These instruments use a crystal monochromator to select the wavelength (energy) incident on the single crystal sample, and then the scattered neutrons pass through a second, analyser, crystal before being counted by a detector. The latter selects the final energy to be detected, hence energy transfer ω. The single crystal can be oriented, varying Q to measure the scattering S (Q,ω) in the Brillouin Zone. The crystal plane calculations allow scans in constant Q, or ω, and hence measure the dynamics or magnetic interactions in solids.

Although there were a number of instruments in operation around the world at the start of the ILL it seemed to be a matter of pride to re-engineer the concept for the new institute. Mounting components on platforms which had air-pads to enable them to be moved precisely was ideal for such instruments, supporting large masses of shielding protecting the detector from background radiation. At the ILL a second aspect was to use distance to effect beam collimation; this alas took no account of the attenuation of thermal beams in air by about 5% per metre. Together these two principles lead to the creation of the "Tanzboden" or dance-floor instrument, where the analyser and sample tables were separate and (in initial studies) aligned using laser beams. In practice adjustable length steel bars were employed. The elements were positioned sequentially switching air on the the component to be moved, say the sample table, plus those further away from the source (analyser and detector modules). A galley wheel was forced onto the floor by air pressure floor and the assembly moved to the new position. The air pressure for the sample table was turned off, and the next component, the analyser crystal table and detector could then be moved together. Finally the detector was set to the correct angle and all air turned off and the counting started. The procedure was sequenced by the CARINE computer. DC-servo motors were used initially for the floating pads with Precilec encoders switched to a single channel decoder electronics control box. While the Precilec coders were very robust it there was no possibility of simultaneous movement. Small step motors were used for the crystal rotation on the monochromator, sample and analyser tables; again the final positions were checked sequentially with Precilec coders. Already at Harwell in 1968 the PLUTO three-axis spectrometer was floated all assemblies on a steel floor, but each axis had a separate digitiser and motor control unit on CAMAC which the individual PDP8 minicomputer could set moving simultaneously.

Replacement of the CARINE system lead to the introduction of Solar 16/40 computers and motor and counting microprocessor electronics connected via serial communications. Wishing more modern flexibility these were successively replaced by PDP11s, starting with IN20.

IN1 Hot source three axis spectrometer

The monochromator goniometer was mounted on rails and was controlled to send the incident beam through a drum at one side containg a collimator to impinge on the sample. This was on a goniometer in a fixed position. Two analysers were foreseen. These were mounted on a rail and consisted of a hollow drum with the detector within which could be moved about the central analyser crystal. In practice the lack of controllable collimation and innate high backgrounds lead to the rapid change to a more conventional analyser table and detector arm, mounted on air pads.

An additional use of the hot source used a cooled Beryllium filter mounted in front of a detector wide an aperture of +-9 degrees for chemists looking at molecular spectroscopy. The pass band of the filter was about 5meV. A second filter was also built with a combination graphite and beryllium filter with a pass band of 2 meV. These units could replace the analyser and detector modules used in classical three-axis measurements. Finally, with reduced demand for hot source three-axis work, the D4 diffractometer, which originally had a monochromator adjacent to that of IN1, became another component which could time-share the IN1 monochromator after a major rebuild. Here addition of a micro-strip multi-detector lead to improved counting rates justifying reduced beam time availability.

IN2, IN20 thermal beam three axis spectrometer

IN2 was one of the first operational instruments. This instrument had a double monochrometer. The first crystal was flat, the second focussed the beam on the sample. The final direction was always across the large sample table around which the secondary, analyser crystal and detector moved. This allowed very large cryostats and magnets to be used. The nominal distance from reactor to sample was about 23 m hence the final intensity was limited by the distance and initial beam tube size. The analyser was mounted on air pads driven using a galley wheel on the floor. Positioning was sequential. The detector inside massive shielding was mounted on separate air pads. The instrument elements were linked with bars which could be extended allowing both small and very high angles to be reached. The very low backgrounds allowed a large range of measurements to be achieved. An often repeated item on budgets was the request for a second analyser arm to be added to the instrument; probably the expected difficulties of optimising usage of multiple arms lead to this being dropped.

Spin echo experiments had been trialed for extending conventional measurements in the guide hall using D10.
A rebuild of IN2 (becoming IN20, Pynn) was a major project. A new 68 tonne monochromator drum was built to house (several) single monochromators, including heusler alloy for polarised neutrons, which could take over some of the functions of D5. The beam tube of D5 was re-used by D3; for precise polarisation flipping measurements shorter wavelengths could markedly reduce attenuation effects.

IN3 thermal guide three-axis spectrometer

A huge floor was laid for this instrument, which progressively shrank with each practical modification. One innovation was the use of a bent analyser crystal and an array of detectors. This allowed certain measurements of flat branch dispersive modes to be measured well, despite the inevitably lower flux on the guide instrument. The enormous floor was progreesivly re-used for other instruments, S18, IN13 etc.

IN8 thermal beam three-axis spectrometer

This instrument has the highest flux, especially after re-engineering a larger beam tube and developing a sophisticated doubly curved monochromator. It is consequently a world class leading spectrometer.

IN12 cold guide three-axis spectrometer.

Place adjacent to the sample position at D11, this used part of the cold guide including IN11. It was the first three-axis instrument with its own mini-computer (Solar 16/40) with a messaging interface to CAMAC electronics. (Lefebvre, Ledebt). The monochromator shielding could be quite modest since the gamma content of the incident beam was small. Nonetheless it was necessary to introduce lead shielding between this instrument and D11, the SANS camera. The instrument (Stirling), being a second generation design is used for a variety of magnetic experiments and can incorporate polarisation and analysis components in its beam line.

IN14 cold guide three-axis spectrometer

Given the demand for experiments on IN12, when the second cold source was added repacing D3, a new instrument close to the reactor was built using the cold flux to a maximum, though with the beam continuing afterwards into the second guide hall. As for IN12 the main aim was magnetic measurements; there were no experiments nearby which were sensitive to use of the superconducting magnets on the sample table.