Hello everyone, I use the nice TPA3255 module of 3e-audio, along with a 36V power supply (unregulated). Works perfectly and very powerful, but I'm having an issue with hum. That was of course not present before putting things in teh case. When boards inputs are disconnected, there is no hum.
With 10cm unshielded cables (on pic : disconnected) from the dupont crimp connectors to the input pot are connected, there is a noticeable hum. It's 50 Hz, not 100. It varries a lot, eg if I short the inputs no difference, if I completely open the volume with inputs shorted it stops, and if I connect the audio ground to the case, it stops.
Right now, the audio ground is not connected to the case. Should I connect the audio ground to the casing ground ? Itself it is connect to the main ground at the IEC connector.
Hints welcome how to cable this input circuit with the ALPS 100k audio pot, especially in regards to ground cabling.
The PCB under the ALPS has both channel-grounds connected together if that matters. The small boards at top left is a timer that shortens the big brown resistor, the goal is to limit the inrush current.
Thanks for your hints !
Charles
With 10cm unshielded cables (on pic : disconnected) from the dupont crimp connectors to the input pot are connected, there is a noticeable hum. It's 50 Hz, not 100. It varries a lot, eg if I short the inputs no difference, if I completely open the volume with inputs shorted it stops, and if I connect the audio ground to the case, it stops.
Right now, the audio ground is not connected to the case. Should I connect the audio ground to the casing ground ? Itself it is connect to the main ground at the IEC connector.
Hints welcome how to cable this input circuit with the ALPS 100k audio pot, especially in regards to ground cabling.
The PCB under the ALPS has both channel-grounds connected together if that matters. The small boards at top left is a timer that shortens the big brown resistor, the goal is to limit the inrush current.
Thanks for your hints !
Charles
Attachments
I made some progress. First, I replaced the input cables provided by 3e-audio (unshielded) by Mogami shielded wire between the volume pot and the inputs. These Dupont crimp connectors are very difficult and fragile... great care and good tools required. This change slightly reduced the hum, but it's still very present, hence not acceptable.
Then as written yesterday, the only real efficient way to drop this hum is to wire the audio ground to the case ground (which is wired to the mains earth at the IEC socket).
The amplifier will be connected to a bluetooth transmitter, which itself is floating, powered by an USB PSU. So I guess grounding my audio inputs to earth is not a problem and won't cause a ground loop.
I'm aware of ground lifts in PA, but I'm hesitant to re-dismount my back-panel to add a switch. Continuing my investigations...
Then as written yesterday, the only real efficient way to drop this hum is to wire the audio ground to the case ground (which is wired to the mains earth at the IEC socket).
The amplifier will be connected to a bluetooth transmitter, which itself is floating, powered by an USB PSU. So I guess grounding my audio inputs to earth is not a problem and won't cause a ground loop.
I'm aware of ground lifts in PA, but I'm hesitant to re-dismount my back-panel to add a switch. Continuing my investigations...
ok issue found. I _have_ to connect the audio ground to the chassis ground, but not directly. Well explained in this document by Andrew C. Russel (Ground Loops, slide 25).
Hi!
Thanks for this link, bookmarked! I'm a bit confused, slide 25 shows a bridge rectifier? I'd be interested to know exactly what you did 🙂
Thanks for this link, bookmarked! I'm a bit confused, slide 25 shows a bridge rectifier? I'd be interested to know exactly what you did 🙂
Last edited:
I connected the audio ground to the chassis ground using 2 diodes in each direction. A cased rectifier is an easier way than 4 diodes. Schematics attached. The result is to have not more than 1.4V voltage difference between the two grounds, respectively between the audio ground and earth.
To understand how it work, imagine current flowing from one of the ground to the other through 2 diodes. And then the same in the other direction.
My schematics misses capacitors on the PSU, and the N/L phases on the IEC sockets, but you get the point.
To understand how it work, imagine current flowing from one of the ground to the other through 2 diodes. And then the same in the other direction.
My schematics misses capacitors on the PSU, and the N/L phases on the IEC sockets, but you get the point.
Attachments
Last edited:
I understand now, thank you! So are your signal grounds going to chassis now? I will be doing the same soon with the 4x100w 3E board and will be using XLR input.
Cheers.
Cheers.
in fact you can see it as the two grounds are only connected if the voltage difference is > 1.4V (2 diodes in series). In my case, as the amplifier and source are both floating, they were "moving around" freely and every inducted 50 Hz was being injected in the amp through induction I guess.
Disclaimer : I'm not aware of every country's safety ruling so I can't pretend this is something you should use in UK. It's from what I understood the safe way to do it. I have seen other schematics using R & C in parallel. I feel that these large diodes are more robust.
Oh and XLR means symmetrical inputs so you might not have hum, but you still should reference your audio ground in some way. The above one worked for me.
Disclaimer : I'm not aware of every country's safety ruling so I can't pretend this is something you should use in UK. It's from what I understood the safe way to do it. I have seen other schematics using R & C in parallel. I feel that these large diodes are more robust.
Oh and XLR means symmetrical inputs so you might not have hum, but you still should reference your audio ground in some way. The above one worked for me.
Nice one, I thought it was more complicated than it was! I'd be interested in your thoughts re the sound of this amp! 🙂
Listening to the amplifier with my Peerless P830987 Tabaq speakers. The amp has an easy job, is very laid back and can drive these small speakers easily. I tried the amp on larger multi-way speakers (Magnat Quantum 508). Impressive bass, much more than my 20W tube amp was capable of. It seems it has a big power reserve, even if I power the 3e-audio module with 36V, less than the 50V they can eat. So if you want to rock your neighbours this amp will do.... but once you are used to "tube sound", these class D amplifiers are like... how should I say, boring ? sterile ? It's not that they aren't good, but I prefer something warmer.
On the other side, very detailed, now listening to Pink Floyd - The Final Cut. Again they do the job without any issue but without panache. Ok, this album recording quality is not something to be proud of.
On the subject of the power supply : I recovered a big toroidal transformer for very cheap, added a diode bridge and 2 x 10'000 uF capacitors. The secondary has 2x 12V, connected in serie I get 24V AC, so about 36V unregulated. Works fine and never hit the undervoltage "fault". Or maybe I would but I wanna keep good relations to my neigbours so I tend not to push them too hard 😉
On the other side, very detailed, now listening to Pink Floyd - The Final Cut. Again they do the job without any issue but without panache. Ok, this album recording quality is not something to be proud of.
On the subject of the power supply : I recovered a big toroidal transformer for very cheap, added a diode bridge and 2 x 10'000 uF capacitors. The secondary has 2x 12V, connected in serie I get 24V AC, so about 36V unregulated. Works fine and never hit the undervoltage "fault". Or maybe I would but I wanna keep good relations to my neigbours so I tend not to push them too hard 😉