ILL - Instrument control
Early days - CARINE and NICOLE
Initially two systems were developed for data gathering.
Crystal monochromator instruments with step-scanning acquisition modes were controlled by CAMAC electronics shared across two Telemecanique T2000 systems CARINE, (IN1, IN2, IN3, D1A, D1B, D2, D4, D5, D6, D8). A single program was active at any one time, others being swapped out to disk. Each instrument had an ASR33 Teletype as keyboard, and had an "APPEL" button to request attention from the controlling system.
Often the scientists would perform offline counts and movements rather than laboriously program alignment sequences. Overall performance lead to poor overall performance with much deadtime between actual counting periods. The basic simplicity led to reliability.
For multidetector arrays, event data were transferred from CAMAC to a Telefunken TR86 system NICOLE. This served D11, D7, IN4, IN5, and PN1. Each instrument had a keyboard, a CRT display, and a control console to show the multichannel data as cuts or surfaces, with knobs and buttons to select scales and modes. The computer system monitored all these devices to respond to experimenters' orders. Data were updated on disk, with the exception of D11, where data were held in memory. The latter were lost when system failure required a system restart. In practice the D11 experimenters often worked in shifts, so were always present to restart the system, and measurements rarely exceeded 40 minutes. As instruments were progressively transferred to more modern replacements and the complex environment simplified the overall reliability improved.
Mid 1970s - Individual PDP11 on new instruments
One of the ILL's PDP11 devoted to data gathering
2023 A. Barthélemy's archives
John Allibon installing a removable disk cartridge RK05 on the DEC PDP11/40 of the diffractometer D8
©1982 ILL
Development of new instruments, IN10, D17 etc lead to adoption of individual PDP11 computers connected to standard CAMAC electronics. A project to replace CARINE and NICOLE led to the introduction of SEMS Solar16/40 individual computers for the 3-axis instruments and D1B with new electronics, and PDP11s for the other instruments post 1978.
Period 1979 to 1991
A PDP-11/23
©http://gunkies.org
EOLE, the VAX 11/750 for the data processing of D19
©ILL, A. Barthélemy's archives
The principal interface between the control computers and electronics was through a CAMAC crate controller which was also linked to the UNIBUS (or through a Quniverter to Q-bus) of the DEC computers. This bidirectional highway could send CAMAC commands and receive data.
As the PDP11, with its general purpose UNIBUS, evolved into a single board computer (PDP11/23_LSI-11 using a cheaper multiplexed Q-bus) it became much cheaper and it was offered for sale OEM by Plessey Ltd, a British company. It could be obtained without too much political bargaining and progressively replaced instrument computers, mostly running the RSX11M system, though RT11 and TSX (a multi-users version of RT11 were also used).
By the mid 80’s 32-bit computers were dropping in price and thus a transition occured for both the hardware and software. This is described in page "Introduction of 32-bit computers".
A VAX11/750 was bought to manage data for the multi-detector single crystal diffractometer D19, which was controlled by a PDP11/24 (a UNIBUS variant of the LSI/11). VAX11/730 computers were installed to control D11 (1985) and D17 (1986) directly without a front-end PDP11. The advent of the microVAX-II single board computer with a Q22-bus lead to price falls and it became very attractive to replace the PDP11/34 bought in 1980, by these more powerful systems.
Apart from deliberately slowing response times in the software for faster cpus, the low level VMS software written for D11 was used unchanged on all subsequent VMS instrument computers.
The network was developed to include ethernet to all instruments.