How to remove hum sound when no cable plugged in?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have the circuit A as power amp preamp part, XLR input connected to DAC output.

729637d1547706302-remove-hum-sound-cable-plugged-capture-jpg


Curcuit A works well when cables are properly connected with DAC on.
But I aware the following situations.

1. when no cable plugged in, there is no noise and hum sound.
2. when XLR cable plugged in and another end not connected to anything/DAC off, there is some noise and hum sound.
3. when an adapter plugged in (RCA to XLR, cold input connected to GND) without cable, there is some noise and hum sound.
4. when RCA cable plugged with adapter with another end not connected to anything/DAC off, there is huge hum sound.


I would like to eliminate these chances of having hum sound.
I wonder if adding the resistor Ra1 and Ra2 like circuit B will help.

And hoping to know if there are any measures to deal with the hum sound too.

Thank you very much.
 

Attachments

  • Capture.JPG
    Capture.JPG
    122.6 KB · Views: 662
I am now using 240K.
will using smaller resistance help?

That's exactly the point - some amplifiers with XLR input use 300R in those positions. However, you need to make sure your anything/DAC is able to drive such a low impedance.

Example - Accuphase A680. XLR input with 300 Ohm input impedance.

Also, using some pro-grade balanced receivers help a lot (very high CMRR).

Example - THAT 1206.
 
i may be mistaken but the op-amps connected to pin 3 should be using the inverting inputs, no?

Not necessarily - depends on the overall topology of the converter ;)
What is important - input impedance and common mode rejection rate (CMRR).
This kind of simple converter normally provides CMRR not too high because of certain resistors value tolerance. Special receivers with laser trimming, like THAT chips, provide very high CMRR.
 
I would guess the problem is twofold:

1. Differential mode input impedance is excessive at close to 480 kOhms - that'll make a first-rate antenna with some unterminated cabling connected.

To solve this while keeping common-mode input impedance high, I suggest a T-type input resistor topology:
Lift R1-R2 junction from GNDS, and insert a third resistor between them and GNDS. Use one of your 240Ks for this new resistor and something around 10k for R1/R2.
The result: Common-mode input impedance has actually increased more than twofold, while differential mode input impedance now is down to a far more reasonable ~20 kOhms, which still remains easy to drive.

The same approach can be used with RF suppression capacitors to limit the effect of mismatch on high-frequency CMRR. 470p+470p + 47p could be used after some 1k in series, for example, though I would also recommend 2x 22-47p directly at the input to chassis due to unavoidable series inductance.

2. Clearly, CMRR leaves something to be desired. Hand-match R3+R4 and R5+R6, respectively.

3. There probably is another problem, possibly a ground routing issue. I wouldn't think it's the pin 1 routing, however I'll emphasize that XLR pin 1 always goes to chassis ground (as directly as possible), not to local signal ground. There may be more lurking where that came from.

4. What is the gain of this circuit? 0 dB? If so, note that balanced connections are often run hotter than unbalanced ones, as typical receivers are somewhat noisy, not to mention the added robustness this provides. Compare maximum DAC output level and power amp input sensitivity - you may find a substantial mismatch. You may want to modify your receiver to provide -6 dB or -12 dB of gain, or attenuate the signal of your present receiver afterwards for that matter (which may give better input noise).
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.