I've wanted to make a decent bluetooth speaker for a while now, and have been slowly collecting the parts to do so.
The goals here are pretty standard:
There are some other features that I'm planning on building in, which make this more geared towards the sort of work I do.
My "reference" for this build is a Bose Soundlink Colour, in red. It's super-compact (as thick as a decent book, and a similar-size front to a CD case) and sounds reasonable at low/mid volumes. Turned up, some LF compression kicks in to rescue the tiny drivers. It hits 87dBC at 1m when playing pink noise. NB - the LF output has disappeared entirely by that point, and it only got to 90Hz to start with.
I looked around online for bluetooth speakers to see if there were any that were worth just buying outright, and saving myself a lot of time/effort. The UE Hyperboom looks pretty good, and the JBL Xtreme looks okay, too. However, I'm pretty sure I can do better.
Here's the prototype far:
In total, there are:
4x Kartesian Sub120_vHP (the ferrite 4" mini-subwoofer)
2x Lavoce TW131.00
2x SB Acoustics 8x5" passive radiators
1x Zoudio AIO438 amplifier
The cabinets are re-used from another project, but they're doing the job for now. The planned cabinet will be around 20cm square at the top, and 35cm tall. The 4x Sub120 drivers will be in pairs on L/R sides, and PRs will be front/back. That way, I've got complete force cancellation and the speaker shouldn't walk around much.
HR sim for 2.83v into each driver:
Which is very good for a 11L (net) cabinet. I did try simulating with other drivers, ranging from Fostex 6" full-range units to 10" PA drivers, to see if anything else would get close. They often managed more efficiency above 120Hz, but I didn't find anything else that would keep up below 100Hz. Since there's a limited amount of voltage swing available, I didn't want to apply lots of LF EQ and run into clipping.
The Zoudio amp allows a useful amount of DSP functionality, so I've been playing with the crossover to make sure the tweeters cover more of the midrange (to ensure wider dispersion) but don't run into trouble from doing so. These Lavoce tweeters have held up well so far, but obviously the centre-to-centre distances are large at the moment, and the vertical off-axis response is weird at best. Final tuning will have to wait until the final cabinet is built.
For now, though, the prototype hits 103dBC@1m playing pink noise, beating the Bose speaker by 16dB. More, if you count the LF where the Bose was ~10dB into limiting. The extra octave of LF extension is nice to have, too.
For batteries, I'm planning on using some 18v-rated (they're more likely to be 20V once fully charged) power tool batteries. The advantages include:
The disadvantage being they're expensive. That's okay, though. I want this project to do well, and sometimes that means throwing some cash at it. I decided against DIYing the batteries after realising that making them easy to swap out would get quite tricky. With the power tool batteries, you can find battery holders/adapters so the battery will slot in as if it's being installed in a tool.
Finally, I'd like to (briefly) discuss one of the not-so-standard features I'll be including:
- The Zoudio amp can be programmed and reprogrammed quite easily, and I'm under no illusion about the efficiency of the Kartesian mini-subs. In order to multiply the usefulness of this speaker, then, I'll include a couple of NL4 SpeakOn sockets and some switches. The idea is that I can disconnect the internal speakers, and power some larger, more efficient speakers. I have plenty of PA speakers around, some of which are very efficient indeed. For the times when I need more coverage/SPL and still need to run on batteries, this little speaker will be rather useful.
I think that's everything for now. The next steps will be to buy up some batteries, and build a better cabinet.
Chris
The goals here are pretty standard:
- Usefully loud
- Decent LF output
- Wide dispersion
- Long battery life, and easily-changed batteries
- Sound good
- Easily carried across a field
There are some other features that I'm planning on building in, which make this more geared towards the sort of work I do.
My "reference" for this build is a Bose Soundlink Colour, in red. It's super-compact (as thick as a decent book, and a similar-size front to a CD case) and sounds reasonable at low/mid volumes. Turned up, some LF compression kicks in to rescue the tiny drivers. It hits 87dBC at 1m when playing pink noise. NB - the LF output has disappeared entirely by that point, and it only got to 90Hz to start with.
I looked around online for bluetooth speakers to see if there were any that were worth just buying outright, and saving myself a lot of time/effort. The UE Hyperboom looks pretty good, and the JBL Xtreme looks okay, too. However, I'm pretty sure I can do better.
Here's the prototype far:
In total, there are:
4x Kartesian Sub120_vHP (the ferrite 4" mini-subwoofer)
2x Lavoce TW131.00
2x SB Acoustics 8x5" passive radiators
1x Zoudio AIO438 amplifier
The cabinets are re-used from another project, but they're doing the job for now. The planned cabinet will be around 20cm square at the top, and 35cm tall. The 4x Sub120 drivers will be in pairs on L/R sides, and PRs will be front/back. That way, I've got complete force cancellation and the speaker shouldn't walk around much.
