next up previous
Next: Call for Observing Proposals Up: IRAM Newsletter 45 (August 2000) Previous: News from the Interferometer

Status of the new IRAM Correlator

The change from five to six antennas in the IRAM interferometer increases the number of simultaneous baselines by 50%, and requires a major upgrade of the correlator. Additional improvements such as eight instead of six spectral subbands, increased spectral bandwidth and full phased-array capability for VLBI have been included in the new design.

The circular support table and central distribution parts have been built. All eight machines have been installed, loaded with their modules and tested (Fig. 2). The analog modules were pre-tested individually before being loaded, but the digital ones, who need the complete correlator environment to work, were directly inserted in their slots.

The first test was to visually inspect the amplitudes and phases on the FFT's of the 15 baselines. This helped to pinpoint a few minor wiring errors which were corrected individually.

A more refined testing method consisted of using redundant configurations of the correlators and let the software compare the results. This allowed to built evidence that no detectable (+/- 1 count) error occurs in any of the 1920 correlator chips of the system. The 4096 steps of the digital delay lines were verified in the same way by purpose-written software.

Some gain dispersion among the 48 analog channels was detected, which limited the usable dynamic range of the system to 9 dB. After a few manual corrections, the range was extended to 13 dB, which is enough to absorb the level variations usually found in the incoming signal.

The 1MHz comb injecting system was installed and successfully delivered the 10 KHz beacon signal for VLBI. The injection level is quite critical and has to be manually controlled. A detailed report on this issue is available on the IRAM web site (http://iram.fr/TA/backend/vlbi/adder/). The adder module design underwent minor modifications and was sent to production.

Endurance testing has not been done so far because of the high indoor temperature of the lab due to Grenoble summer conditions ($25^\circ$C to 29$^\circ $C over the last weeks).

  
Figure 2: Six-Antenna (15 baselines) correlator for the PdBI during the laboratory tests.
\begin{figure}\begin{center}
\mbox{\psfig{file=rondcomp2.ps} }
\end{center}
\end{figure}



Marc TORRES


next up previous
Next: Call for Observing Proposals Up: IRAM Newsletter 45 (August 2000) Previous: News from the Interferometer
bremer@iram.fr