I put together an amp a few months ago, and noticed there was a bit of hum through the speakers in the midrange when powered on. Other amps I've built before had no hum, so I assumed the large amount of bias current in this amp was the culprit. Not sure why I thought this in retrospect. Recently, I hooked up another amp with much lower bias and there it was again, the same hum. This time, I had to figure the issue was the power supply, since it was the common component in both of the hum producing amps. I used a spectro on my phone to find the major spike was at 180Hz, another lower spike at 240Hz, and some more multiples of 60Hz trailing off from there. Sounds like classic ground loop (I think?). I'm just not sure what's producing it.
I connected just the power supply to mains and hooked it up to a scope. No chassis, just had the boards on my bench. When Earth ground is connected to GND out the smoothed side of the capacitor bank, I can see the 180Hz spike on my scope AC coupled with the probe on +, shield on GND:
When the mains ground is disconnected, no spike:
This behavior persists regardless of whether or not live and neutral are connected.
Just to sanity check myself, GND should be connected to Earth ground so that fault current has a safe path, right?
Here's the soft start and rectifier board schematic:
It's connected to the capacitor bank via the Positive-Common-Negative terminals:
The capacitors are Mlytic 4-poles. H1-4 are case screws. I connected one of the screws to Earth ground via an alligator clip for my test.
Would really appreciate any help tracking down the source of the hum.
I connected just the power supply to mains and hooked it up to a scope. No chassis, just had the boards on my bench. When Earth ground is connected to GND out the smoothed side of the capacitor bank, I can see the 180Hz spike on my scope AC coupled with the probe on +, shield on GND:
When the mains ground is disconnected, no spike:
This behavior persists regardless of whether or not live and neutral are connected.
Just to sanity check myself, GND should be connected to Earth ground so that fault current has a safe path, right?
Here's the soft start and rectifier board schematic:
It's connected to the capacitor bank via the Positive-Common-Negative terminals:
The capacitors are Mlytic 4-poles. H1-4 are case screws. I connected one of the screws to Earth ground via an alligator clip for my test.
Would really appreciate any help tracking down the source of the hum.
I don't know what changed, but the spikes are gone. The spectro is flat now powered and unpowered. All I did was flip the boards and transformer over for a few seconds.
Thanks, I'll give that a try. Whats' the reasoning for this? I thought these snubber caps were standard practice.
bucks bunny said:
Thanks, I'll give that a try. Whats' the reasoning for this? I thought these snubber caps were standard practice.
bucks bunny said:
That's so strange, I took the basic power supply schematic from Doug Self's PA book and didn't think much of the safety caps. Never would have thought they were the cause of the "noise". But indeed they were. I removed them and the amp is silent when powered on with no input signal. Thanks much for your help.