Well, it might be that the 12AT7 is introducing some hum due to heater-cathode leakage, but that would most likely increase that 1mV reading. The other possibility would something like a cold solder joint on the last filter cap connections creating some extra ripple from the high voltage supply, but again that would probably show on the meter. Did you have speakers connected when you measured the 1mV, and if so could you hear the hum? That might indicate that the meter is not sensitive to higher hum frequencies, i.e. 120Hz and 180Hz.