大象传媒

During this period at Derby Racecourse Gunsite in 1941 George Johnston, by theis time a Lance Sergeant RA)with others on the site did experimental work with the Sperry Predictor, and unwittingly wrote the "Drill for Enemy Engagement Unseen with Radar" which was the standard drill for night engagement of bombers for the rest of the war. George also designed apparatus for use with it.

((When using radar control at night, you had a plotting board which was a grid with the gunsite at the centre and you plotted the targets across the grid, and predicted where they would be from these various plots. When plotting every 10 secs, considering the course of the plane, and future prediction etc. you had to produce predictions within 7 secs, as there was a 3,2,1 count. In 7 secs the plotter had to decide where it was and what was happening to the plot. (When they were experimenting with the Modified Sperry Predictor, George et al also developed a semi-automatic version, in which information was fed in, and a travelling light on a screen was used, not to predict the target position, but the rate of change of target course in relation to Eastings and Northings on a rate factor which could be set into the
Modified Sperry Predictor. But George found their own group could plot as quickly manually as with the semi-automatic system, but this was not true generally: George et al were able to do it because they had specialist gunnery knowledge. Later all the plotting was done by ATS, who did not have the gunnery knowledge needed.)

The Original Sperry Predictor was a cunning device which was not intended initially for Ack Ack at all, but for fast targets at sea. But by a modification it could also deal with targets within an arc. (One of the problems with ballistics, apart from the theory in vacuo, is that different atmospheric pressures, wind speeds and directions all affect the flight of the missile. George had to study this applied physics at length on every gunnery course, but particularly the instructor's course at Manorbier. The target seen from a given point is pivoting on that point, but not only pivoting, its position in relation to that point is going so far East, so far North, i.e. resolving into Eastings and Northings from a given point. If you get a number of plots on that course, you also have the north and east components of the speed of travel: they will only be equal if it is going at a 45 degree angle to your fixed point.

The Original Sperry Predictor was for visual targets: what happened with it was that when you followed the target, the predictor was in fact the fixed point, so as the predictor swung round there were two sets of motors determining the Eastings and the Northings, and these were then resolved into a rate on a thing like a gramophone turntable: according to whether it was near the centre or near the edge so you got a varying rate. If you applied and fed in ballistic information into that instrument for a particular gun and particular sort of shell, it would give the quadrant elevation at which the missile and target should meet. And this was continuous: this information was fed to dials on the gun, so the layers on the gun didn't look at a sight but followed the dial, matching their pointers, and kept the gun moving evenly to keep the gun in line with the dial. 15 core cables took the information to the gun. It also determined the fuse automatically, rather than setting it with a key. We had to do hand setting of fuses on gunnery courses, but there was automatic fuse setting: you put the fuse on and it was automatically set. The fuse was then put into the breach: the delay was allowed for. There were the 2 dials and the gun layers followed the dial round. This was satisfactory in day time, and this was the position when we started.

How we came to write the drill for night engagement with radar happened like this. One day an ordinance truck arrived at the racecourse site and they said "We have brought you a Sperry Predictor". We said "We have got a Sperry Predictor". "Well you've got another one". This stayed in Quartermaster's store for about a week. Then an I. G. (Commissioned Instructor in Gunnery, with a red band round the head) came down, a Major. He said "We want you to try this predictor, and unless there is a blanket ban on firing for RAF purposes, if you can engage a target with it we want you to engage it." (Sometimes there was an absolute ban because there were fighters, and sometimes had "only fire when instructed to".) He said "This is the Castrated Sperry". We looked at him with raised eyebrows, so this chap then explained what they had done to it.

On the Sperry, when the little ball came across the "gramophone disk", there was another little ball on top of that ball and that was where the rate was fed off into the predictor. This upper second ball had been removed. The other one actually could operate the predictor, but would not produce a rate. There were two little knobs on the side which allowed you to set in a wind correction as a rate in Northings and Eastings and what we were told was that we were to do the plotting from the radar and produce a rate and feed it into the predictor by this wind mechanism. This moved the arm on the disk, which was no longer being propelled by the other ball.

Hence when we got a target and got it worked out, when the guns were being used semi-automatically, after the initial salvo, you could do 2 things. Either plot it and calculate that that was the salvo point and give the guns a bearing etc. off it. Or, and this is where we had to make these rather clever scales: eventually we got some that you could lay onto the grid on the plotting table and read off the results continuously. We made all sorts of things. Then we called eastings x and northings y and set them into the predictor, and provided the target kept on that course, the predictor was churning out all the time the future position: the quadric elevation for the guns, so you were able to establish continuous gunfire at night, rather than salvo fire. With salvo fire you could only get about 2 salvos off anyway, and if you were a mile out you missed it. If they were following continuously, once they had fired the first salvo, the guns then went onto independent fire which meant some guns fired more rapidly than others. The point is that they were not all firing at same time, so you got a continuous pattern of gunfire along the predicted course of the plane. And that was the drill which we wrote out in the middle of the night once, when the I.G. came back to see what we had done. George was on duty at the command post: he always went down there at night and spent the night in the command post. The Gun Position Officer only came down if there was an alarm. There were 2 sets of alarms, one when the radar went on watch and we were down in Command Post. But this was not a turn out for the guns, but when the guns turned out the GPO came down. They were on watch and Tom Riddle came down and said "the I.G.'s arrived" (it was quite late in the evening), "and he wants the drill we've worked out for the predictor." George said "Drill sir? Which drill?" "We've done a lot of things: he's OK now, he's in the mess with a drink, but we'll have to show him something in the morning. When he goes to bed I'll come down and we'll see what we can knock up." While he was away Tommy Jones and George wrote down a few notes of the most likely things; there were many possibilities they had been playing about with: No 1 does this No 2 does the other. They wrote this out. Tom Riddle was quite happy; he was a site Captain and he said "You'd better get the predictor detachment" (they had got to carry the drill out) "out early in the morning and make sure they all know which thing they are supposed to do". After breakfast he brought the I.G. down and went through all the drill a couple of times, and we gave him his piece of paper. Off he went quite happy. Ten days later a copy of that paper was received on the site as the Provisional Drill for the engagement of enemy targets by Modified Sperry Predictor - by this time the ATS were involved, so it was no longer 鈥渃astrated鈥�. And so it remained, with only very minor modification, for the rest of the war, until they got the No10 Predictor and direct Radar control.))

Out of the development of the drill, we went down to Manorbier to do a spell of trials with the Modified Sperry Predictor at the Trials Wing there.

(As told by George Johnston to his son 1990)

漏 Copyright of content contributed to this Archive rests with the author. Find out how you can use this.