I have an RPI4 (2GB) + DigiAMP running with some CHN50s. I'm building the speaker boxes using 9ply.
The entire lot is powered using a SMPS - a mean well 90W 19V medical class 1 power supply. The SMPS is reasonably clean but I noted that the RPI is raising the noise floor when attached causing a set of peaks across the spectrum (approx 4Hz apart). So I decided to make a LC low pass filter (not entirely designed but simply with the bits that I have around).
The power from the SMPS -> L -> C C C C C -> DigiAMP+ onboard regulator -> RPI4.
The LC is a big Coilcraft 470u 3+Amp shielded inductor (these have been used in 400V power before). I then have two WIMA MKP 1uf and a WIMA 0.22 FKP plus two 25V 1000uF no-name low ESR caps salvaged from a dead SMPS.
Pk-Pk AC noise is now at ~43mV at the speaker. However I have a stronger 14.17kHz peak that rises to about 60dbV. On investigating this doesn't appear at the power supply, which is quiet as a mouse on the input side of the LPF. It also shows on the power point of the DigiAMP+ next to the power jack.
So I'm thinking that the RPI or DigiAMP+ has something onboard that's causing some noise at 14.17kHz. Has anyone experienced this?
Now I could test this without the RPI4 connected, to see if it's the originator but I have a feeling this may be a little more involved. It seems too low at 14kHz to be CPU power saving, so I'm thinking this may be one of the aux/HDMI output?
The entire lot is powered using a SMPS - a mean well 90W 19V medical class 1 power supply. The SMPS is reasonably clean but I noted that the RPI is raising the noise floor when attached causing a set of peaks across the spectrum (approx 4Hz apart). So I decided to make a LC low pass filter (not entirely designed but simply with the bits that I have around).
The power from the SMPS -> L -> C C C C C -> DigiAMP+ onboard regulator -> RPI4.
The LC is a big Coilcraft 470u 3+Amp shielded inductor (these have been used in 400V power before). I then have two WIMA MKP 1uf and a WIMA 0.22 FKP plus two 25V 1000uF no-name low ESR caps salvaged from a dead SMPS.
Pk-Pk AC noise is now at ~43mV at the speaker. However I have a stronger 14.17kHz peak that rises to about 60dbV. On investigating this doesn't appear at the power supply, which is quiet as a mouse on the input side of the LPF. It also shows on the power point of the DigiAMP+ next to the power jack.
So I'm thinking that the RPI or DigiAMP+ has something onboard that's causing some noise at 14.17kHz. Has anyone experienced this?
Now I could test this without the RPI4 connected, to see if it's the originator but I have a feeling this may be a little more involved. It seems too low at 14kHz to be CPU power saving, so I'm thinking this may be one of the aux/HDMI output?
So unplugging the RPI4 from the board - the 14.17kHz spike remains. So this is definitely something on the DigiAMP+ board.
From my understanding the TAS5756m switch frequency is 384kHz. The clock for i2c mode is 400kHz but the serial mode can be 8-48kHz which is about the only TAS5756m clocking that gets close. However it's not used as far as I know unless the firmware doesn't disable it.
From my understanding the TAS5756m switch frequency is 384kHz. The clock for i2c mode is 400kHz but the serial mode can be 8-48kHz which is about the only TAS5756m clocking that gets close. However it's not used as far as I know unless the firmware doesn't disable it.
Ok, so 14.17kHz was coming in through the oscilloscope ground lead. Back to reality. Nothing going on.. it didn't happen. #facepalm.
Now back to the wood working.
Now back to the wood working.
Last edited:
