The MINT Backend PrincetonCMB

  Main
  · Science
  · Instrument
    · Backend
      · Channelizer
      · Correlator
         · Design
         · Enclosure
         · Fabrication
         » Performance
    · Receiver
    · Optics
    · Data Acquisition
    · Local Oscillator
    · Base
    · Site
  · Analysis
  · Results
  · Publications
  · Team

The Correlator Performance

There is, of course, no performance details on the correlator because it is not yet built. The prototype board consisted of just the digitizer unit using basically the same chips. Unfortunately, a synchronization error allowed only half of the digitizer unit to work at one time, effectively limiting the speed of the digitizer to 500 MHz sampling or 250 MHz bandwidth. The digitizer responded, nonetheless, correctly to sine wave and noise input. We were also able to connect the correlator chip to two digitizers and produce a correlation function. For a complete description, see Toby's thesis.

In support of the correlator design, we have characterized some of the rf components, namely the filter and the attenuator.


pdf xls

The critical specification on the attenuator is the change in phase for each attenuator setting. Measurements on an HP Vector Network Analyzer show that the phase change is less than 3 degrees at 500 MHz for all setting except the highest, and the highest is less than 10 degrees.


pdf xls

The bandpass of the anti-alias filter cuts off at an acceptable 500-600 MHz, with a sharp falloff.


pdf xls

A zoom out of the bandpass shows that for frequencies up to 18 GHz, the rejection is at least 20 dB out of band. The critical rejection is at 4.5 and 5.5 GHz, the location of the LO's for the channelizer.