Search

This article describes an improved rig for the dynamic calibration of skinfold calipers. The new unit is 5% lighter and almost 60% smaller than its predecessor (Carlyon et al., 1996, 1998) with a 9.5 mm solid aluminium base and a quick release caliper mount providing stability to both the rig and caliper. Automation of the gap controller with an electric motor standardizes the jaw opening and closing velocity, thereby enabling hands-free operation. Frictional losses in the moving components of the rig have been reduced by replacing the main bush of the swing arm with a bearing, reducing the mass of the swing arm, adding a support wheel to the end of the swing arm, and replacing fishing swivels with a universal joint to allow for changes in the opening screw angle as the caliper's arm moves through its arc. This rig can also be adapted to different types of calipers by changing the position of the load cell, microswitches, and the caliper mount. A universal mounting bracket that can be secured to almost any table supports the rig in a vertical plane when calibrating the load cell. To demonstrate the versatility of the calibration rig, preliminary data are presented for the upscale and downscale jaw pressures of seven Harpenden and seven Slim Guide calipers. Am. J. Hum. Biol. 14:721–727, 2002. © 2002 Wiley-Liss, Inc.

The components of our rig, which is used for the dynamic calibration of cycle ergometers, were purchased for $4,080 (USA); interfacing the instrument with a computer and printer increased the cost by an additional $2,550 (USA). The rig is essentially a torque balance and was therefore validated against the primary measurements of: angular velocity (rev.min-1), distance (m) and mass (kg). Summing the inaccuracies associated with these 3 measurements yields a worst case error of only 0.3%. The coefficients of variation for a dynamic calibration rig and mechanically braked Monark cycle ergometer assembly ranged from 0.6 to 3.2% for 6 trials at each of the following power outputs: 117.7, 176.6, 235.4, 294.3 and 353.2 W. The instrument therefore has a high degree of precision.

In this study we measured the accuracy of the following types of cycle ergometer against the criterion of a dynamic calibration rig (DCR): 35 friction-braked (Monark), 5 research-grade air-braked (Repco) and 5 electromagnetically braked (2 Siemens, 1 Elema-Schonander, 1 Ergoline, l Warren E. Collins). Monark ergometer power outputs over the range 58.9–353.2 W significantly (P < 0.001) underestimated those registered by the DCR with mean accuracies of 91.7–97.8%. The least accurate individual reading for each of the six up-scale (0–353.2 W) power outputs ranged from 81.6␣to␣91.6%; corresponding down-scale (353.2–0 W) accuracies were 85.1–92.5%. A hysteresis effect was furthermore evident for this ergometer in that up-scale measurements were significantly (P < 0.05) greater than down-scale ones. In addition, when the oldest [mean (SD): 11.3 (2.3) years old] and newest [1.4 (0.8) years old] eight ergometers were compared, the latter were significantly (P < 0.05) more accurate over the range 117.7–294.3 W. Apart from the two lowest power outputs of 47␣W (62.2–96.0% accuracy) and 127 W (88.0–97.7% accuracy), the individual up-scale and down-scale accuracies of the Repco ergometers ranged from 98.0 to 104.2% for power outputs of 272.7–1137.8 W and the means were not significantly different from those of the DCR. There was also no evidence of hysteresis. Except for the initial power output of 50 W (40 rev/min: 83.8–99.2% accuracy; 60 rev/min: 93.2–122.6% accuracy), the␣individual accuracies of the electromagnetically braked ergometers ranged from 89.3 to 101.4% over the up-scale range of 100–400 W, and none of the means were significantly different from those of the DCR. The variability of individual errors for the preceding data emphasises that all cycle ergometers should be validated against the criterion of a DCR if accurate power outputs are required.

Two-terminal integrated circuit temperature transducers (Analog Devices' AD 590) were used to fabricate skin and rectal probes for the monitoring of mean body temperature. They are inexpensive, robust, easily constructed and of low mass. The skin sensor system described is capable of being solidly attached regardless of the body contour such that it remains integral throughout profuse sweating and vigorous movement. Both types of probe are stable, accurate to within +/- 0.05 degree C over the physiological range of measurement and exhibit 100% response times of approximately 60 s to a square wave stimulus of 10 degrees C.