weltyj
New Member
I have a Fosi BT20A amp, plus Boston Acoustics CR65 speakers for a space I will be working out in. Lots of other stuff in the room so I'm not looking for perfection.
When I connect the headphone output of the radio (FM) to the line input of the BT20A, I get horrible noise, generated by the switching (and harmonics I suspect) from the BT20A that are present on the line input jack traveling through the audio cable, creating radio frequency EMI for the radio's front end.
So I made a little PI network filter, and breadboarded it up. As untidy as it is (see picture) -- seems to work great if I get a good antenna signal to the radio.
I have a good spec sheet for the inductors I have, so I could put in the appropriate parasitic capacitance etc. for modeling.
I have no spec sheet for the junkbox of capacitors I have, so I just used some values from spec sheets for other polyester caps.
I had to make some other assumptions:
1) Impedance of the line out on the BT20A is 10k ohms
2) Impedance of the headphone input is 1k ohms
3) Switching frequency of the BT20A is 200kHz or higher (I have no clue what it really is).
I wrote a perl script to evaluate all possible combinations of inductors and capacitors I have. Goal was to have flat dB curve thru the audio range and get ~ 30dB attenuation at 200kHz, and at least 50 dB attenuation in the FM frequency range.
For curiosity, I also modeled what would happen if a line-level output from a radio was attached through this filter (the red line). Takes a pretty big hit on the 20 kHz end of the audio frequency content going back to the amp in that case. I intend to use headphone output nearly all the time anyway so I'm not overly concerned.
Before I get out the soldering iron -- what mistakes did i make? I am a total novice at this so I really appreciate any corrections or words of wisdom.
TIA,
Jeff
When I connect the headphone output of the radio (FM) to the line input of the BT20A, I get horrible noise, generated by the switching (and harmonics I suspect) from the BT20A that are present on the line input jack traveling through the audio cable, creating radio frequency EMI for the radio's front end.
So I made a little PI network filter, and breadboarded it up. As untidy as it is (see picture) -- seems to work great if I get a good antenna signal to the radio.
I have a good spec sheet for the inductors I have, so I could put in the appropriate parasitic capacitance etc. for modeling.
I have no spec sheet for the junkbox of capacitors I have, so I just used some values from spec sheets for other polyester caps.
I had to make some other assumptions:
1) Impedance of the line out on the BT20A is 10k ohms
2) Impedance of the headphone input is 1k ohms
3) Switching frequency of the BT20A is 200kHz or higher (I have no clue what it really is).
I wrote a perl script to evaluate all possible combinations of inductors and capacitors I have. Goal was to have flat dB curve thru the audio range and get ~ 30dB attenuation at 200kHz, and at least 50 dB attenuation in the FM frequency range.
For curiosity, I also modeled what would happen if a line-level output from a radio was attached through this filter (the red line). Takes a pretty big hit on the 20 kHz end of the audio frequency content going back to the amp in that case. I intend to use headphone output nearly all the time anyway so I'm not overly concerned.
Before I get out the soldering iron -- what mistakes did i make? I am a total novice at this so I really appreciate any corrections or words of wisdom.
TIA,
Jeff
