Hello guys,
New to this forum.
Looking for help in getting rid of a terrible buzz from my miniDSP-nanoSharc-based digital preamp build (or what to call it). I use USB/optical input to the nanoSharc board and outputs go directly from D/A to my class D amps.
I first had a simple, OK working layout/wiring, attached as layout_old.png (I know, made in Paint; not an electrical engineer…). Most Components were purchased from Audiophonics so should be suitable for audio applications, I guess. There was always a very faint hiss/buzz but not really a problem when connected to my speaker system. The A/D is a very simple breakout board with only one ground terminal, and it was not in practical use in this case but it was always powered and communicating with nanoSharc via I2S.
Then I installed a RIAA card from ebay (to hook up my turntable via the A/D) and all hell broke loose. Buzz/hum/hiss no matter what I did. In my attemps to get rid of this I made a new wiring, attached as layout11_new.png. That made the problem less prominent but it is still a big problem. The noise sounds mostly like an ugly buzz, not a humming sinus.
My theories behind the changes I made were the following:
1) Make all DC return currents go back to the source, primarily from DC-DC regulators feeding the A/D.
2) Make analogue signal returns go back to the source.
3) Separate digital and analogue ground as far as possible, but combine them in one point in the middle and connect that point to the chassi.
4) Remove electrical connection between nanoSharc and chassi, to avoid a ground loop.
5) Break ground in I2S wiring to both A/D and D/A, also to avoid ground loops.
6) Connect rear and bottom chassi with a ground wire to have entire chassi at the same potential.
7) Make AC and DC Cables twisted pair (didn't make a difference).
I am now hoping for some tips on what to try next? I think the buzz has always been there in the first layout (without RIAA card) but perhaps so heavily reduced that I didn't notice it so much. Any thoughts from anyone? The big difference I can think of is that in the first version, the cards were joined with ground cables in the I2S wiring only, and the connection to the chassi was at the nanoSharc and to the rear chassi only, with a bad connection to the rest of the chassi where the power components are placed (?).
PS: There's no difference when connected to USB, optical or Power amps compared to not connected to anything. I listen through headphones through 3,5mm jack on D/A. DS
New to this forum.
Looking for help in getting rid of a terrible buzz from my miniDSP-nanoSharc-based digital preamp build (or what to call it). I use USB/optical input to the nanoSharc board and outputs go directly from D/A to my class D amps.I first had a simple, OK working layout/wiring, attached as layout_old.png (I know, made in Paint; not an electrical engineer…). Most Components were purchased from Audiophonics so should be suitable for audio applications, I guess. There was always a very faint hiss/buzz but not really a problem when connected to my speaker system. The A/D is a very simple breakout board with only one ground terminal, and it was not in practical use in this case but it was always powered and communicating with nanoSharc via I2S.
Then I installed a RIAA card from ebay (to hook up my turntable via the A/D) and all hell broke loose. Buzz/hum/hiss no matter what I did. In my attemps to get rid of this I made a new wiring, attached as layout11_new.png. That made the problem less prominent but it is still a big problem. The noise sounds mostly like an ugly buzz, not a humming sinus.
My theories behind the changes I made were the following:
1) Make all DC return currents go back to the source, primarily from DC-DC regulators feeding the A/D.
2) Make analogue signal returns go back to the source.
3) Separate digital and analogue ground as far as possible, but combine them in one point in the middle and connect that point to the chassi.
4) Remove electrical connection between nanoSharc and chassi, to avoid a ground loop.
5) Break ground in I2S wiring to both A/D and D/A, also to avoid ground loops.
6) Connect rear and bottom chassi with a ground wire to have entire chassi at the same potential.
7) Make AC and DC Cables twisted pair (didn't make a difference).
I am now hoping for some tips on what to try next? I think the buzz has always been there in the first layout (without RIAA card) but perhaps so heavily reduced that I didn't notice it so much. Any thoughts from anyone? The big difference I can think of is that in the first version, the cards were joined with ground cables in the I2S wiring only, and the connection to the chassi was at the nanoSharc and to the rear chassi only, with a bad connection to the rest of the chassi where the power components are placed (?).
PS: There's no difference when connected to USB, optical or Power amps compared to not connected to anything. I listen through headphones through 3,5mm jack on D/A. DS
Attachments
One thing I came to think of: could I have wired up the DAC board in the wrong way? Please see attached picture as part of the spec for the board. This is the type of transformer I have:
Transformer R-CORE 30VA 2x15V + 2X9V
Puissance 30VA
Secondaire : 2x15V / 0.5A (15-0-15V)
Secondaire : 2x9V / 0.85A (9-0-9V)
Audiophonics support claimed I should use 15V-0-15V for the analogue part and 9V-0 for the digital part of the board. This generic DAC board is also sold on Ebay, and that page recommends the following transformer:
30W R-CORE Transformer
Primary: 0-110V
0-110V
Secondary: 0-9V 0.4A
0-9V 0.4A
16V- 0-16V 0.6A
This feels contradictory. The Ebay recommended transformer should give 18V AC over the two terminals when used with 110V AC mains. At present, I measure 9V AC over the two terminals, but should I in fact have 18V AC?