HR sim for 2.83v into each driver:
Which is very good for a 11L (net) cabinet. I did try simulating with other drivers, ranging from Fostex 6" full-range units to 10" PA drivers, to see if anything else would get close. They often managed more efficiency above 120Hz, but I didn't find anything else that would keep up below 100Hz. Since there's a limited amount of voltage swing available, I didn't want to apply lots of LF EQ and run into clipping.
The Zoudio amp allows a useful amount of DSP functionality, so I've been playing with the crossover to make sure the tweeters cover more of the midrange (to ensure wider dispersion) but don't run into trouble from doing so. These Lavoce tweeters have held up well so far, but obviously the centre-to-centre distances are large at the moment, and the vertical off-axis response is weird at best. Final tuning will have to wait until the final cabinet is built.
For now, though, the prototype hits 103dBC@1m playing pink noise, beating the Bose speaker by 16dB. More, if you count the LF where the Bose was ~10dB into limiting. The extra octave of LF extension is nice to have, too.
For batteries, I'm planning on using some 18v-rated (they're more likely to be 20V once fully charged) power tool batteries. The advantages include:
- Easy to swap out
- Easy to charge
- Warranty
- Can use them for power tools
The disadvantage being they're expensive. That's okay, though. I want this project to do well, and sometimes that means throwing some cash at it. I decided against DIYing the batteries after realising that making them easy to swap out would get quite tricky. With the power tool batteries, you can find battery holders/adapters so the battery will slot in as if it's being installed in a tool.
Finally, I'd like to (briefly) discuss one of the not-so-standard features I'll be including:
- The Zoudio amp can be programmed and reprogrammed quite easily, and I'm under no illusion about the efficiency of the Kartesian mini-subs. In order to multiply the usefulness of this speaker, then, I'll include a couple of NL4 SpeakOn sockets and some switches. The idea is that I can disconnect the internal speakers, and power some larger, more efficient speakers. I have plenty of PA speakers around, some of which are very efficient indeed. For the times when I need more coverage/SPL and still need to run on batteries, this little speaker will be rather useful.
I think that's everything for now. The next steps will be to buy up some batteries, and build a better cabinet.
Chris
Veey cool Chris. This is my garden boom box. 2 X Tectonics BMR. In stereo but off axis relative to each other to attempt to help dispersion. Single 6.5" vifa sds 165 , ported. Powered by a 12v SLA. Tpa3116 3 channel with an 'active filter '. Cheap and cheerful but suits us for garden listening to decent levels.
Always amazes me how good it sounds outside when no reflections .
Always amazes me how good it sounds outside when no reflections .
Attachments
Here’s my outdoor “boombox” that’s in progress. Waiting on battery bms from China. 
It’s 2x Dayton sd215 sealed in 11L. Braced very well. Lined with 1/4” cork paneling.
1x Tc9fd-18 in triangle pyramid enclosure.
Using 3x 60w tpa3118 pbtl amps and 20v tool batteries for now.
Minidsp 2x4 doing the xover at ~250hz. +9db bass shelf at 60hz iirc. Plays low.
It’s 2x Dayton sd215 sealed in 11L. Braced very well. Lined with 1/4” cork paneling.
1x Tc9fd-18 in triangle pyramid enclosure.
Using 3x 60w tpa3118 pbtl amps and 20v tool batteries for now.
Minidsp 2x4 doing the xover at ~250hz. +9db bass shelf at 60hz iirc. Plays low.
IIRC, the absolute maximum input voltage for the AIO438 is 26VDC, and 2xSLA in series would hit about 29V fully charged. Plus, lithium batteries are much smaller & lighter for a given amount of energy storage.
BMRs look cool! How do you find the dispersion with 1x up-firing and 1x forward? I could probably replicate that with the tweeters. The woodworking looks excellent, too. Was that CNC'd?
Eric, nice project. I bet the 2x 8"s have a lot more thump than most bluetooth speakers.
Chris
Funny enough I had let this project sit for a month- then we visited a friend who was showing off his new Sonos roam.
Well I got home and finished this up after a night or two (to where it is, ughh). Short answer is yeah it shakes my (small) house. It’s pretty unique sound for me bc it keeps a nice amount of low bass at quiet volume, not used to hearing that in a portable speaker. The dsp really makes it possible. I need to do outdoor measurements for real once it cools off a bit.
https://www.diyaudio.com/community/threads/omnidirectional-work-in-progress.303941/page-5 some OMI ideas here.
could 3D print some kind of omidirectional horn for a compresion driver + midranges as apple homepod? (would solve the multiple source issues).
could 3D print some kind of omidirectional horn for a compresion driver + midranges as apple homepod? (would solve the multiple source issues).
