· By Yerko SEPULVEDA
This has been an intense time for testing the individual components and putting them through their paces.
Our DSP software allows us to rapidly add elements to our program that affect the thump intensity output force, response over frequency ranges, headphone volume and EQ, etc.
This is our current DSP schematic:
The original BackBeat G1 had a low-pass filter only in the signal chain. This provides a frequency response between approx. 20-200Hz. It also had a natural frequency resonance around 60-80Hz. The rumble response is significantly more pronounced around these frequencies.
For G2, we've included a compressor in the transducer thump chain. Much like a compressor in an audio signal chain, it will smooth out the thump response over a wide frequency range, so the higher resonance levels will be lower, and the lower will be higher.
EQ setting are also easy with the DSP. We have a low pass filter for the thump response range. We'll also have the future ability (through the app) to change EQ settings. For example, you'll be able to add extra bass in the headphone output if you're using small ear bud headphones.
We're close to finishing up our programming testing which involves:
-verifying component functionality.
-programming bootloader and softdevice onto the main BT chip.
-flashing the firmware.
The firmware is the custom software that will make BackBeat function. You will interface with the firmware through the membrane switch buttons and also eventually though the app.
We have also completed the design and are in the middle of constructing the programming and test fixture for the PCB's. This is a HUGE part of the process since the PCB is the brain of the whole unit and this process will ensure the same quality across all boards.
You may have seen our 3-blade spring design previously. Through our physical testing we could tell there was an issue with the spring around 60Hz. It tended to wobble sideways and not in its intended up-down motion. That wobble caused the magnet to hit the voice coil, making noise and slightly damaging the coil.
We performed a finite element analysis (FEA) of the spring and magnet combination. Sure enough we found a natural modal frequency at 61Hz:
The solution is quite simple and that's to go to a 4-blade design:
The only change we need to make is to add a post in the enclosure to accommodate the 4th blade. FEA runs show there aren't any undesirable modal frequencies in G2's operating range. The shape of the enclosure, the magnet, the voice coil all remain the same.
Thats it for now, watch for our next update in a couple of weeks, we have a lot more cool stuff to report!