Testing For Noise on Summing Buss

[a soBer Newt]

I am working on a Sony MXP that the owner says has noise on the main stereo buss. My thought was to pull all the modules except the master section to eliminate any noise from the channels, there are send returns that feed the stereo buss but these can be muted. My question is with the channels pulled out there is nothing terminating the buss so the ACN would be just picking up noise. Should I tie the ACN input to ground with a resistor, or should I just have a feed resistor? Testing would be done with an Audio Precision 2522.

 

[JohnRoberts]

To test for 'bus' combining amp noise floor it needs to be tested in circuit as it would be normally used.

Mute all the sundry sources feeding the bus.

JR

PS: this complaint often gets resolved with the customer finding a noisy send feeding the bus he was not thinking about.

 

[JohnRoberts]

If you find excessive bus noise, the next step is to check direct outputs from the channels feeding the L/R bus for noise. If they don't have direct outputs, perhaps assign the sources one at a time to some other bus (like an Aux/EFX send).

Lastly check the physical bus node for a stray solder bridge or inadvertent impedance to ground that will increase the noise gain of a virtual earth sum amp. These bus shorts are rare and have variable impact. They can seriously degrade the bus performance, so check sum amp frequency response and THD.

JR

 

[Matt Syson]

Is the 'noise' appearing in both sides? A stereo noisce could be the two physical buses shorted together, Otherwise pretty unlikely unless a source routed to both. Also important to define 'noise' as it seems most recording engineers are strangely deaf when complaining about 'noise'. Is it 50/60/100/120Hz seems to get many stumped.

 

[JohnRoberts]

Is the 'noise' appearing in both sides? A stereo noisce could be the two physical buses shorted together, Otherwise pretty unlikely unless a source routed to both. Also important to define 'noise' as it seems most recording engineers are strangely deaf when complaining about 'noise'. Is it 50/60/100/120Hz seems to get many stumped.

sounds like you've been there seen that.....

JR

 

[porkyc]

MXP3000?
I'm sure there is a "pan mute" switch that removes that monitor path from the stereo mix buss.
Having found said switch, it may be one module that is given gip.
Mute the echo returns as well. (We've all been burnt by that!)
PC

 

[Brian Roth]

I fought with an "odd ball" intermittent on a MCI JH-636. At random intervals (which could be many minutes apart) a "bang" would appear on sends 3/4...which were a stereo pair driving a Lex 224 reverb. Since both buses were being hit equally, that pretty much ruled out the two bus masters having the exact same problem at the exact same time.

In that design, each I/O strip had a buffer after the 5/6 send pot which then fed the 5/6 pan pot. I hit on the idea of hard panning 1-18 inputs hard left to bus 5, 19-36 hard right to bus 6 and waiting/watching 5/6 send outs. Aha! The noise was showing up on only send 5. So, I panned 1-9 inputs hard left, 10-18 hard right. Kept "dividing and conquering" until narrowing it down to one I/O. Replaced the 5534 buffer opamp being fed from the 5/6 send pot on that channel. Desk's owner was thrilled <g>.

Bri

 

[porkyc]

MXP3000?
I'm sure there is a "pan mute" switch that removes that monitor path from the stereo mix buss.
Having found said switch, it may be one module that is given gip.
Mute the echo returns as well. (We've all been burnt by that!)
PC

Also, put all of the EQs into the channel path to start with. Minimize your audio electronics and increase them slowly until it goes rats.
PC

 

[Rob Flinn]

I had noise on the 2 mix of a trident trimix that was very difficult to track down. Eventually I traced it to a poorly soldered electrolytic cap in the path.