Thanks Chris. I can't take any credit for the box construction. I think it was a wine gift box set or something. I did recess the BMRs with my router.
The dispersion is good. I know not ideal with L and R, I probably should mix the stereo and send mono to both but it's fine for garden use. If it's facing you can can pretty much walk around it 180° and the sound changes very little . Vertical and horizontal Polars must look pretty good.
It doesn't go too loud before I assume the 6.5 reaches its limit but plenty enough to annoy neighbours! And like your cool project Eric the LF is good at low level. It just uses a '2.1' Chinese amp.
If it were me building something like this, I would at least use some very high efficiency neo magnet drivers like PRV 6MR500s mids, Selenium ST200 tweeters and some smaller Peerless long excursion sub drivers as a 3 way design. A good class D dsp plate amp that runs off a medium sized 12 - 24V LiFepo battery would give you plenty of volume and last a good few hours, plus it would sound great while doing so.
Power tool batteries are a great idea. I have alot of Ryobi tools that use 18V 4Ah Lipos and they can be had for 50 bucks on sale. They also make 40V batteries for larger stuff, which could net you some decent power output.
The more I think about it, the more I like the omnidirectional idea. I'll do some thinking in that direction. One way to do it (given the DSP available) would be a small midrange (SB65) and a very small tweeter (something in the 1/2" range), both up-firing. Crossover before the SB65 gets narrow, and boost the VHF of the tweeter to get the 90-degree response sensible.
Some testing to be done, but moving to 3-way would also widen the dispersion in the 600Hz-1kHz range, since the 4x 4" drivers are starting to get narrow there.
The place(s) that I'd be using this in anger have their nearest neighbour approx 500m away, and a substatial chunk of that is forest. I'd also use it at home, but obviously not at full power.
Chris
That sounds like a great idea you could do the top two ways passive to keep it within your amp channel count. I do this on my tops and it works quite well with the only downside I have found been that limiters can't be set independently for each driver which could reduce maximum posible output. The passive crossover can also be simplified as EQ can be done in the DSP.
Well, if it's omni and there's basically zero physical separation between the channels, I think mono would be fine. I don't think there'd be much to gain from trying to keep stereo.
I did try to see if a stereo effect could be done from a single box, but that sort of thing relies heavily on reflective surfaces nearby, and those surfaces need to act fairly predictably. Since this speaker will be used in many different environments, anything relying on reflections will fail in some/most situations.
I like the idea of a tiny 3-way because it means the drivers would be operating in a range where they're acoustically small, and would therefore have very wide dispersion without any need for reflectors etc.
Chris
I did try to see if a stereo effect could be done from a single box, but that sort of thing relies heavily on reflective surfaces nearby, and those surfaces need to act fairly predictably. Since this speaker will be used in many different environments, anything relying on reflections will fail in some/most situations.
I like the idea of a tiny 3-way because it means the drivers would be operating in a range where they're acoustically small, and would therefore have very wide dispersion without any need for reflectors etc.
Chris
You can get a nice stereo effect with a small box if you radiate the HF in both L/R directions with super tweeters in addition to a forward radiating tweeter covering a wider band crossing lower to a mid driver. I've built a few portable amplified speakers this way and they sounded very real from further away with recorded live music.
A synergy horn would actually be ideal for larger applications but labor intensive to design and heavier compared to a traditional direct radiating multi way design. They would also be more directional and larger as a horn loaded design.
In all cases, the higher the efficiency the longer the battery life at a given volume level and also the higher max SPL. You would almost certainly need to use a passive radiator setup instead of a physical port, being the box is relatively small for the tuning frequency needed.
If you want an omni radiating tweeter, just mount the tweeter on top firing upwards and place a cone pointing down onto it used as an omni reflector. You would need mids facing in all directions to keep the box smaller and have even power response.
A synergy horn would actually be ideal for larger applications but labor intensive to design and heavier compared to a traditional direct radiating multi way design. They would also be more directional and larger as a horn loaded design.
In all cases, the higher the efficiency the longer the battery life at a given volume level and also the higher max SPL. You would almost certainly need to use a passive radiator setup instead of a physical port, being the box is relatively small for the tuning frequency needed.
If you want an omni radiating tweeter, just mount the tweeter on top firing upwards and place a cone pointing down onto it used as an omni reflector. You would need mids facing in all directions to keep the box smaller and have even power response.