need help get rid of main hum

Discussion in 'General Electronics Chat' started by bug13, Jun 2, 2016.

  1. bug13

    Thread Starter Well-Known Member

    Feb 13, 2012
    1,208
    38
    Hi guys

    I got this circuit here, it work great if I power the circuit with a 9V battery, but I got main hum if I power it with 12V switch mode power supply. I have a voltage regulator to regulate to voltage to 5V in both case. The balance signal is transmit through a 20m cat6 cable.

    I am using ZSR500 5V regulator, with 1u and 100n cap as close to the input and output.

    How can I get rid off the main hum??

    Thanks guys!

    Capture.PNG
     
  2. shteii01

    AAC Fanatic!

    Feb 19, 2010
    3,397
    497
    Low Pass Filter?
     
  3. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,546
    1,252
    Most switching power supplies require a minimum load to regulate well, and you circuit might not be enough. This can cause excessive noise and power-line artifacts on the supply output. Note that with an input stage gain of over 42 dB, it will not take much hum on the power rail to appear at the output.

    Also, at 3 kHz the chip has less than 20 dB of negative feedback, making the circuit much more susceptible to external noise sources. At 20 kHz the chip has only 40 dB of forward gain, so the circuit is running with no negative feedback at all at the higher audio frequencies.

    ak
     
  4. #12

    Expert

    Nov 30, 2010
    16,343
    6,828
    With that much gain, you better start with a metal box for it. And AK is right. You ran out of gain-bandwidth product. Looks like you need two stages of amplifier.
     
  5. bug13

    Thread Starter Well-Known Member

    Feb 13, 2012
    1,208
    38
    It looks like I don't understand this part of a opamp well enough, can you point me to some reading material, or some keywords that I can google and learn more about this. Thanks.
     
  6. bug13

    Thread Starter Well-Known Member

    Feb 13, 2012
    1,208
    38
    I will try that, thanks #12.
     
  7. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,546
    1,252
    What is the rest of the circuit, and what is the overall circuit function?

    ak
     
  8. bug13

    Thread Starter Well-Known Member

    Feb 13, 2012
    1,208
    38
    There
    Note: this part is powered by a separated switch mode power supply.

    Test with good sound: Tx power with 9V battery, Rx power by a separated 12V switch mode power supply.

    Test with hum: Tx power with 12V switch mode, Rx power by separated 12V switch mode power supply

    Capture.PNG
     
  9. MrAl

    Well-Known Member

    Jun 17, 2014
    2,440
    492
    Hello there,

    When you say "main hum" do you mean the hum from the 'mains' power in your home or office, like 50Hz or 60Hz hum that is often a problem in audio circuits?

    If so then you might look up ways to do this in tube amplifiers, which used to use a potentiometer to inject part of the mains line voltage into the audio circuit (a very small part of course). I dont remember it well enough to post a schematic though. The 'mains' line voltage however does not come from the mains itself, but from a transformer secondary, which in the old days they used one of the filament heater transformers that were already present for the tubes.

    Alternately, a band stop filter, also called a 'notch' filter, works very well sometimes. I once made a 60Hz version that took out most of the 60Hz hum in my circuit, and it was a passive notch filter no op amps, called a "Twin Tee" notch filter.

    If the 'hum' is from the power supply though, then you need to filter the power supply better, probably with one small inductor and at least one nice size filter cap.
     
  10. Sensacell

    Well-Known Member

    Jun 19, 2012
    1,132
    267
    It sounds like your differential amp has poor CMRR. (common mode rejection ratio)

    The ability to reject common-mode voltages (hum, typically) requires a high CMRR differential amp.

    An alternate concept would be a simple audio transformer, these have great CMRR.
     
  11. bug13

    Thread Starter Well-Known Member

    Feb 13, 2012
    1,208
    38
    When power with a 12V switch mode power supply:
    From my scope, there are spikes at the output of R10, while it's reasonably clean at the output of R5. And those noise are in sync with the spike from the power supply.

    When power with 9V battery:
    Output of R5 and R10 is reasonably clean.

    So I guess I need to filter the power supply.

    Then I tried add 100uH inductor in series and 220u, 100n, 10n, 100p in parallel, spike are reduced (as see in a scope), but hum/noise is still very lound (same as no filter before)

    NOTE:
    also tried multi stage gain, noise still there.
     
