How?
Are you trained to design and manufacture and test and guarantee the safety of ClassII products?
Are you allowed to impart that training to unskilled Builders?
Big loop area in your input signal pair. Particularly at the RCA connection.
twisted pair would be better, screened twisted pair better still, tiny loop area at the termination the best of all.
The use of the bolt on tag forces the termination into big loop area. Soldering the return to the surround right next to where the hot comes out of the barrel can be much lower loop area, if you ensure the flow and return are closely coupled at the connection.
The speaker pair should also be twisted to minimise loop area.
Trained to design and manufacture, yes, at least in principle. Guarantee definitely not.
I don't think there is any legislation preventing sharing our knowledge, as long as we don't say "if you do it this way, it will be safe". I think we all keep emphasizing the message of "if you don't know what you are doing, don't!".
You were quicker than me. 🙂
Fully agree with the advice - those loop antennas are noise and oscillation waiting to happen.
And funny enough I happen to be typing this at the Tivoli conference center in Copenhagen. 🙂 At least here in the hotel the power outlets have 3 pins.
I also have three pins all over my apartment, as most other danes. The earth pin is just not used. In the wall only two wires are connected to the outlet.
I have tried twisted pair from input RCA to board, but with no audible difference. What is loop area? Is it the "area" between two parallel cables?
What is "bolt on tag"? Soldering the return..... This whole sentence is a mystery to me. What part of the amplifier are you referring to - can you draw it.
And most important, is it safe?
Sorry, but I can´t believe this.
On your picture I can see a green/yellow cable from mains inlet screwed to the chassis, so it is class I, you make me confuse a little bit.
"bolt on tag" = that metal tag with a hole in the middle and bolted under the nut of the RCA socket.
The Signal Return/cold is soldered to the end of the tag some distance away from the Signal Flow/hot.
It's the gap/area between the flow and return wires that forms an aerial and emits or receives interference.
Twisted pair minimises that gap by keeping the flow and return close together. It also has the advantage of swapping the flow and return in each half turn of the twist.
Interference on one half turn is cancelled by the opposite phase interference on the next half turn.
It's the combination of low loop area and phase cancellation that makes twisted pair a very good connection.
The ultimate is screened twisted star quad used for ultra low signals that need maximum attenuation of interference.
I use hand made unscreened star quad for longer interconnects. The slight disadvantage of star quad is a doubling of capacitance compared to twisted pair.
The Signal Return/cold is soldered to the end of the tag some distance away from the Signal Flow/hot.
It's the gap/area between the flow and return wires that forms an aerial and emits or receives interference.
Twisted pair minimises that gap by keeping the flow and return close together. It also has the advantage of swapping the flow and return in each half turn of the twist.
Interference on one half turn is cancelled by the opposite phase interference on the next half turn.
It's the combination of low loop area and phase cancellation that makes twisted pair a very good connection.
The ultimate is screened twisted star quad used for ultra low signals that need maximum attenuation of interference.
I use hand made unscreened star quad for longer interconnects. The slight disadvantage of star quad is a doubling of capacitance compared to twisted pair.
That was just to avoid distortion in the class d picture gallery. If you post a picture of an amplifier without clearly visible safety earth then hell freezes over😀
ok - now I understand. I did try that first both on signal and speaker out. But there is no audible difference, between the short parallel cables, and twisted pairs, so I thought why bother and just made it with parallel cables.
A mobile phone next to, or on top of, a holey chassis is a good test for interference attenuation.
It's not my first Class D. One from CAD audio needed twisted pairs, or else i got radio noise. The modules from audiopower seems to be quite immune, even without a case, its quite but not as quiet as in the cabinet. It's the cabinet that's makes the noise almost dissapear, not twisted pairs
- Status
- Not open for further replies.