search

dankamongmen / counterforce

leet monitoring tool for schwarzgerät, running on its 5.5" amoled frontpiece
GNU General Public License v3.0
3 stars 1 forks source link

improve ambient noise detection #44

Open dankamongmen opened 1 year ago

dankamongmen commented 1 year ago

we're currently using a KY-038 on an arudino mega directly off the 5V line. at a minimum, use the 3.3 alternative voltage for a cleaner signal. we probably ought use better hardware entirely. also, replace the jumper wires with solid core.

let's move to 3.3V on solid core at least, and see how that improves things. i've got some MAX4466s we could replace the KY-038 with. if those don't provide sufficient improvement, maybe an I2S MEMS is in order.

dankamongmen commented 1 year ago

yeah, i just went through with speaker-test -Dpulse at different volume levels. there was no correspondence i could see to KY-038 voltage levels. yet we do know there is something which leads to reasonableish-looking plots; perhaps it's just responding to noise inside the machine? let me try leaving the doors open while repeating this experiment...

dankamongmen commented 1 year ago

nope...over a good range of noise, i saw:

[schwarzgerat](0) $ mosquitto_sub  -u mora3 -P mora3 -t \#        | grep mic
{"mic":352.4}
{"micmax":363}
{"mic":352.46666666666664}
{"micmax":363}
{"mic":352.46666666666664}
{"micmax":363}
{"mic":352.46666666666664}
{"micmax":363}
{"mic":352.53333333333336}
{"micmax":363}

what's weird is that the averages and the maxen are both so consistent (admittedly, this is pink noise)....

anyway, let's try with the MAX4466.

dankamongmen commented 1 year ago

it does seem to track the PSU fans pretty closely....[shrug]