The equipment used to perform experiments on the human elbow has seen many revisions and many groups of people have contributed to the existing version. This paper is an attempt to unravel years of different ideas and views to enable the reader to understand the current state of affairs. The apparatus consists of 5 main components:
Servo valve, manifold, Rotac and torque transmission equipment
The position control system
The support structure for the equipment and subject
The computer data acquisition system
Each system will be dealt with in turn and in the greatest possible detail. If any changes are effected by the reader, please add them to the descriptions to avoid the frustration associated with unraveling someone else's handiwork.
The system started out with a 1000 psi system located on the third floor. Noise and inconvenience required a move of the offending equipment to the basement. Plumbers (rather than hydraulic specialists) installed stainless steel pipes from the basement to the third floor. A hydraulic service company (MCS Servo Inc.) modified that system to accommodate a 3000 psi pump and installed supporting brackets with rubber bushings to reduce vibration. Further modifications include the addition of the robotics laboratory to the existing system and individual solenoid control at each post. Collectors at each post and shut off valves concluded the modifications. Flushing of the system was performed during the initial installation.
The elbow actuator and ankle actuator are similar but significantly different. The ankle actuator in its current incarnation is a 3000 psi system installed in approximately 1989 and in regular use since that time. The elbow actuator has seen limited use since its conception circa 1976. The elbow actuator system has a 1000 psi maximum using 3/4" pipe and size 12 pipe thread fittings (NPT). Input and output pipes are labeled on the fittings (Both butterfly fittings are the same gender, so the fittings are interchangeable but the lines are NOT interchangeable with regards to the servo-valve.). Take care to connect high and low pressure properly. Components of the elbow hydraulic system have been in use since 1967.
Within the control box of the elbow hydraulic system, there is a 1 A fuse (now replaced with a 5A) to prevent overload. The solenoid of the elbow hydraulic system has been connected across the first switch of the control box - Power ON/OFF. Connecting across the light associated with this switch met with no success as the voltage drop was too great.
The Rotac rotary hydraulic actuator is a 360° model and will not continue in one direction if given a constant input. The actuator is joined by a torque tranducer (strain gauge) and a shaft supported by a double roll ball bearing.
The servo-valve was re-calibrated and overhauled in August of 1992. The reported cause of the problem was tampering with the motor couple within the servo itself. DO NOT take the servo apart as recalibration is very expensive. The servo is dated 1967 and replacement cost is high so care is suggested when dealing with the servo-valve. Also in 1992, the system was flushed for a second time to remove any debris remaining in the elbow actuator branch of the system.
Connection of the servo-valve is a little less than obvious. The problem
arises when "standard" connections on the servo do not correspond to
"standard" connections on the controller module. The problem was
solved by a unique wire connector. DO NOT USE these connectors for any
other application. The motor coils are connected in parallel, implying that A
and C are connected to positive and B and D are connected to negative. So, at
the controller module end of things, only 3 of the 5 pins available are used.
Controller Module Servo-Valve
(five pin mil spec) (four pin mil spec)
A A & C
B B & D
C ground/shield
Detailed specifications are available from the manufacturer in pamphlet form.
Military specification (or "Cannon") connectors have 4 or 5 connecting pins in this apparatus. All male connectors are labeled clockwise starting with A denoting the pin closest to the keyway. Female connectors are connected in a counter-clockwise path.
The potentiometer used is a standard 1 revolution, 50kO
variable resistor. When facing the POT (shaft away from the viewer), +10V is
supplied at 8 o'clock, -10V at 7 o'clock, and the signal (variable voltage/resistance)
at 6 o'clock. Orientation may change but note that the order will always remain
the same. Since there was no standard from the potentiometer end of the
connecting wire, the connections were made as per the gain module and are as
follows:
Mil Spec Function Wire
connectorA +10V red
B -10V black
C signal white
D ground/shield green
Faced with two conflicting standard elements, the only
method of attaching the torque transducer and the gain module without modifying
one or the other was to create a unique connecting wire. Once again, DO NOT
USE this connector for any other application as the connections are NOT obvious.
TORQUE TRANSDUCER GAIN MODULE
FUNCTION WIRE WIRE PIN FUNCTION
excit. +ve A C 20 excit. +ve
excit. -ve D A or B 28 excit. -ve
signal +ve B E 12 signal -ve
signal -ve C D 15 signal +ve
Note that despite a contorted pathway, the functions at the two ends do correspond. The only discrepancy is the sign of the signal, which is not significant.