---+ Overlapping DAN channels

Motivation

Using DAN at 195 kHz with channels separated by less than 195 kHz may be an option we consider. Here, I show some of the effects associated with DAN channels that overlap in frequency space.

Digitization Noise

The following plot shows the nuller amplitude. Note that the dominant noise source is larger than 1/sqrt(12). Since the only type of noise that could be greater than 1 bit in the standard dongle setup is due to digitization, this shows that the digitization at the accumulator is ~15x larger than the digitization noise due to the FIR.

Fig. 1: Timestream of nuller amplitude.

Noise Shaping

Since the accumulator has an "extra" bit that is not being fed back on, we see some noise shaping. The following plot shows the PSD of Fig 1.

Fig. 2: PSD of nuller amplitude.

Only some of this noise makes it through the feedback loop.

Fig. 3: PSD of residual signal.

-- TijmenDeHaan - 09 Aug 2010

Topic attachments
I Attachment Action Size Date Who Comment
PNGpng dan_digitization_noise.png manage 14.0 K 2010-08-09 - 16:13 TijmenDeHaan  
PNGpng dan_digitization_tod.png manage 13.0 K 2010-08-09 - 16:13 TijmenDeHaan  
PNGpng dan_low_gain_nulling.png manage 12.1 K 2010-08-09 - 16:13 TijmenDeHaan  
PNGpng dan_separation_spike.png manage 10.2 K 2010-08-09 - 16:13 TijmenDeHaan  

This topic: ColdFeedback > WebHome > DanOverlappingChannels Topic revision: r1 - 2010-08-09 - TijmenDeHaan
© 2020 Winterland Cosmology Lab, McGill University, Montréal, Québec, Canada