|
Post by flacman on Apr 13, 2018 13:08:36 GMT
Volume pot causes 10-20mV noise at the output. But I do not understand why the noise has a frequency of 66Hz? How does 50Hz noise get into the pot volume? And how does it become 66Hz? ps: I'm using a simple regulator power supply. Darlington bjt + zener. I see 0mV ripple of my power supply. My amp's gain is 7. -- ps: Noise is loud when I turn volume up. When turning the volume down, the noise disappears. It seems that the 0R output impedance of the volume pot has killed this 66Hz noise.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Apr 13, 2018 17:04:46 GMT
Are you using a single supply voltage ?
Did you decouple the opamp close to both power pins ?
|
|
|
Post by flacman on Apr 14, 2018 4:27:39 GMT
Are you using a single supply voltage ? Did you decouple the opamp close to both power pins ? Im using dual supply. Yes, i have decoupled it, by 2200u// 100u // 0.1u. The center point is connected to ground (star point). I connected the ground with 200x200mm chassis, the noise was decreased a bit. Then, I touch my finger at ground (my feet are on the floor) the noise drops to near 0m. Is there still "earth ground" shortage? I test with Sensheiser Momentum, it is quite sensitive to mV. I do not want to connect earth ground (this is inconvenient) and also do not want to use 100ohm output resistor devider. Any other way to kill that noise or "virtual earth/chassis ground" circuit? ps: And I still do not understand how the 66Hz (or 50Hz) noise can get in. And how the 0R output impedance of the volume pot can kill it. -- sorry for my bad English.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Apr 14, 2018 8:24:24 GMT
Did you connect the metal outside part of the potmeter to signal ground ? When the volpot is set to 0 do you still get noise (hum ?).
Are the 0.1uF decoupling caps wires as short as possible and running directly from the PS pins of the opamp to signal ground or directly from pin V+ to pin V- ?
When you connect the scope to a 50Hz source (output of a mains transformer) does the waveform calculate as 50Hz ?
The scope signal suggests the scope power supply itself also contributes. Is the scope grounded ?
|
|
|
Post by flacman on Apr 14, 2018 13:09:26 GMT
Did you connect the metal outside part of the potmeter to signal ground ? When the volpot is set to 0 do you still get noise (hum ?). Are the 0.1uF decoupling caps wires as short as possible and running directly from the PS pins of the opamp to signal ground or directly from pin V+ to pin V- ? When you connect the scope to a 50Hz source (output of a mains transformer) does the waveform calculate as 50Hz ? The scope signal suggests the scope power supply itself also contributes. Is the scope grounded ? When the volpot is set to 0, the noise (hum) disapeared But, looks like I'm mistaken, the volume pot is not related to this issue. It seems the magnetic field noise/hum has gone into line in cable. After the cable is shielded, the hum decreases dramatically. But it seems that the Earth grounding actually eliminates it. (Why ??) I still do not want to add the Earth grounding wire.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Apr 14, 2018 14:28:53 GMT
Line level signals must always be shielded. Also when a metal enclosure is used with a transformer inside it is important to connect the metal enclosure to the signal ground. Preferably from the star ground in your case. The RCA's grounds must be isolated from the enclosur. When the star ground and input RCA ground is connected via the enclosure the magnetic field outside of the transformer will induce a current in that loop which causes an audible voltage.
|
|