I'm working on an amp project using op-amps feeding into BD139/140.
I'm at the measurement stage and have noticed noise on one of the channels that I'm unsure of the origin.
Looking at the following, we have noise at 50Hz, which is a given, but then the left channel has spikes at 100Hz intervals:
Any pointers on what might cause such a pattern of spikes?
I'm at the measurement stage and have noticed noise on one of the channels that I'm unsure of the origin.
Looking at the following, we have noise at 50Hz, which is a given, but then the left channel has spikes at 100Hz intervals:
Any pointers on what might cause such a pattern of spikes?
"Buzz", as Rayma said. Power line harmonics from rectifiers, or bad shielding, grounding, fluorescent and LED lamps, motors.
Change things to see what makes it worse, or better.
Change things to see what makes it worse, or better.
The case is metal?
One channel means specific to that, so check earthing and shielding first, an open or weak shield can cause this.
If common supply, fault would be on both channels, if the problem was in the supply.
Here the issue is that it is picking up noise signals, could be from the input cables too.
Had that happen, broken shields, changed the co-axial cable, solved.
One channel means specific to that, so check earthing and shielding first, an open or weak shield can cause this.
If common supply, fault would be on both channels, if the problem was in the supply.
Here the issue is that it is picking up noise signals, could be from the input cables too.
Had that happen, broken shields, changed the co-axial cable, solved.
The power supply is a small, filtered SMPS +/-15v and common to both channels.
The amp isn't in a case yet, and the layout and grounding are symmetrical so I dont quite get why it would only be one side and not both.
I swapped cables around and the noise stayed on the left channel.
I'll have a closer look at the grounding on that side and see if I can find anything untoward.
The amp isn't in a case yet, and the layout and grounding are symmetrical so I dont quite get why it would only be one side and not both.
I swapped cables around and the noise stayed on the left channel.
I'll have a closer look at the grounding on that side and see if I can find anything untoward.
Finally managed to spend a little more time with this one.
I had a good look over the board and can't see anything obvious to the LH channel other than a little stray solder ball. Removed this and no change.
I improved the ground connection from the PSU to the amp - previously there were 2 ground lines (+15 & GND, -15 & GND), which is not best practice so I now have a single +15/GND/-15 line.
Tested again and little change.
I had provision for a ground loop breaker on the PSU board so I populated this. No change.
Before and after changes looks like this: (Ignore the ~3.5KHz bump, that is particular to my Focusrite 2i4).
And the latest run split by channel looks like this:
The power supply is a RAC-10-15DK/277 implemented per the datasheet, with a CRC after.
Is everyone of the opinion that this is a 50/100Hz power supply harmonic?
I had a good look over the board and can't see anything obvious to the LH channel other than a little stray solder ball. Removed this and no change.
I improved the ground connection from the PSU to the amp - previously there were 2 ground lines (+15 & GND, -15 & GND), which is not best practice so I now have a single +15/GND/-15 line.
Tested again and little change.
I had provision for a ground loop breaker on the PSU board so I populated this. No change.
Before and after changes looks like this: (Ignore the ~3.5KHz bump, that is particular to my Focusrite 2i4).
And the latest run split by channel looks like this:
The power supply is a RAC-10-15DK/277 implemented per the datasheet, with a CRC after.
Is everyone of the opinion that this is a 50/100Hz power supply harmonic?
You swapped cables and the noise stayed on the left channel, so something in the signal path.
Or a noisy cap in the feedback loop.
Rare case is bad chip.
You have not specified which op amps are being used.
Check that as last resort.
Or a noisy cap in the feedback loop.
Rare case is bad chip.
You have not specified which op amps are being used.
Check that as last resort.
I went back over all the solder joints on the LH channel just to be sure and also made sure that the decoupling capacitors were connected properly by measuring the capacitance of the rails. All good there.
I swapped the NE5532 from side to side and no change.
I added 100uF decoupling on the rails (I had only put the mandatory 100nF close to each opamp). The PSU itself has a filtered output with the bulk capacitance there, but I have added the extra capacitance to the amp board now.
The noise still shows predominantly on the LH channel:
This is the before and after the latest changes:
So it is quieter, but I'm still not satisfied with it.
I am running a DC servo for this amp. Is it possible this might have anything to do with it?
I swapped the NE5532 from side to side and no change.
I added 100uF decoupling on the rails (I had only put the mandatory 100nF close to each opamp). The PSU itself has a filtered output with the bulk capacitance there, but I have added the extra capacitance to the amp board now.
The noise still shows predominantly on the LH channel:
This is the before and after the latest changes:
So it is quieter, but I'm still not satisfied with it.
I am running a DC servo for this amp. Is it possible this might have anything to do with it?
Can you post a schematic for everybody to see?
I have little experience with those, but the experienced members can give their opinion.
I have little experience with those, but the experienced members can give their opinion.
You need to look for differences in the layout and wiring, not for differences in components.
Tom
Tom
