On April 3, a major upgrade was started on Plateau de Bure, which included a complete change of the LO distribution system, continuum detectors, and the replacement of 2 ``old'' 3-mm SIS receivers by dual-frequency systems covering the the 3mm and 1.3mm bands on antennas 1 and 2. These improvements had major consequences on the control software, since the new hardware (receiver and LO system) is entirely controlled by VME micro-processors linked to the central computer by Ethernet.
After less than 2 weeks of intense activity, the array was back into operation at 3mm and the first 1.3mm tests were performed. First fringes at 230 GHz were actually obtained on April 12, at 12:15 on 3C84, between antenna 1 and 4 (Figure 2), but effectively ``seen'' only a week later because low signal to noise prevented immediate identification and other urgent tests needed to be performed.
Figure: First 1.3mm fringes on April 12. An uncorrected 10ns delay
offset causes the strong frequency dependence of the phases.
Figure: 1.3mm observed on April 18, from 3C279, on a 160m baseline:
Amplitude and phase as a function of time.
On April 18, just before a bad weather period, fringes were again obtained
at 230 GHz between antenna 1 and 4 on several sources: 3C273, 3C279, NRAO530
and 3C345. Higher signal to noise allowed immediate checks of pointing and
focus differences between the 3 mm and 1.3 mm receiver. The baseline length
was 160 m, which corresponds to the highest spatial resolution ever used at
Plateau de Bure. Because of the poor weather conditions and long baselines,
phase stability was bad (even at 3mm), but there is an excellent correlation
between the 3mm and 1.3 mm phases and even between the phases and the
total power output of the 1.3 mm receivers (See ``Phase correction'' section,
page 
).
The new receivers have typical receiver noise temperatures of about 45 K in the 3mm band (with a rejection of 6 dB of the image sideband), and 50 K in the 1.3mm band (210 to 250 GHz).
Final software modifications were carried out between April 23 and April 26, and the Plateau de Bure interferometer was fully back into normal operation on April 26.
