However, my friend and I were doing some tube rolling. We put a RCA 5692 RED base input tube instead of the 12AU7. My friend notices some Hum Noise even when the pot is at minimum and no audio signal is coming from the DAC. I switch back the input tube to 12AU7. The noise was a lot less and appears only when the step attenuator was at 90% and more. So nothing to say about the 12AU7. By the way, the 5692 needs an adapter 6SN7 to 12AU7 in order to run it. ( I was using an adapter JEMOSA from Amazon)
Checking the datasheet of each input tube.
12AU7 : heaters in parallel : 300 mA at 6.3 Vac
5692 : heaters in series : 600 mA at 6.3 Vac
Is it possible that the bigger amps use on 5692 heaters is affecting the audio signal?
This is speculation, but the higher current draw of the 5692 might not be fully satisfied by the transformer, which is to say the heater may not reach full operating temperature and thus the cathode also may not reach full operating temperature.
Tubes are designed to operate space-charge limited*. operating in this regime allows the cathode to produce the design current/voltage operating characteristics of the tube. However, if the cathode is operated too cold, the tube enters the temperature-limited operating regime, which both changes the current/voltage characteristics, and introduces addition thermal noise in the electron beam. This will probably cause reduced output, increased distortion and increased noise.
This, by the way, is essentially what happens to a tube at end of life. Because its cathode coating that promotes electron emission has become depleted, the tube gradually switches from space-charge limited to temperature-limited operation (commonly called "goes TL"). And so the impact of operating a tube with too low a heater current should produce the same impact as a tube at end-of-life.
This may explain what you are hearing, assuming the tube is being starved for heater current.
*In case you're interested, the space-charge limited condition is where the cathode is hot enough to created a cloud of electrons above the surface of the cathode (the "space charge"), when the cathode is biased to draw current, the current comes from this cloud rather than directly from the cathode surface. As the bias voltage is raised the voltage-current ratio follows a smooth mathematical formula I = P*U^(3/2) (Child–Langmuir law), P is the Perveance, a constant, dependent on the tube geometry. If so much current is drawn from the cathode that the space charge cannot be maintained fully, the some electrons start to be drawn directly from the cathode surface. This transition from space-charge limited to temperature-limited behavior is never uniform across the cathode surface, because there is variation across the surface of it's ability to emit electrons (variation in work function - which is hidden as long as the space charge cloud is fully populated). So as soon last the cathode starts to go TL, the Child–Langmuir law is no longer followed and the tube progressively deviates further and further from design behavior as more and more excess current is drawn. Furthermore, electrons emitted directly from the surface have more, and a broader distribution of, transverse momentum (transverse to the electron beam direction), and that expresses directly as random noise imposed on the audio signal.
