buzz in micpre

[salomonander]

Hey there,
a client brought me his micpre that has an obvious buzz going on. Regardless of what mic is used - or a dummy resistor.
i replaced all electrolytic and tantalum caps but no luck. I rewired the psu and twisted all ac lines. No luck.
By chance i found a solution, but im not certain that its legit. There is a big cap (right after the rectifier) in the psu that is connected by two wires to the pcb. If i simply take those two wires apart - the buzz is gone. As soon as they get close to each other, the buzz starts again. Obviously i could now lay the cables in a way that they do not touch. But im still wondering if this is a legit solution and im baffled on how this ran clean in the first place.

any ideas?

thanks!

 

[Khron]

Thicker wires, and twisted together? And ideally as short as feasible...

 

[salomonander]

Thanks Khron,
i just tried thicker wires (1.5mm - enough?) and it did improve. It doesnt like twisting thought. I still have to keep them a little apart. Its better than before though….

 

[Khron]

What shape is that big electrolytic in, though? Does it still test in spec for capacitance, and is its ESR within sane values? Is the rectifier bridge still fully operational (all four diodes ok)? And is the buzzing only in the audio signal, or mechanical, or both?

 

[salomonander]

I will have to test the rectifier. But the psu gives the correct voltage.
The big cap i replaced today - as all other caps. Everything brand new in that department. The buzz is only in the audio signal. Nothing mechanical. Turning the transformer also changes nothing.

 

[salomonander]

Ps: the big cap you can see on the photo right next to pcb and transformer.

i cant really test the rectifier since the pcb holes are so small that its impossible to desolder anything without breaking….

 

[Khron]

You don't need to desolder it; you can use the diode-test mode on a multimeter to test each of the four diodes in there, in-situ.

 

[salomonander]

The rectifier tests fine

 

[salomonander]

Ok i tried around more and im lost honestly.
By separating the wires it does get better. But its not gone.
Found out two things by accident:

- the buzz is gone almost when lifting the psu pcb a little and simply holding it in the air…..

- there are three LM395T. When connecting one of my multimeter probes to the pcb mass and putting my other probe on the left leg of any of these, the buzz is magically gone. Multimeter is in DC voltage mode….

 

[Khron]

Maybe it's time to hard-wire a 1meg resistor between that LM395 pin and ground, if it cures the issue, and call it a day..?

 

[Audio1Man]

You are fighting a losing battle. The CORRECT WIRING is to charge the capacitor 2 twisted wires from the bridge rectifier to the cap terminals, and then 2 twisted wires to the discharge load. NEVER SHARE THE CHARGING & DISCHARGING WIRES.

Duke

 

[transient]

Have you put a scope on it to check for RF? Bear in mind that the capacitance and/or resistance of your scope probe may kill an oscillation if it's there. I see what looks like film capacitors spread over the board, but perhaps they are the wrong type and/or in the wrong place. There could be a problem with the board layout or a manufacturing fault too.

 

[salomonander]

You are fighting a losing battle. The CORRECT WIRING is to charge the capacitor 2 twisted wires from the bridge rectifier to the cap terminals, and then 2 twisted wires to the discharge load. NEVER SHARE THE CHARGING & DISCHARGING WIRES.

Duke

thanks duke.
just to make clear that i understand you correctly.…
instead of using a single cable to the plus and single cable to the minus terminal of the capacitor, i should use two per pole that i twist (separately per pair)? if so, what awg is apropriate?

i did try adding the resistor but i only had a 1/4 watt and it did litte… could the low rating be the issue?

 

[salomonander]

Have you put a scope on it to check for RF? Bear in mind that the capacitance and/or resistance of your scope probe may kill an oscillation if it's there. I see what looks like film capacitors spread over the board, but perhaps they are the wrong type and/or in the wrong place. There could be a problem with the board layout or a manufacturing fault too.

I dont trust the design at all. Nor the build. It was a prototype from some gear shop. The star ground did not even have connection to the chassis a complete mess. So there is a good chance that circuit is messy too. But im honestly not engineer enough to fix a flawed design…
The client so far used the pre with a condenser mic. No problems. Its with his dynamics and quiet sources that the buzz becomes obvious.

 

[RoadrunnerOZ]

From the photo your DC cables to the capacitor run right across the edge face of the toroidal transformer and may be picking up interference from that. When you pull the wires apart you are maybe just moving them away from the transformer.

 

[salomonander]

From the photo your DC cables to the capacitor run right across the edge face of the toroidal transformer and may be picking up interference from that. When you pull the wires apart you are maybe just moving them away from the transformer.

Thanks! I did try rotating the transformer - no luck

 

[RoadrunnerOZ]

Try re-routing the wires to the capacitor - you may need to replace them with longer wires and keep them away from the transformer but if you release them from the saddle mount and tape them away from the transformer you can see if that works. Rotating the transformer is not going to help, only moving it away, but that’s the hard option - easier to replace the wires than move the transformer - it’s just too close to those wires.

 

[salomonander]

Try re-routing the wires to the capacitor - you may need to replace them with longer wires and keep them away from the transformer but if you release them from the saddle mount and tape them away from the transformer you can see if that works. Rotating the transformer is not going to help, only moving it away, but that’s the hard option - easier to replace the wires than move the transformer - it’s just too close to those wires.

Thanks, i did release the capacitor. But i can still hear the noise no matter where i place it and regardless on where the wires are.
I will try the two wires per pole next.
im starting to wonder if the psu simply is not well thought out. It might be the phantom power that messes with things. When i disconnect the 48v outputs from the pcb it gets better.

 

[RoadrunnerOZ]

Thanks, i did release the capacitor. But i can still hear the noise no matter where i place it and regardless on where the wires are.
I will try the two wires per pole next.
im starting to wonder if the psu simply is not well thought out. It might be the phantom power that messes with things. When i disconnect the 48v outputs from the pcb it gets better.

In your initial post you indicated that separating the capacitor wires got rid of the noise??
Edit: Also are the tops of any of the three big capacitors swollen? They should be flat. Or are they ones that you replaced? Did you replace the one on the flying leads?

 

[JohnRoberts]

Thanks, i did release the capacitor. But i can still hear the noise no matter where i place it and regardless on where the wires are.
I will try the two wires per pole next.
im starting to wonder if the psu simply is not well thought out. It might be the phantom power that messes with things. When i disconnect the 48v outputs from the pcb it gets better.

Managing hum in low noise designs can be layout related. If you can get it quiet with alternate lead dress, do what works.

JR