Ekco AI Mk 20 radar

overscan (PaulMM)

Staff member
Dec 27, 2005
Reaction score
AI Mark 20 (Green Willow)

Chris Poole - July 2007

Red Steer (ARI-5919) evolved from AI Mark 20 - code named Green Willow which was the designation given to the EKCO fire control radar developed as a back up to the AI Mark 23 radar being designed by Ferranti for the English Electric Lightning (then known as the P1), which was under development in the early 1950's.

English Electric P1
Photograph of English Electric P1 at RAF Henlow (unknown Internet source)

The contact to begin development of AI Mark 20 was awarded in late 1953 and was for a limited number of sets (probably no more than 5) and was almost certainly given to the company because of the existing work going on with the 'Blue Sky' (Fireflash) radar system where to a certain extent the performance parameter were similar although AI 20 was to be physically larger.

The granting of the development contract by TRE was as a result of problems Ferranti were having getting the Ai Mark 23 (Air pass radar) configured and working, which to be fair the them was as much due to changing operating requirements by the RAF, one of which was the decision to adopt the 'Firestreak' infrared homing missile and not helped by the fact that the whole fighter contact was running late (almost a British tradition??).

The design was to be unique since the radar had to sit in an enclosed housing, which formed the 'shock-cone' inside the nose of the aircraft used to slow the air entering the engines down.

The AI-20 team was headed up by Cyril Drew supported by Bill Graville, Ron Beaven (CRT unit), Henry Cox, Mike Skinner, John Radlett, Ted Smith and Ian Walker (the personnel involved changing periodically).

The strobe unit subsystem was done by John Yarrow (assisted by Henry Cox) and his team from 'top lab' where they were working on project 'Blue Sky'.

Most power supplies, certainly the Transformers and wound products were designed by Ian Walker and John Clark and some additional re-design work was possibly done by Norman Jerrum (ex TV and Radio designer) at Southend.

Gibby and his team including Eric Alden were responsible for the mechanical design.

Norman Wall and his team were responsible for The 'IF' unit.

Hugh Green and his team, which included Harry Robinson and Tommy Morgan (although there was at least one other), headed the waveguide design team.

Ray Reeves designed the Klystron control unit.

The Radar Labs draughtsmen were under the leadership of section leader Jack Timms, together with Horace Fowler, Jim Bye and a couple of others. For secrecy, this D.O. section was in a separate partitioned off part of the main D.O. and due to the secrecy, non-participating Radar folks were not welcomed! Note Gibby had his own draftsmen assigned to him, one of which was Stan Seager, later joined by Len Newborough.

Cyril and his team were in ensconced in 'her ladyships bedroom' since this room had the space to house the chassis of the equipment, which was mounted on a special rotatable jig, which in turn was on casters since this equipments was physically larger than any unit made before and was technically challenging to say the least. This laboratory also had an advantage of having large 'French style window doors, which opened up a large area such that the transmitter could be powered up pointing outside over the top of the canteen into free space although usually this power was put into a dummy load.

AI-20 was to be a high-powered 'X' band system of about 100Kw output using a using a 5C22 hydrogen thyratron modulator. Interestingly the operating frequency and the Pulse Repetition rate remain secret to this day (2007) due to its later incarnation as 'Red Steer'.

Much new ground was broken in the design and development of AI-20 not least of which was the use of honeycomb structure in the main chassis (to save weight) but also in the waveguide run where part of the waveguide run was machined as two halves out of solid aluminium and then mated together. This entailed a very high degree of workmanship due to the very tight tolerances.

'Red Steer' waveguide
Detail of 'Red Steer' waveguide run showing complexity of arrangement
Photograph courtesy Steve Milnthorpe.

Another requirement was that the whole of the radar system had to be pressurised due to the high altitude at which the aircraft was expected to operate (in excess of 50,000 ft) and here a fibreglass shroud with inflatable seals was developed.

Pressurised Shroud
Photograph of the pressurised shroud - courtesy of Maurice Wedd

The indicator unit designed by Ron Beaven also presented many challenges since this unit was the first design to use a high brightness radar display - several times brighter than any previous display. (This 'high brightness display was needed due to the high ambient light levels at high altitude) To solve this problem, a 2 1/2" diameter 25Kv projection CRT tube was selected and scaled up into a 5" diameter 30Kv direct view tube.

Red Steer MK1 indicator unit, very similar to AI-20 except for additional control functions

In the era before transistors, there was no feasible way to generate 30Kv inside the un-pressurised indicator. The solution the design team used was to peak rectify the magnetron pulse in the pressurised main unit and pipe 30Kv DC through aircraft cables to the indicator. This of course meant that the team had to design their own sealed high voltage cable and plug/socket assemblies. Surprisingly while some reliability problems were expected as well as having 30Kv flowing through the aircraft structure, the system worked like a charm and no reliability problems or failures occurred.

In terms of the scanning head, AI-20 was high speed Spiral Scan and needed a lot of power to drive it (the main motor was a 10,000 RPM 1/2 HP BTH machine). This presented many engineering challenges due to the complexity of having a scanner mechanism rotating and 'nodding' at some 1,00 RPM. To overcome the risk of this unit shaking the rest of the system to bits, special 'Barry mounts' were developed. Unfortunately no scanner head is known to exist nor has it been possible to find a photo.

The scanning head took 18 revolutions to spiral out from 'dead ahead' to an arc of 45 degrees and a further 18 revolutions to spiral back thus performing a full scan pattern in about 2 ¼ seconds

Another ground-breaking development was an AI-20 intercept simulator intended to produce a synthetic intercept sequence. This development was undertaken in a separate laboratory (top lab) with the assistance of Ray Reeves on a part time basis. This was to prove very helpful in reducing the overall development time.

By 1955, flight trials were underway from RAF Defford (the home base of RAE Malvern) and during the trials the system performed impeccably, however only two or three AI-20 models were made since by this time AI-23 was also beginning trials so the program was halted and put on ice.

While no published details are available of the system performance, it is known that there was a 95% probability of a 'lock-on' of a 'hunter' sized aircraft at 7 miles.

In 1956, a requirement was issued for a 'tail warning radar' (TWR) for the trio of RAF V Bombers (Valiant, Victor and Vulcan) and it was realised by RAE (promoted by one of the EKCO liaison officers - Jerry Steer) that AI-20 was a well-engineered small radar, which, with limited modifications would be an ideal basis for the tail warning radar for the V bombers. This was selected and after some re-engineering became 'Red Steer' (ARI-5919).

Archived from http://www.ekco-radar.co.uk/poole/ai20.php

Similar threads