There's some RF energy eminating from the INPUT of the DC-DC converter -- I posted on this 2 years ago -- you can eliminate it by using a small value choke on the input lead -- if you're using Pete's board you could replace the fuse with a choke.
I'm also outlined possibility of noise injection from Murata DC-DC converter in one of my posts (can't verify that in real life because I'm abroad away from home).
Did you used ferrite or permalloy choke? Approximately what value?
Last edited:
So, the recommendation is to replace fuse F1 with a 12mH choke? Does Pete know about this?
Hmmmm, I don't know if I did extensive testing when selecting the choke -- 12mH was just something I pulled out. i have one of these Murata DC-DC converters in the preamp stage of a Pioneer SX-838 which I rehabilitated.
12mH is a lot! I think that lowest DCR and lowest inductance that does the job is all that's required. I don't know how to figure that out though.
I'd like Pete's opinion before I go modifying his circuit. (as if I had one of those power supplies lying around anyway...)
Me too, Pete? Do you think this is the source of my noise?
much appreciated,
Andrew
After all, I killed all gremlins which haunted my box.
As I wrote before in one of posts, I had oscillation (or injected noise, hard to state clearly), strange tiny peaks in frequency response curve, and absolutely unsatisfactory THD+N even in loopback mode (input and output connected together on 33K resistor terminals).
I don't know which part of the tweaks actually solved problem(s), but they are dirty cheap anyway.
1) I occasionally burned U1 DRV134PA, and replaced it with THAT 1646P08-U.
2) Added the following ceramic capacitors in order to suppress noise:
1uF - 5V supply (soldered straight on USB socket PCB terminals);
0.1uF - parallel to C22, C7, C8 (near Murata switch mode DC-DC converter);
1uF - parallel to C13, C18, C11.
0.1uF - high quality ceramic NP0 type
1uF - ceramic Z5U type
Choke would be the last option if anything above doesn't work, DC 500mA 10-12mH is quite big, it doesn't fit nicely.
4 terminal chokes (from many switch mode PS) are common mode types, they are not suitable for this application.
Now in loopback mode I've got the following (1.9V from ESI Juli, balanced connections, 192000 mode, 191072 FFT):
1 KHz: THD - 0.0075%, THD+N - 0.082%
20 Hz: THD - 0.041%, THD+N - 0.19%
20 KHz: THD - 0.048%, THD+N - 0.094%
SMPTE IMD 250Hz/8KHz - 0.035%, MD2 - 0.034%, MD3 - 0.0038%
50 Hz noise: -104dB
100 Hz noise: -120dB
150 Hz noise: -120dB
Next step will be building foil shield inside plastic box in order to get rid of annoying 50/100 Hz noise. Its not that bad, but could be better.
As I wrote before in one of posts, I had oscillation (or injected noise, hard to state clearly), strange tiny peaks in frequency response curve, and absolutely unsatisfactory THD+N even in loopback mode (input and output connected together on 33K resistor terminals).
I don't know which part of the tweaks actually solved problem(s), but they are dirty cheap anyway.
1) I occasionally burned U1 DRV134PA, and replaced it with THAT 1646P08-U.
2) Added the following ceramic capacitors in order to suppress noise:
1uF - 5V supply (soldered straight on USB socket PCB terminals);
0.1uF - parallel to C22, C7, C8 (near Murata switch mode DC-DC converter);
1uF - parallel to C13, C18, C11.
0.1uF - high quality ceramic NP0 type
1uF - ceramic Z5U type
Choke would be the last option if anything above doesn't work, DC 500mA 10-12mH is quite big, it doesn't fit nicely.
4 terminal chokes (from many switch mode PS) are common mode types, they are not suitable for this application.
Now in loopback mode I've got the following (1.9V from ESI Juli, balanced connections, 192000 mode, 191072 FFT):
1 KHz: THD - 0.0075%, THD+N - 0.082%
20 Hz: THD - 0.041%, THD+N - 0.19%
20 KHz: THD - 0.048%, THD+N - 0.094%
SMPTE IMD 250Hz/8KHz - 0.035%, MD2 - 0.034%, MD3 - 0.0038%
50 Hz noise: -104dB
100 Hz noise: -120dB
150 Hz noise: -120dB
Next step will be building foil shield inside plastic box in order to get rid of annoying 50/100 Hz noise. Its not that bad, but could be better.