I received a 6 channel headphone amp from a friend of a friend asking me to take a look at it as a couple of channels have hum on them. I figured it wasn’t a power supply thing since only couple channels had the hum on them.
I took a listen and found that channels 1-3 were quiet. Then channel 4 had a very light hum, 5 a bit louder and 6 the loudest. I felt it wasn’t a coincidence that as you get closer to the power transformer the louder the hum gets. Being these three.
It has a single center tapped secondary winding and each side of the winding goes into a 1N4001 and their cathodes are tied together providing a single rail of 30.8V. So at the cathode of one of the diodes I tapped in with one of my power supplies.
With my power supply running the amp it functioned perfectly and not a hint of hum on any of the channels, dead quiet. So this told me it had to be the power transformer. There is a steel sheet barrier, but I used copper tape help seal any opening and create a better shield. First wrapping the power side in kapton for safety.
This really hasn’t helped at all, maybe slightly reduce the hum in channel 4 which was pretty low already, but 5 and 6 are still pretty loud.
Can anything be done?
Dan
I took a listen and found that channels 1-3 were quiet. Then channel 4 had a very light hum, 5 a bit louder and 6 the loudest. I felt it wasn’t a coincidence that as you get closer to the power transformer the louder the hum gets. Being these three.
It has a single center tapped secondary winding and each side of the winding goes into a 1N4001 and their cathodes are tied together providing a single rail of 30.8V. So at the cathode of one of the diodes I tapped in with one of my power supplies.
With my power supply running the amp it functioned perfectly and not a hint of hum on any of the channels, dead quiet. So this told me it had to be the power transformer. There is a steel sheet barrier, but I used copper tape help seal any opening and create a better shield. First wrapping the power side in kapton for safety.
This really hasn’t helped at all, maybe slightly reduce the hum in channel 4 which was pretty low already, but 5 and 6 are still pretty loud.
Can anything be done?
Dan
You need ferrous metal to block a magnetic field. For chuckles and grins, remove the power transformer and set it outside the case. See if that drops the hum.
I had an electronic crossover with hum. I modified the power supply and also bought a new Hammond transformer encased in steel. Mounted that on the rear panel so it looks nice, the leads come in under the transformer. Cured all supply noise as good as using external regulated power supplies.
I had an electronic crossover with hum. I modified the power supply and also bought a new Hammond transformer encased in steel. Mounted that on the rear panel so it looks nice, the leads come in under the transformer. Cured all supply noise as good as using external regulated power supplies.
Maybe dirty or corroded grounds on the connectors? If magnetic field problems then waving a piece of steel plate around may have some effect on it. Copper likely not so much. Its more for electrostatic shielding. That said, some transformers have copper bands to help cancel flux.
The copper band is a shorted turn. It blocks electrostatic noise pretty well.
I have found poor grounds causing trouble with chassis rivets and poor solder connections to lugs. That would typically affect all channels as audio common is hopefully isolated from chassis ground by 10 to 100 ohms. You an induce noise by making your own direct chassis connections.
Sometimes a steel plate is very effective at blocking flux, other times it isn't. That's holding it and placing it between the noise source and victim circuit. Grounding that plate and seeing the difference tells you more about what is going one (try each way).
I have found poor grounds causing trouble with chassis rivets and poor solder connections to lugs. That would typically affect all channels as audio common is hopefully isolated from chassis ground by 10 to 100 ohms. You an induce noise by making your own direct chassis connections.
Sometimes a steel plate is very effective at blocking flux, other times it isn't. That's holding it and placing it between the noise source and victim circuit. Grounding that plate and seeing the difference tells you more about what is going one (try each way).
I'm looking at that PCB and those channels are near not just the transformer, but also the rectifiers, filter capacitors and regulators. What does this hum sound like? Can you record it and post a short .wav file? Is it 60Hz, 120Hz, more of a buzz with lots of harmonics? What does the output look like on a scope trace?
I'm trying to thing what "magnetically" could happen to the transformer that would cause it to affect only the channels near it, and I'm coming up empty. I'd check the rectifier diodes and filter caps, and touch up all the solder joints in that area.
I'm trying to thing what "magnetically" could happen to the transformer that would cause it to affect only the channels near it, and I'm coming up empty. I'd check the rectifier diodes and filter caps, and touch up all the solder joints in that area.
