and distortion reducing mounting techniques. To do this one driver has the motor system facing outwards while the other, mounted parallel, has the motor within the enclosure. The 2 drivers are mounted to the box and to each other by threaded rods and nuts; with 12 mounting points per driver this should provide very rigid coupling between the drivers.

The sub, with and without grilles, person for size reference:



On the near side you see some 'boxes', which are the amplifiers and electronics which would as such be hidden under the removable grilles, "set and forget" style. The connectors would be on the back as usual and the power switch is on the front brushed aluminium strip. On the far side is a semi-circular bulge, which is just 9L of enclosure volume shifted into the otherwise unused space to reduce the footprint.

My housemate kindly produced this render of the design in 3DS max, showing it situated in a corner:




Thanks for looking! Please post any thoughts you have on the design

[chris661]

Looks good, but I wouldn't (personally) expect 20Hz out. I know my own hearing drops off rapidly below 30Hz, so that even with large headphones, 20Hz is inaudible.

Ideally, you'd experiment with the Q and f3 of the LT circuit, so you can find the best settings to integrate with your room. Some would say that Q=0.5 would give better results, others prefer different roll-offs. It's personal preference, but don't choose something because it models flat.
Of course, reducing the amount of gain in the low frequencies would allow you to turn it up more (as the woofer isn't using more excursion to go lower). Some reading for you (if you haven't seen it already) - ESP - The Linkwitz Transform Circuit

Apart from those (relatively small) things you might want to think about, it looks like a well thought out design - my first subwoofer was something not dissimilar, but with a pair of 8" drivers.

Interested to see how this goes
Chris

[Dr.EM]

Thanks for looking!

I just tried my hearing with large over-ear headphones and I'd say for me 24Hz is the lowest I can really hear a tone at, below I think I'm mostly hearing distortion, no real 'bass' feel. I wouldn't expect 20Hz to be strong but should easily be over 100dB continuous output level in a room, so I'd expect it to be audible, if any music has it in of course! Definitely a subwoofer designed with music and accuracy in mind, not massive output, but at the same time I'd like to think you wouldn't be missing anything provided you aren't playing extremely loud.

Interesting point on the Q factor, I hadn't actually looked at it, only the frequency! Which is odd as I build sealed cabs to have around 0.6 usually. Will play around with it and see if it has any benefit, though I think it's something you can only decide on once you know exactly what room gain will do. I've seen the ESP article and the info on Linkwitzs site, used to have a spreadsheet that would calculate component values too, which was useful, will have to find that again.

I think, overall, room gain and corner loading will influence an optimal design quite substantially, which is unfortunate as they are slightly unpredictable factors! A few adjustable settings would be included if this were built, for maximum versatility.