The amplifier I built about two years ago has one channel causing crackling sounds when no signal is present. I also noticed crackles when there was a signal. My suspicion is there is a leaking component which is causing the generation of these unwanted sounds. I am posting to ask what tests I should do before I open the amplifier's box to look inside. The amplifier is a large amplifier using four pairs of {2SA1943, 2SC5200} with their emitters connected to the rails albeit through 0.22R wirewound resistors. The collectors are of transistors on each rail as connected together and feed energy to the speaker through a 0.1R resistance. The latter, is connected to ground to provide some negative feedback for the driver stage. When I was in the final stages of this amplifier build I had to add 3.3R resistors to the driver's emitters. The resistors are in series with the 0.1R resistance. Without the 3.3R resistance, both channels clamped onto one rail whenever I tried to adjust the amplifier's quiescent current. The latter was impossible without these resistors.
Please note, although the schematic has DC offset correction circuitry I omitted this in my build.
Please note, although the schematic has DC offset correction circuitry I omitted this in my build.
Crackling sounds are usually due to a rotten contact or to popcorn noise. I'm not aware of any effect that causes transistors that were originally fine to start producing popcorn noise, so I would expect it to be a rotten contact somewhere.
Using junk speakers, short the inputs and poke around inside the amp with an insulated tool
to see whether the noise can be triggered. If not, try freeze spray.
to see whether the noise can be triggered. If not, try freeze spray.
Thanks for your replies.
Looking again at the schematic I found two grounding errors which have a potential to cause this crackling. The cascode's and the current mirrors' potential dividers and grounded to signal ground. This causes a current of several milliamps to flow through the ground separating 10R resistance developing a small voltage across it. Since these potential dividers are connected to the rails, they inject rail noise and disturtbances into the 10R resistance causing the inputs to sense it. According to my reasoning this is the cause of the malfunction. Another side effect of this grounding error is a somewhat high DC offset (10mV, LTSpice's simulations suggest 22mV).
Remembering what I did while building the circuit, I remember I used a PCB to contain the input stage and VAS. This is entirely grounded to signal ground, which corroborates my suspicion the mentioned potential dividers are grounded incorrectly.
Finally, more LTSpice simulations (thanks to Mooly for telling me about it), show the distortion figure is improved if I ground these potential dividers to power ground.
This little extra "surgery" will require the addition of another wire to feed the power ground to the above mentioned PCB.
Nevertheless, during these two last years, this amplifier has been a very nice piece of equipment which offered me quite a good service.
Looking again at the schematic I found two grounding errors which have a potential to cause this crackling. The cascode's and the current mirrors' potential dividers and grounded to signal ground. This causes a current of several milliamps to flow through the ground separating 10R resistance developing a small voltage across it. Since these potential dividers are connected to the rails, they inject rail noise and disturtbances into the 10R resistance causing the inputs to sense it. According to my reasoning this is the cause of the malfunction. Another side effect of this grounding error is a somewhat high DC offset (10mV, LTSpice's simulations suggest 22mV).
Remembering what I did while building the circuit, I remember I used a PCB to contain the input stage and VAS. This is entirely grounded to signal ground, which corroborates my suspicion the mentioned potential dividers are grounded incorrectly.
Finally, more LTSpice simulations (thanks to Mooly for telling me about it), show the distortion figure is improved if I ground these potential dividers to power ground.
This little extra "surgery" will require the addition of another wire to feed the power ground to the above mentioned PCB.
Nevertheless, during these two last years, this amplifier has been a very nice piece of equipment which offered me quite a good service.
As this amplifier separates signal ground from power ground with a 10R resistance, I would like to ask how the non-inverting and inverting inputs should be grounded. My confusion arises from the fact, that the output is grounded at the power ground and the inverting input signal is taken from the voltage across the output. Should the inverting signal be grounded to signal ground or power ground? Simulations show, that when the inverting input is grounded to signal ground, a signal similar in shape and frequency to the output flows in the 10R resistance. This is confusing me.
I am posting a screenshot of the schematic around the input stage.
I am posting a screenshot of the schematic around the input stage.