Hi Ben,
Depending on component leads and traces, a hum can be induced. Same mechanism as distortion and loop area in an amplifier.
Depending on component leads and traces, a hum can be induced. Same mechanism as distortion and loop area in an amplifier.
The transformer CT should go directly to the bulk supply caps. Any access to that path except at the caps will create a ground hum voltage. If the transformer CT is soldered to a ground plain, then that is your problem. You may have to re-rout that pin with a jumper to the caps.
the progressively increased hum level by the channel location/distance to the PSU hints the pulsating rectifier current is rippling out through the GND system on the PCB, suggesting failure of the filtering caps following the rectifier diodes, as @nigelwright7557 suspected. When the rectifier filtering caps have a severe capacitance loss, the linear regulators can starve and drop out of regulation momentarily during the 120Hz cycle, significant pulsating current can emerge on the GND circuit on far side of the linear regulators, by charging/discharging the regulators output capacitors. If the PCB design did not have such scenario accounted for, this pulsating GND current can back-drive the headphones and make them hum.
Well, no. The filter capacitors are probably fine. The charging current in good filter capacitors is what you need to be careful with.
Very easily checked with an oscilloscope, and a basic first test when chasing this kind of fault.
Very easily checked with an oscilloscope, and a basic first test when chasing this kind of fault.
I think I have come to a point where I am going to stop on this. I couldn’t find anything on a scope, I’m not sure if it’s currently AC coupled or DC coupled right now, but I could not see anything on the humming channels versus the non-humming. I use my meter to check for any type of AC ripple, and I found the differences between any of the channels. I don’t get that. I removed the main filter caps and they measure perfectly fine. Each of the voltage regulators also have additional capacitance to help smooth their 12 V supply and they all measure perfect. I did remove the transformer from the chassis and moved it back as far as way as I could and it did indeed reduce the amount of hum.
I removed and checked capacitors throughout the audio portion of the two channels, and nothing significant was found. The entire amplifier runs on 12 V, the 30.8 V immediately goes into three separate regulators, all 7812s. So theoretically, I could just install a barrel jack and this could be run off of a 18 V or 24 V computer supply or wall wart. It draws very little current, it is a 20 W consumption device. The center tap of the transformer does go to a ground plane, but that ground plane is also connected to the bulk capacitors, at least the negative side. This has been very frustrating, and I would like to ultimately fix it, but it’s just not worth it right now. I’m not getting anything for my time and I would likely be able to solve it faster by installing the barrel jack. If anyone else has suggestions, let me know, maybe an idea of why I can’t see anything on my scope???
Dan
I removed and checked capacitors throughout the audio portion of the two channels, and nothing significant was found. The entire amplifier runs on 12 V, the 30.8 V immediately goes into three separate regulators, all 7812s. So theoretically, I could just install a barrel jack and this could be run off of a 18 V or 24 V computer supply or wall wart. It draws very little current, it is a 20 W consumption device. The center tap of the transformer does go to a ground plane, but that ground plane is also connected to the bulk capacitors, at least the negative side. This has been very frustrating, and I would like to ultimately fix it, but it’s just not worth it right now. I’m not getting anything for my time and I would likely be able to solve it faster by installing the barrel jack. If anyone else has suggestions, let me know, maybe an idea of why I can’t see anything on my scope???
Dan
The problem maybe PCB layout. Connect a lab PS and if the HUM goes away, the next is to verify CROSSTALK as the PCB layout & TX field is causing the problem.
Post the schematic.
Duke
Post the schematic.
Duke
Hi Duke, you may not have seen my post above, but in my first post, the third photo down shows it hooked up to a lab power supply. And I did also mention that with the lab power supply, there was no noise.
I can’t imagine they would’ve released a device like this that had audible noise in two channels, so something had to have changed. As for a schematic, I’ve had no luck tracking one down.
Dan
Hi Dan
I have seen many products where they have difference between channels.
1. connect with a lab PS , verify NO HUM
2. Turn on the AC Power and see if HUM is present? (Note tx output is not connected) This checks for FIELD induced to PCB.
contact me duke.aguiar@ieee.org
I have seen many products where they have difference between channels.
1. connect with a lab PS , verify NO HUM
2. Turn on the AC Power and see if HUM is present? (Note tx output is not connected) This checks for FIELD induced to PCB.
contact me duke.aguiar@ieee.org
Last edited: