How many faults can a newly-manufactured gear have? Let's just say that there are rather many…. Intended to be used both on the production line - to catch faults as they occurred - as well as in the inspection department, the Parkson gear testing machine was able to check spur, bevel, spiral and worm gears. Introduced in the 1940s, it continued to be available well into the 1950s and many are still in service today.
The machine was of simple construction and very easy to use - yet capable of identifying a wide range of inaccuracies. While, as with all analogue measuring equipment, operator experience played a large part in getting the best from it, even trainees are reported to have coped well and been able to discover most of the significant errors. The main problems able to be identified included errors of eccentricity, tooth thickness, rolling action, and centre distance.
Although some changes were made to the appearance of the machine during its production run - and several different models offered - all operated in the same way. The basic model consisted of a bed in cast iron upon on were carried a pair of gear-carrying arbors (5). One arbor (4) was mounted on a slide that could be locked to set the initial distance between the gears - the makers referred to this as the "adjustable carriage". The other arbor was mounted the "floating carriage" (E), this being carried on a flat caged needle-roller bearing and to give a very free but limited range of travel. For external gears, the "floating carriage" could be spring-loaded by an adjustable plunger (6) the tension of which could be adjusted by a screw. For internal gears, another adjusting nut (3) was provided. Two regulating nuts (7) were provided that acted against the spring loading so that distance between the gears could be adjusted precisely or used to limit or prevent movement of the floating carriage.
Attached to the "adjustable carriage" was a ruler scale that overlapped a vernier on the floating carriage, this arrangement allowing the distance between them to be measured to within 0.001" inches or, on the metric model, 0.02 mm.
The two slides, as shown in the illustrations below, could be fitted with various attachments to test bevel, worm, spiral and spur gears - these being mounted on the centre bosses (2) and located by a dowel pin.
A typical example of using the machine would be to mount a perfect example of the gear on one arbor and the one to be tested on the other. The adjustable carriage was moved to being the two gears close together and then locked. The spring-loaded screw (6) was then adjusted to give the desired engagement pressure, the pair of gears being slowly rotated to see how closely the gears could mesh without binding. If the test gear was perfect when the pair were rotated slowly together, the dial indicator would not move - so proving that there was no variation in their centre distances (and the precise distance also shown on the ruler and vernier scales). If the teeth were eccentric to the bore, or their depth and thickness were not the same - or they were bruised or otherwise imperfect - movement of the dial indicator needle would also show this. This quick check could be followed by many others to determine what remedial action might be taken to improve the settings of the blank and gear-making machines.
As an optional extra, a recording device could be fitted to all models, save for the worm-and-wheel version of the 14 and 36-inch sizes. The recorder traced two lines around a circular chart, the inner line showing the irregularities of the gear, the outer giving the variation from a true circle that a perfect gear would provide.
To accommodate the wide variety of gears that needed to be tested - from very small ones machined into a long shaft to crown wheels and pinions from car transmission systems - several essential accessories were offered; these are all shown below.
Parson also offered to make whatever changes to a machine or its accessories a customer might need for special applications--hence, occasionally, non-standard machines are encountered that will differ from the standard range..

This is the 9-inch between model for sour and helical gears on parallel centres. As an optional extra, as shown fitted, a recording device could be fitted to all models, save for the worm-and-wheel version of the 14 and 36-inch sizes. The recorder traced two lines around a circular chart, the inner line showing the irregularities of the gear, the outer giving the variation from a true circle that a perfect gear would provide.

The 9-inch between centres model equipped for checking worm and wheel gears. The worm was mounted horizontally in bushing as, as in the picture below, between centres. The standard accessory could accommodate worm and wheels gears from 1 inch to 8.5 inches between centres

The ubiquitous 9-inch model this time set up to check spiral gears. The standard accessory could accommodate spiral gears from 1 inch to 8.5 inches between centres

A special model designed to check motor car crown wheels and pinions. Suitable for either straight or spiral teeth, the machine could take crown wheels up to 15 inches in diameter. This pictures shows the machine fitted with the optional recorder

A very much larger (1008 lbs / 458 kg) Parkson gear tester fitted with a recorder and designed to take spur and helical gears on parallel axes. This machine was produced as in two models the smaller of which could take gears up to 24 inches in diameter and the larger (shown below) up to 36 inches. It could also be equipped to check spiral gears by a bracket, mounted on an adjustable carriage, that carried a horizontal arbor - or one set to any angle ordered by the customer.

The 36-inch capacity Parkson gear tester fitted with the optional-extra "centre attachment" to hold spur and other gears on parallel centres. The accessory could hold shafts up to 18 inches long with gears up to 11 inches in diameter. With the accessory removed - it was held on by screws - it could hold larger gears as shown by the 24-inch model

The 36-inch capacity Parkson gear tester fitted with the optional-extra "centre attachment" to hold bevel gears. In addition to the usual ruler graduations and vernier fitted to the horizontal slides, they were also fitted to the vertical - this arrangement allowing the distances from the faces of bosses on the gears to the apex of their pitch cones to be read both vertically and horizontally.

The 36-inch capacity Parkson gear tester fitted with the optional-extra attachment to hold worm and wheel and spiral gears. The worm was held in bushes secured by caps, so allowing a quick change from one size to another. The slide was also able to be moved a few degrees either side of horizontal so that the angular contact of the worm threads could be adjusted. The slide's horizontal position was located by a dowel pin and the amount of swivel controlled by a handwheel working through screw thread indicated on a degree scale

A picture showing, in detail, the bearings used to support the shaft of the worm. The nearer bearing has its cap removed and the bushing in place. To speed up changes of job a clever system was used - the bearings caps were held on hinged bars and clamped by means of an inclined face that acted as a cam