General DIY audio project thread

Well i am very interested in music, I actually play the violin for 8 years now. I listen to a lot of classical music but also more modern. Fucking hate rap like xxxtention and company, sorry.
I got into speakers but not the DIY scene. I own Minirig speakers from the UK. They are portable BT speakers and I would say they are one of the best out there. I have 2 minirigs and 3 subwoofers lol.

1 Like

Wow thank you for the rapid ID.

Anything else I should be wary of when desoldering and resoldering them? Reground to same area or a cleaner looking one?

1 Like

Do you own a multimeter, can you measure with it? If not then you’re pretty limited in your ability to diagnose the possible problem spot.

I would personally guess that the power supply section right where your AC power comes in to be the problem block and something has gone wrong in it that it doesn’t anymore filter the supply lines and that noise is making it’s way to the amplifiers on their power rail and therefore to the speakers.

EDIT:
Perhaps one of the filter capacitors are burnt out inside and have gone open-circuit AKA not doing anything.

Do you know of a decent dual Bluetooth speaker amplifier for ND65-8?

Anyone have experience with fixing broken earphone wires?

My approach that yields the best results to date is unwraping the colored cables from the non-conducting fibres, cutting the fibres off, then burning the colored coating off with a flame (I use my stove, lighter would also work), and then twisting the uncoated wires together to make them bond. Usually you want 3-5cm burnt for them to make a strong connection when twisting together. After that I tape them down with electrical tape. Lasts okay for a month or two after then.

I’ve tried soldering, but the solder doesn’t stick to the wires at all. Maybe I’ll get better results now that I have solder paste, but I doubt it.

P.S. Anyone know of a good yet cheap MMCX cable? All the ones I’ve tried get a loose connection where they connect to the earphones after a few months. :frowning:

A decent one? I don’t know. I have mostly dealt with the cheap chinese BT modules, so I don’t necessarily know enough to give a recommendation to a “decent” once

You could try the M38 variant of this super cheapo bluetooth module, as it has BT connectivity and integrated amps and go from there.

1 Like

These are awesome. Super inspiring, will look at this when I have money again, looking to have some good monitors for my desktop

How’s the sound quality overall?

The solder not sticking is most likely caused by there still being coating left on the wire and/or the open flame will char the enamel coating to the wire and you will need something slightly coarse to very gently sand it off after that.

This repair operation is gonna be tricky, if the coating on the wires is polyamide based (kapton), as it will not burn off with just the heat of the soldering iron with a solder blob at the end. Now if it is polyurethane then it will just burn off from the heat of the soldering iron and those are much easier to do re-work on.

1 Like

I’ll be using the M18-module (smallest) for my upcoming DSP+multi-channel amp build, but haven’t tested how it is sound quality wise, just tested it a little bit to make sure it worked. Line output was pretty low, so you might need some higher gain in the front end of your signal chain to compensate, but if you’re using the module with the integrated amps then I have no clue, as I have not personally used that one.

Here’s a couple YT videos with M38 board and you can search for other modules there as well.

1 Like

Dope thanks. I might have skimmed over it but did you get your speakers from parts express or something like that

I’d love to order from parts express, as they have excellent selection and excellent prices, but due to them being located in the states, the shipping and import taxes to Finland raise the price too much and I have therefore bought most of my better and expensive elements from Europe @ https://www.soundimports.eu/en/

1 Like

Ah gotcha. Yeah I guess I have an advantage there

Just had time to read now, really cool, would love to listen and see how the 3D printed walls handle vibrations and all

Since you are going to ever increasing level, I say you must get a measurement microphone, you can by luck get a speaker to sound good by ear, but never do a proper tuning considering power response, that is what really matter for a speaker to sound good, with good bass but not overwhelming, with great detail but not fatiguing after a few minutes of listening

This is the design for the passive and active versions of my boombox crossover, 29 different measurements done for it, I should have done double that since the speaker its not symmetrical, its a lot of work

But man, today was the first day I gave a listen to it, and the only thing I could do was grin from ear to ear

Passive crossover

Active crossover

It’s not an ideal material, but it’s cheap´, I have it on hand and I can use it now. I don’t have tools or working area to fabricate out of wood, so 3D-printing is currently the easiest way for me to build the enclosures. Maybe in the future, if I get my own company going I’ll be able to get a work shed somewhere and buy a nice table router and start making stuff out wood and aluminium too. I’ve had my eyes for a long time on a http://www.cncrouterparts.com/pro4896-4-x-8-cnc-router-kit-p-253.html , but budget and no place to put it into is the limiting factor currently.

