Interesting hum problem

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I am building a per-amp based on A-15 Liteaudio DIY kit and their relay input selector board. I placed everything in a steel box that is grounded to the mains but there is no signal ground connection to the box. In other words - two transformers (one for the relay borad, other for the pre-amp board) are creating virtual grounds that are interconnected but they are not connected to the box (ie mains earth). Signal is passing through the relays and a 100K pot to the preamp board and then to the output connectors. The pot is a motorized Alps.

The problem is: without any input and with my headphone amp (battery powered) connected to the output I hear very bad hum noise.

My qyestion is - where is the hum coming from since there are no obvious ground loops (at least not obvious to me).

Am I doing something wrong and will the hum go away if I connect the preamp to a properly grounded power amp?
You need to connect the box to signal ground, otherwise any voltage present on the box will be capacitively coupled to the circuit. Connected to mains ground it will certainly have undesired voltages on it. Sure fire recipe for hum. Sort of like using shielded cables with the shields floating or connected to mains ground. You may need some sort of low value resistor, diodes, thermistor, or other isolator, for best performance. I'll usually sacrifice a few dB and go with a nice safe direct connection.
What about letting the box to float (no earth connection) and connnecting the signal ground to the box. Would that be any better?

In other words - what is the quietest solution for the unbalanced output in this arrangement? I know that the decoupled signal ground is the safest (I am using that in my power amp).
Connect 0V (ie your "star" ground) to saftey earth through a loop breaker - see here for good info. A pair of diodes in inverse parallel will do instead of a bridge - 1N5408 should be fine.

You could probably get away with just the resistor and capacitor too - the resistor would still let enough fault current flow to blow a fuse, and more than enough to trip an RCD/GFCI.
Thanks for that - I already new about that trick since I used it on my power amp.

I had to do it again since obviously my "floating ground idea" did not really work. Still a lot to be discovered.

It is working fine know - I am not sure what to expect when I connect it to the power amp.

I ll see.

Thanks again.

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.