Transformer R-CORE 30VA 2x15V + 2X9V
Puissance 30VA
Secondaire : 2x15V / 0.5A (15-0-15V)
Secondaire : 2x9V / 0.85A (9-0-9V)
Audiophonics support claimed I should use 15V-0-15V for the analogue part and 9V-0 for the digital part of the board. This generic DAC board is also sold on Ebay, and that page recommends the following transformer:
30W R-CORE Transformer
Primary: 0-110V
0-110V
Secondary: 0-9V 0.4A
0-9V 0.4A
16V- 0-16V 0.6A
This feels contradictory. The Ebay recommended transformer should give 18V AC over the two terminals when used with 110V AC mains. At present, I measure 9V AC over the two terminals, but should I in fact have 18V AC?
Attachments
Thank you for your answer, Duke! I'll try and do this and maybe also try holding some alu foil sheets between the modules and see if they interfere with each other.
One thing I haven't dared try yet is grounding the center tap for the digital part of the DAC board; I don't know if this could break anything? The transformer typically recommended for this board doesn't even have a center tap for the 9V part so this will probably not improve things... At present, there seems to be a rather unstable ~5V DC difference between the center taps of the 9V and 15V windings but perhaps this is normal and to be expected considering the equipment powered by it (?).
I've read about using a "loop breaker circuit" to reduce buzz, consisting of a resistor, capacitor and two diods. Where should this be connected in my case? Between the center tap of the 15V winding and ground?
One thing I haven't dared try yet is grounding the center tap for the digital part of the DAC board; I don't know if this could break anything? The transformer typically recommended for this board doesn't even have a center tap for the 9V part so this will probably not improve things... At present, there seems to be a rather unstable ~5V DC difference between the center taps of the 9V and 15V windings but perhaps this is normal and to be expected considering the equipment powered by it (?).
I've read about using a "loop breaker circuit" to reduce buzz, consisting of a resistor, capacitor and two diods. Where should this be connected in my case? Between the center tap of the 15V winding and ground?
Looking for some more tips and advice. I have twisted the primary and secondary wires. Is it normal to feel something similiar to an electric shock when touching an earth ground point and coming in contact with the insulaton of the twisted primaries? :-o There's nothing wrong with the actual wires; I have used mains voltage installation wires, FK 1,5mm2. Does the twisting generate some electromagnetic field in my fingers in this case? :-o It's difficult for me (as a newbie) to understand how that could be good thing (?) in reducing hum. They are routed in a straight line through the air inside the box, not alongside a chassi wall. Should they be? Or even routed inside a braid connected to the chassi?
Best regards
/Martin
They are routed in a straight line through the air inside the box, not alongside a chassi wall. Should they be? Or even routed inside a braid connected to the chassi?Best regards
/Martin
Which two terminals? If it is a center-tapped 18vac winding (actually looks like it might be 30vac RMS, center tapped. That what AC15v-0-AC15v would seem to imply, as seen on the transformer label in your diagram), then it should measure 18vac RMS from leg to leg (or from the label, 30vac RMS leg to leg - under full load, it may measure a little higher under no load). If not, you might check to see if the transformer primary could be strapped wrong. If there is only one way to connect the primary and the no load secondary voltage is only half of what it should be, then that would start to look like they may have sent the wrong part. Is there any color coding on the primary leads? (Again from the diagram, it looks like the primary may be a 230vac, center tapped winding, with red being the center tap connection. If so, the primary resistance should measure about half from red to black, as compared to measuring from blue to black.)
Last edited:
The first thing I notice is you have signal and power grounds connected to the chassis, this is most critical in RIAA preamp as the input signal is mV's and gain will be around 40dB.
I would isolate the signal grounds from the chassis and power grounds.
Also cable routing in the enclosure will be important if you have switching PS's in the same enclosure. Switching PS's can be incredibly noisy especially cheaper ones that run typically at lower switching frequencies with little or no filtering.
Make sure your power cabling is well away from the signal cables and IF they must cross do so at 90°, do not run them parallel.
I would isolate the signal grounds from the chassis and power grounds.
Also cable routing in the enclosure will be important if you have switching PS's in the same enclosure. Switching PS's can be incredibly noisy especially cheaper ones that run typically at lower switching frequencies with little or no filtering.
Make sure your power cabling is well away from the signal cables and IF they must cross do so at 90°, do not run them parallel.
You should feel nothing if you touch the primary mains wire insulation and chassis, I would investigate this and fix it as it is potentially dangerous. It's also not a good idea to be touching mains wiring in the equipment without disconnecting the power.
Twisting wires was invented by Alexander Graham Bell it reduces electromagnetic interference. It's commonly called twisted pair or UTP.
It was used by the telephone companies for decades until the recent laying of fibre. Here in Aus there is still loads of UTP in the ground for communications.
You need to twist all the unshielded wiring to make it most effective.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.