The recent F/V conversion with an ADC question got me to thinking about an IRIG-B decoder about which I once speculated. For those not familiar with IRIG-B, the demodulated signal is 1 KHz with groups of five cycles at a peak voltage to indicate a binary one and five cycles of 1/2 peak voltage to indicate a binary zero. It seemed as if almost any ADC could be used to have an adaptive decoder that could handle variations in the peak voltage. The only complication I could see at the time was the difficulty of detecting the peaks without some sort of zero-crossing detector.
Would it be feasible to sample at say 4 KHz and adjust the time window until the sampling is synched to the signal? You'd slide the window until getting near zero readings on alternate samples?