Hi
I'm seeking some advice and any help would be much appreciated.
I have built a simple DAC that I am very happy with. The sound quality is excellent. There is just one thing that could be improved.
It has single-ended outputs that output about 3.2V peak to peak. These are directly coupled from the output stage, which is a differential amplifier (the INA2137) via a 33ohm series resistor (to somewhat protect against accidental shorting of outputs).
My issue arises when I have the DAC plugged into a mixing desk (an older Mackie model). It works fine until I power down the DAC, and then those channels start to pick up a lot of mains hum.
The simple fix is of course to turn down the volume on those channels, but I wonder if this could be fixed by adding a pull-down resistor to my outputs. It's only a short cable run of a few feet, but with no power my DAC outputs are high impedance / floating and that's enough to cause mains coupling into those lines.
I guess it's more common practice to add pull-down resistors on device inputs, but I don't want to mess with the Mackie.
I'm wondering what's typically done in this situation with pre-amps, DACs and the like. If pull-down resistors are added, it would solve this problem, but could affect the efficacy of the output stage, which would now see a lower output impedance when in use than it otherwise would. And I guess, more importantly, the desk, or other gear I plug the DAC into would see a higher input impedance.
I could:
1. Do nothing, and mitigate the hum by leaving the DAC powered on or only open those channels on my mixing desk when I'm using them.
2. Add a pull-down resistor. In which case, what value suits this application, and would I also add a capacitor in-line so I'm only pulling down A/C?
3. Something else I haven't thought about.
Thanks in advance for any suggestions.
Craig
I'm seeking some advice and any help would be much appreciated.
I have built a simple DAC that I am very happy with. The sound quality is excellent. There is just one thing that could be improved.
It has single-ended outputs that output about 3.2V peak to peak. These are directly coupled from the output stage, which is a differential amplifier (the INA2137) via a 33ohm series resistor (to somewhat protect against accidental shorting of outputs).
My issue arises when I have the DAC plugged into a mixing desk (an older Mackie model). It works fine until I power down the DAC, and then those channels start to pick up a lot of mains hum.
The simple fix is of course to turn down the volume on those channels, but I wonder if this could be fixed by adding a pull-down resistor to my outputs. It's only a short cable run of a few feet, but with no power my DAC outputs are high impedance / floating and that's enough to cause mains coupling into those lines.
I guess it's more common practice to add pull-down resistors on device inputs, but I don't want to mess with the Mackie.
I'm wondering what's typically done in this situation with pre-amps, DACs and the like. If pull-down resistors are added, it would solve this problem, but could affect the efficacy of the output stage, which would now see a lower output impedance when in use than it otherwise would. And I guess, more importantly, the desk, or other gear I plug the DAC into would see a higher input impedance.
I could:
1. Do nothing, and mitigate the hum by leaving the DAC powered on or only open those channels on my mixing desk when I'm using them.
2. Add a pull-down resistor. In which case, what value suits this application, and would I also add a capacitor in-line so I'm only pulling down A/C?
3. Something else I haven't thought about.
Thanks in advance for any suggestions.
Craig
Try is the obvious answer... see what a 1k to ground does for it. Adding it before the 33 ohm will mean the level at the user output is unchanged by the extra load but the difference in level change is next to nothing. 1k will only pull a milliamp or two at signal levels and the chip has a capability of around 60ma I think.
I would say no need for a cap as the output should be at zero volts anyway although its worth checking that is the case i.e. that you are not relying on input coupling caps in other equipment.
I would say no need for a cap as the output should be at zero volts anyway although its worth checking that is the case i.e. that you are not relying on input coupling caps in other equipment.
That chip can drive 600 ohms (in fact the distortion seems to be lower at 600 ohm load than 2k load according to the datasheet, so 1k to ground is a great idea!).
The 33 ohm output resistor is isolate the chip from cable capacitance and prevent oscillation, the chip can handle continuous short circuit.
The 33 ohm output resistor is isolate the chip from cable capacitance and prevent oscillation, the chip can handle continuous short circuit.
Thanks for the suggestions. I will try the 1k resistor. There’s no DC on the output so yep, just 2 extra components for L and R. I could do something more elaborate with relays but hopefully not necessary. And you’re right Mark, the 33ohm series resistor is more about cable isolation than preventing shorts.