I am a really big fan of the modular concept of framework in general and especially of the FW16. Additionally the open information access to mechanical and electrical data makes it easy to develop hardware. When the first reviews went online, I saw the LTT video and heard the comment the the speakers had room for improvement. Once I got my FW16 I thought the speakers are fine, but why not give the the modularity a try to give the FW16 exceptional audio performance - and the idea was born to add front facing audio “input” modules.
The idea is simple: find a speaker that fits the size constraints and drive it with an appropriate DAC and amplifier. For the speaker I just searched for the largest inductive speaker on digikey at the time, that would fit the chassis. The CS38-02P50-02-1 would fit easily and with a rated output power of 2W and 3W peak it uses all the available height for some serious air pressure. As a DAC/amp combo the MAX98357A was chosen. Its main advantages are the simplicity of only having one chip that receives digital I2S signals, it being able to run directly with 5V from the notebook and it not needing many additional components on the PCB. It is also well documented by Adafruit. Now the only thing missing was a carrier PCB and a USB to I2S bridge.
This last task was made very easy by framework due to the availability of KiCAD files on their GitHub, which I am very thankful to have. I just had to add the DAC/amp chip and the speaker to the already existing input module example, as the example already includes a RP2040 which can be used as a USB audio bridge.
After ordering the components, assembling the PCB and getting the pico-playground audio DAC example running, I was able to successfully test the PCB. I was both internal and the additional speaker at the same time, and as expected, the high frequencies (including voices) were clearer with the new speaker. The lower notes are still better to be left to the original speakers, as the new PCB lacks the needed mechanical coupling for them and starts to rattle in the chassis below approx. 400Hz. But this should be easy to handle by adding appropriate dampening and software equalizer to steer the frequencies. The current idea is to let the built in speakers handle the lows to mid frequencies, which they are well capable of and add the new speaker for the mid to high frequencies, which have a large benefit from the front facing direction.
This the current state of the project and now I am curios what you guys think.
Do you like the idea? Do you have any improvements?
Things that are still in progress:
- improved synchronization of speakers
- specific EQ curves for internal and additional speakers
- improved mechanical coupling to prevent rattle