  12. Sensacell

    Well-Known Member

    Jun 19, 2012
    1,132
    267
    Try connecting the battery AND the Ground from the AC power supply. (leave the + unconnected)
    Power from battery, but still grounded to the same way.

    That will tell you what's up
     
  13. bug13

    Thread Starter Well-Known Member

    Feb 13, 2012
    1,208
    38
    I am home now so I can't test it. What does it tell me if I only connect the AC power supply ground, and it produce noise?
     
  14. bug13

    Thread Starter Well-Known Member

    Feb 13, 2012
    1,208
    38
    Note: I have also added a 1k load resistor in the circuit.
     
  15. Sensacell

    Well-Known Member

    Jun 19, 2012
    1,132
    267
    Tells you that it's common mode ground noise, not noise on the power supply.
     
  16. merts

    New Member

    Apr 1, 2016
    8
    0
    Essentially you have a mic. amplifier which is prone to noise.
    Shield the amp.in a metal case after filtering as suggested.
    Position your supply as far as possible away from the amp.
    This will prevent the transformer from inducing any hum into the circuit.
    Good luck.
     
  17. ian field

    Distinguished Member

    Oct 27, 2012
    4,415
    784
    According to the TS's schematic; the electret mic has no supply - might not need so much gain if it had.
     
  18. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,546
    1,252
    Since both inputs to U4 are derived from a single point (the output of U1), I see no reason for a fully differential amplifier there. All of the signal damage already has been done. If you want to use a differential amplifier to improve signal quality, it has to be at U1.

    And, from post #7, what is the overall circuit function?

    ak
     
  19. bug13

    Thread Starter Well-Known Member

    Feb 13, 2012
    1,208
    38
    Cool thanks for replying and I will test it next week.

    Not that I don't want to shield the amp, but I shielding it might be very difficult in my case, but I will discuss with others and look into the possibility of doing it. Or at least try it out in a metal box next week. Thanks for your input.

    The electret mic has a 5V regulated supply through 2 1k resistors, 100n and 1u decoupling cap between two 1k resistor. Sorry I didn't draw in my circuit(hense the confusion), instead I just use an voltage source to simulate the mic.

    I am not sure how to get an differential signal out from a electric mic, that's why I start with something I can understand.

    I started as single ended signal, it work well if I use a separated twisted pairs. But we want to carry all the signal in a cat6 cable. Along with my audio signal, the cat6 also carry two digital line, 12V power, and another two analog signal. And it picked a lot of noise from the two digital line and two analog lines.

    That's why I move to a differential pairs. My differential Rx can reject all the common mode noise from the other 6 wires in my cat6, only if I power my mic preamp from a 9v batter.

    And we don't want to power our preamp with 9v battery permenantly, that's why I I tested it with 12V plug pack, that leads to the noise problem from my 12 plug pack.

    It's the same project I have been working on (I already have two posts asking about how to select electric mic, and switching between electric mic with relays, which you and others have already helped). The requirements has changed, now we want to be able to run the mic up to 20meters (was 5-10m) with a cat6.

    We have a VoIP module, we want to make an adaptor/add on board, which can provide up to 6 sets of mic, speaker and other control signals. A user at the far can can control which mic and speaker can be active. And only one set of mic and speaker can be active at one time.

    I may have a lot of assumptions here, so if anything is not clear, feel free to ask.

    Thanks.
     
  20. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,546
    1,252
    Only if it is the first thing in the door. Your pseudo-differential signal, made by inverting the single-ended mic signal, inverts both the audio and the noise, so there is no cancellation in the diff amp stage.

    For a 2-pin electret microphone element, consider this: +5V > 1K > mic + terminal > mic - terminal > 1K > GND. The mic element is terminated in 2K, a very common value, and the 2 terminals have a true differential signal (plus some DC). This can drive the diff amp stage U4 directly:
    Delete R5, R6, R7, R8, C3. Keep C4, C7. Change R15 and $17 to 10K. This gets you differential cancellation and some gain.
    Output drives a single resistor-capacitor 2.6 kHz lowpass filter, replacing the effect of R5-R8 and C3.
    Next comes re-using U1 as a single-ended, non-inverting stage. This makes up the rest of your gain. Sharing the system gain across two opamps will greatly improve bandwidth, so you don't run out of gain-bandwidth product.

    ak
     
Loading...