Salas DCG3 preamp (line & headphone)

Hi,
I have been investigating the issue of harmonic noise in my DCG3.

I have posted below again a picture of the spectrum measured at the output with the preamplifier on and the inputs shorted.
I was able to find the source of the peak at 50Hz, these are the 230VAC coil relay that I use to turn on the amplifier. After bypassing these relays, the 50Hz harmonic disappears completely (unfortunately, I did not take a screenshot ) however, all other harmonics (150Hz, 250Hz, 350 Hz ...) remain unchanged.
I tried changing the position of the transformers, inserting shields (copper and steel) between the transformers and the electronics boards, the harmonics remain the same - without the slightest change.
Does anyone have an idea what could be their cause? maybe the rectifier bridges? - In the bridges the diodes are bypassed with parallel 100nF capacitors.
 

Attachments

  • switched_on_input_shorted_no_relays.png
    switched_on_input_shorted_no_relays.png
    24.2 KB · Views: 19
Can the rest be a feature of the measurement loop?
I've been thinking about that, too. But I'm not sure that the way it behaves supports such a thesis.

With everything plugged in and the preamp off, there is complete 'silence' on the graph. Harmonics appear immediately when the DCG3 is turned on, even before the output relay is triggered, and remain unchanged after it is triggered.
Does this behavior match?

How to possibly check whether these harmonics are related to the measurement circuit or to the preamplifier itself?
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Could be magnetic field interference on the internal signal cables and/or the chassis/terminals. Since the little spikes appear before the signal output relay acts. Try connect/disconnect the chassis with the signal ground? Let's not forget its not easy to nail it because we deal with mainly -120dB below reference spikes here i.e. very very low level harmonic noise to start with.
The other completely harmonic noise free measured examples shown here use either DCSTB or UBiB 1.3 supplies and have the transformers flat on the bottom panel are their main differences to your build. If it has anything to do with what regs. But I would first be suspicious of the transformers different placement and the much mains AC wiring near the signal connectors.
 
  • Like
Reactions: 1 user
Thank you for your input.
I tried connecting and disconnecting the ground to the chassis without any apparent change in harmonic levels. However, I have now tried using other signal cables and got spikes lower by 5 to 8 dB, perhaps, as you write the interference is induced in the cables partly or entirely.
 
  • Like
Reactions: 1 user
diyAudio Chief Moderator
Joined 2002
Paid Member
That's a significant reduction. Congrats. Maybe redressing the normal coax can further refrain it from picking interference. Star-Quad microphone cable is the best against magnetic interference, though in a balanced XLR situation. In SE it loses the essential common mode rejection element. You can nonetheless also try Star-Quad somewhat creatively for SE. One way shielded etc.