Long post
I’ll start with my personal journey through the audio design landscape, starting roughly a year ago at the end of 2018. Not that I hadn’t done any audio projects before that, but this was when my second downward spiral started into audio electronics, cabinet design and driver selection began proper.
Believe it or not, it all started with a 1€ promotion sale from a local hardware/hobby new location opening. Namely with this kit that one of my work colleague bought like 4 units. And handed me one to assemble for the office.
Now, this was at the same time cool as heck and awful as hell. As a concept it was brilliant, bluetooth and a cardboard “enclosure” that snapped together with magnets, but the reality was “Well this is cool and functional and all… but it sounds like absolute shit” and from there I started to design new enclosure for this thing in the first place, and I switched the 3 alkaline batteries for a single 18650 cell, which worked great.
Now this thing still had everything original electronically. Electronics module and drivers. It was better, but it still sounded like shit (no low-end at all), so from there it was time to choose new drivers and make a new enclosure for them. I’m pretty sure it was around this time I downloaded and installed Winisd and got my first touch into enclosure design and driver selection.
I choose as my first chosen drivers were a pair of Peerless PLS-65F25AL04-04 and designed a completely new enclosure for them, but still used the bluetooth amp module, which was functional.
Now this was a huge improvement for the low-end performance in relation, but it still wasn’t enough. After couple of weeks I decided I needed more LOW END, so again I went hunting for drivers and had now discovered one of the BEST places to find the Thiele/Small-parameters for thousands of different drivers at http://www.loudspeakerdatabase.com/ and like 99% of the drivers have a Winisd ready parameter file ready to just download and plug into the program to get, so suddenly it was really easy to compare different drivers to each other.
Now at this point I still didn’t want to break the bank, so I decided to go for a reasonably priced pair of Dayton Audio’s ND65-4.
I unfortunately don’t have any pictures with the drivers fitted, but I just used the same enclosure and made a new face plate for the NDs. I also printed a new enclosure with higher infill percentage in case you wonder the color change of the enclosure.
Now it was better than the with the Peerless drivers, but still, the low-end was lacking.
Ok, so more low-end… At this point I had maxed out the capabilities of the original bluetooth amp module that had come with the original cardboard kit. I needed more tools to achieve this… It was time to design my first own audio amp! I studied the bluetooth amp module for this purpose quite long and discovered that it used dedicated class-D mono amps for the left and right channels. Upon closer look it seemed that there was just like only 5 external components including the amp IC… Well that seems simple. Couple weeks later I have my first audio amp PCB designed and ordered. A low voltage mono Class-D amp.
I couldn’t find any IRL pictures of the board, but here’s the Kicad project’s 3D renders
Now I had alongside the amp IC added a little simple op-amp filter block in front of it (for bass boosting), but had designed it completely wrong due to being a pleb in op-amp analogue circuits… But the amp itself WORKED. And I had now designed my first ever own amp, even if it was as simple as it could be. I spent multiple weeks tinkering with the little amp learning about gain, single-ended and differential input signals, single-ended and bridge-tied loads and much more related to designing audio amps, thanks to this little, simple, board.
Ok, so I know had some audio electronics ideas to get me forward with my objective of getting more low-end on my speakers. At this point I had also realized that the little drivers simply were not going to be able to move enough air to produce the low-end, so I decided to build a completely new enclosure for the next speaker and make it a two-channel design, with a dedicated low-end driver and a re-use the DA’s ND65s for the medium and high frequency. At this point I was prioritizing a small enclosure, because I wanted it to be portable and battery powered. I ended up with a Tang Band W3-2108 micro-subwoofer for the low-end driver and it is quite amazing for it’s size what it can do in a reflex enclosure.
First mock-ups of the front plate and reflex tube.
It was at this point I realized what “port noise” / “port chuffing” meant. The tube was woefully too small surface area wise and it was then when I also discovered the other problem with reflex tubes in small enclosures. You end up with really long tubes… and the packaging becomes an absolute nightmare, as you try to stuff a figurative anaconda of a reflex tube into the enclosure, which also then eats up the internal volume lowering the low-end response…
After many different iterations of port design and packaging I found a good compromise between port length, diameter and packaging. Now I could start designing a dedicated two-way amp for the speaker.
It might be packaged, but there are no electronics inside despite the volume and power button.
So okay, I need to design a new amp for the two-way speaker, I need to get more low-end and also more power/volume, but I still want to use a single lithium series cell pack, so I need to boost that voltage higher for the amp ICs, to get more power to the drivers…
So boost converter from 3,6V -> 12V, op-amp section to make an active cross-over for the high- and low-drivers and dedicated amps for both. Result is shown below.
No electrolytic caps fitted
And it works pretty dang well. It could do with another HW iteration, but once again I learned a lot from this small project and designing multi-way speaker and an active crossover. The boost converter starts to choke at higher powers, but it’s capable of driving the TB’s three incher in a very nice way.
At this point I sort of did a little smaller speaker side project with another Peerless driver I had ordered earlier just because it looked interesting… Well I designed an enclosure for it and finally actually integrated all the electronics inside it as well, which I partially took out later after discovering that they cased some air leakage from their mounting holes (note to self, in the future, put panel mounted components into their own separated section to isolate them from the rest of the enclosure to avoid air leaks).
Look at that reflex port packaging though.
I used the amp board for the two-way speaker, but assembled only one of the class D amps and skipped the crossover in the op-amp section to get full-range, although I did use the bass boost section in the low-end driver op-amp section to get a bit more bass out of this thing.
This thing is pretty neat for it’s size and would be “complete” if I redesigned the front panel, so the panel mounted components didn’t have air leaks. only the micro-USB connector for the lithium cell charger is un-populated at the moment, so you have to pop the back open to charge it. It also doesn’t have over-discharge protection, so the board usually discharges the pack to 2.3 Volts, if you just use it long enough or forget it on.
So that’s what I have done over the past year, but this post is already long enough so I’ll stop here and tell in the next one what I’m working on now…