It allows me to easily just experiment with different designs and hold up if while staying in the lower
audio pressures.

I’ve been going by ear for now with the: “if it sounds good to me, it’s good” mentality, as I don’t have any anechoic measurement chambers available to measure the actual response. One of my work colleague has calibrated measurement microphones from his earlier cabinet measuring days, but also doesn’t have measurement chamber. Besides because I build mostly miniatures, battery-powered and portable stuff the environment is always changing and with that the end response, so I don’t see the goal of getting reference chamber measurements to be that high of a priority.

I’m going the direction of DSP, well first of all, because I want to add one more thing to my skillset (toolbox) and I find the concept interesting, I want to tinker with it. They also allow me to do the filter building much more cheaply, flexibly while conserving PCB space.

The particular DSP IC also has only digital outputs (PWM or I2S) and I want to test some digital input amp ICs. Allows me to stay completely in the digital realm after initial front-end AD-conversion.


All in all, I don’t approach my audio builds with too much seriousness, but rather just try to make something cool and interesting that also sounds good in the process.

1 Like

I understand you, working with MDF makes such a mess, lately I’ve been offloading the cutting job to be made on CNC, a bit more expensive, but the end result is much nicer, and no mess made

I want to try 3D printing some day, need to get some new material, ABS would certainly warp in my printer with a big part

You actually don’t need a chamber to do it, all the ones I’ve posted are done in my living room, you use gatting to cut off the reflections, and them merge near field measurement to get the full spectrum response

The DSP really makes the difference, I can’t live without one, for small portable speakers if you don’t rely on additional processing to shape the response you can go without, but for stationary system, it with a measurement microphone make day and night in room correction, I had to add one in every system in my house after seeing what it could do in my bedroom

I’ve seen some board with this approach of keeping the digital domain, really should help with the noise floor, the off the shelf boards don’t have a good one, for the TDA7498 that I’m using, it has too much hiss at high gain settings, I’m trying to find a way to fix this

But for me it’s all for fun also, once you get serious it loses the joy, or when problems start to happen that you can’t figure it out

2 Likes

if sounddesign counts as diy audio :stuck_out_tongue:

2 Likes

Ok, so I dedicated a mammoth post to my past year of audio tinkering and journey, so now I’ll dedicate this post to my current project.

I’m building a portable, battery powered boombox with bluetooth, DSP power and USB-C + Power Delivery charging for the battery pack. Based on my testing on the earlier two-way speaker with the Tang Band W3-2108 micro-subwoofer made me realize that it is possible to get good bass even in small enclosures. The biggest problem is that max volume is going to be limited, as you will soon hit the limits of excursion for the driver to be able to move more air for more volume.

From there I spent couple days looking for good driver candidates for the low-end drivers. I also wanted to get more volume for the low-end so I decided to look for bigger surface area driver, so they have more potential for more volume. After many hours of looking through http://www.loudspeakerdatabase.com/ I ended up with Visaton W130X-4 DVC, which is a 5 inch driver with a dual 4 ohm voice coils.

Here’s the simulated response from Winisd for a closed, reflex and a closed w/ PR enclosures around 4 litre end volume. I chose to go with a Closed enclosure with a Passive Radiator to extend the low-end a bit further and which I then plan on leveling it out with the DSP block in the amp board.

I ordered the the speakers and the passive radiators from https://www.soundimports.eu/en/ . The Passive radiator is Dayton Audio’s DSA135-PR, a 5 inch with an aluminium cone. Once I got the drivers I quick prototyped a test enclosure for testing the low-end. I drove it with the two-way amp, with the high side still being routed to the ND65-4 mounted in the old speaker.


It was pretty promising, it sounded good (thicc) and I believe it will sound even better once I’ll be able to correct the low end response a little bit via the DSP. I did notice that the PR was having a lot higher excursion compared to the active element and I might have accidentally chosen a too small one, but if that becomes a limiting factor in the future I can just design a new backplate for the enclosure that’ll be able to fit a bigger PR.

I was first planning on making the enclosure from wood and doing some kind of a vinyl wrap on it with a aggressive pattern to give it a bit more aggressive look. Below is a concept render from blender.

Unfortunately I got a bit of a cold feet with going first directly to the wooden enclosure, due to not having materials at hand and no tools. I then decided to build the first one by 3D-printing, so I modeled the project a bit further and ended up with a 3-piece construction for the whole thing. It’ll have 2 enclosures for the low-end drivers with a center console in the middle housing the middle-high drivers, battery, electronics and user interface. Below a 3D-render of this.


The first enclosure is currently printing at this moment of writing this post.

I had also done some concepting and prototyping for the user interface. I ordered an encoder with an integrated switch, found an unused 4x20 character LCD screen with a blue backlight, ordered a dedicated pushbutton switch w/ a Red LED for the power button and then a 2-position rocker switch to select between 3,5mm audio jack and bluetooth-module for the audio source, this rocker switch also has a Blue LED inside which I plan to use to indicate the state of the bluetooth-module’s connection. I can then build a menu structure for the LCD screen and encoder input, so I can fiddle with the DSP-module’s settings on the fly. If the power button and audio input rocker switch don’t work out so well I can also build the same functionality through the menus on the LCD screen and encoder.


I suppose that is what is currently going on in the driver and enclosure side of the project, so let’s take a closer look into the planned electronics that the middle console will be housing. Of course the electronic design will be much busier due to greatly increased feature set that it will support. We have multiple audio inputs, fully digital audio processing, user interface, USB-C & PD battery charging and much higher audio output powers compared to my last amp board. I will be splitting the electronics into 2 boards, the audio PCB and battery charging + management PCB. The planned battery pack would be a 5S li-Ion or Li-Po pack.

Below is the currently planned block diagram for the audio PCB. I plan on having dedicated amps for the Visatons and then drive the 2 ND65-4s from a single amp in stereo.

Current Kicad schematic overview.

And here’s the currently 97% done layout design for the board. I need to add some connection points for the battery input and speaker outputs. Also this is just a prototyping board and is done on a 2-layer PCB, so that is why I don’t have all the planned amp blocks put on it. It’s a proof of concept board and if it works, then I will likely make another iteration with a 4-layer board and a full output section.

The “TAS5612 class D amp with PWM digital inputs” should be able to pretty easily drive the Visatons, as it’s rated for +200 Watts @ 2 ohm loads and I plan on putting the dual 4 ohm voice coils in parallel for 2 ohm end impedance. Although the biggest power limiting factor will be the battery pack voltage itself, but @ 20 V pack voltage it should be in theory be able to push 100W RMS to the speaker, so I think the first real limiting factor will become the speakers suspension, but that’s why I’m building this preliminary board to find out :wink:

  1. 3,5 mm jack input
  2. Bluetooth-module
  3. LCD-screen connectors
  4. Encoder-connector
  5. Power button + Led -connector
  6. Audio source rocker switch + Led -connector
  7. TAS5612 class D amp
  8. TAS5612 class D amp
  9. STM32F030 Microcontroller
  10. AK5720 24-bit 96 kHz audio AD-Converter
  11. Analog signal switch
  12. 3V3 linear regulator
  13. 5V Buck SMPS
  14. 12V Buck SMPS
  15. STA309 Audio processor

Now I have been pretty budget oriented in choosing all the components and the total for all the ICs on the board come to ~25€ with the 2 TAS5612s taking up 13€ of that sum. Of course there are a lot of other components along with those, but considering these are single - low volume amounts the price is pretty good, no bulk discounts.


I need to get started on the USB charging and BMS board. It’s gonna use a simplified implementation of USB Power Delivery, because I’m not familiar with the protocol yet to make my own software stack for it yet, so I’m gonna use a cheaper MCU and use an interface IC to the USB PD communication.

1 Like

Did you ever get this figured out? Mines doing the same thing. The noise is coming from one of the coils marked 330, and I think because the circuit downstream of it is drawing too much current. Some small component might be shorted. Any luck?

Had forget about this thread

Here a demo and build video I’ve made from my speaker

2 Likes

@Pedrodemio your like the MacGyver of Brazil
image

1 Like