A strain gauge sensor lets a machined body be deformed by the load it is measuring and reads that deformation electrically. Bonded foil gauges change resistance, a bridge circuit turns the change into millivolts, and a calibration converts the millivolts into inch-pounds.
Every other torque sensor in this catalogue is placed in the torque path: something drives one end, something resists the other, and the sensor sits between them. This one goes somewhere different — underneath.
A motor pushes back on itself. Whatever torque a motor delivers to its load, it applies an equal and opposite torque to its own casing. That reaction has to be held by something, and what holds it is the mounting. So if you instrument the mounting rather than the shaft, you are measuring the same torque from the other side. Published descriptions of reaction torque measurement put it exactly this way: the measurement is taken at the point where the torque is transferred into the ground.
What that buys is access. You can instrument a machine that already exists without cutting into its drivetrain. The motor keeps its own output shaft, its own coupling and its own alignment; nothing is disturbed downstream; and the sensor goes in during a mounting operation rather than a rebuild. On equipment that is awkward to take apart, or where the drivetrain arrangement is not yours to change, that is often the difference between measuring and not measuring.
It also leaves the rotating system alone. Anything inserted into a shaft adds inertia and adds torsional compliance, both of which change the behaviour of the thing you were trying to characterise. A sensor under the motor adds neither. Guidance on reaction measurement notes exactly this — the arrangement is simpler and mass effects are negligible, where a rotating sensor needs precision alignment and inertia compensation.
The cost is that the reaction path is now the measurement path. The mounting has to be the only thing holding the casing still. Anything else that restrains it — a rigid conduit, a stiff power cable, a cooling line, a second bracket someone added for convenience, a foot resting on the frame — carries part of the reaction around your sensor. The torque does not disappear; it simply stops being measured, and because those restraints vary with temperature and handling, the share that goes missing is not constant.
And the equality holds only while things are steady. Reaction measurement assumes negligible angular acceleration — that is how the technique is defined. During a genuine speed change the motor's rotor is being accelerated as well as driving the load, so the reaction the mount feels and the torque arriving at the load differ by that inertial share until the speed settles. At constant speed the two are the same number.
Sensors of this kind are used for electric and internal combustion motor testing, switch and control mechanism monitoring, automotive steering, suspension and drivetrain evaluation, industrial mixing, pumping and conveyor machinery, material characterisation, research and development, quality control, turbine and generator monitoring, calibration reference work and aerospace assembly.
The bridge behaves conventionally. What matters on a mount-mounted sensor is the mechanical arrangement, and there are four things about it worth knowing before anything is bolted together.
Signal. The bridge is nominally 350 ohms and the supply 10 VDC, and rated torque returns 2 mV/V. Because the sensitivity is a ratio, the actual signal scales with whatever you excite it at; run the standard supply and the top of the range sits at roughly two hundredths of a volt. Every model is certified clockwise and counterclockwise, which matters here because a motor under test may be driven in either direction and because braking reverses the reaction.
Everything goes to the sensor, and nothing goes past it. This is the whole discipline of a reaction mount installation. The device bolts to one face, the mounting plate to the other, and every other connection to that device has to be flexible enough not to carry torque. Route power and instrumentation cables in gentle loops rather than straight rigid runs. Use flexible hose for coolant or air rather than hard pipe. Remove any bracket, stop or support that is not part of the intended reaction path. Whatever restrains the casing, other than the sensor, is measuring torque you will never see.
Give the shaft clearance. Where a motor's own shaft passes through or near the sensor, published installation guidance is explicit that it must pass without touching the inner diameter — contact there transmits force directly and corrupts the reading. Establish that clearance at the design stage, and remember that a shaft moves slightly under load and can grow with temperature, so a gap that exists cold and unloaded may not survive the test.
Two shafts, two flange faces, and two body sizes. The lower three capacities share one set of dimensions and the upper three a larger one, with the drive interface growing between them. Take the figures from the model you intend to buy rather than from the series, and design the mounting plate and the device adapter to suit — on this configuration those two plates are as much part of the instrument as the sensor is, because any flexibility in either of them shows up as compliance in the reaction path.
Stainless and sealed, which is unusual in this family. The body is 17-4 PH heat-treated stainless steel with bonded foil gauges sealed against industrial environments — a more robust specification than the anodized aluminium used across our other reaction sensors. That is a deliberate match to where this configuration is used: bolted into machinery, under motors and pumps, on production plant rather than a laboratory bench. It suits dust, coolant mist and everyday moisture; sustained wash-down and corrosive service remain worth raising with us before ordering.
Specifying a reaction mount sensor is mostly about being honest concerning what the machine can produce and what else is attached to it. Our application engineers deal with motor and machinery installations regularly and would rather see the arrangement before it is built.
Size against what the machine can do, not what it normally does. This is the most common error on motor work. An electric motor's torque at stall or locked rotor is typically several times its rated running figure, and a reaction mount sees every bit of it — there is no clutch, no slip and nothing to absorb it. Direct-on-line starting, a jammed load, a sudden reversal or a braking event can all produce far more than the nameplate. Establish the worst case the drive is capable of and keep it inside the safe overload figure, which is a survival limit rather than a working allowance.
Then audit everything else attached to the device. Walk round it and list every solid connection to the outside world — power cable, conduit, instrumentation, coolant, air, drain, any bracket or stop. Each one is a candidate for carrying reaction torque around your sensor. Making them flexible is design work best done before installation, and it is far cheaper than diagnosing a reading that is quietly ten percent low.
Ask whether your measurement is steady state. If you are characterising a motor at constant speed, holding torque, monitoring a pump or verifying a switch mechanism, a reaction mount gives you the number directly. If the interesting part is what happens during acceleration, braking or a rapid transient, understand that the rotor's own inertia is part of what you will record, and talk to us about whether a reaction arrangement suits the test you actually want to run.
Treat the adapter and the mounting plate as instrument parts. Both sit in series with the measurement, so whatever springiness they have is added to the reaction path and shows up as a slower, softer response than the sensor itself is capable of. Flat faces, generous section and a proper bolt pattern — a plate specified as a bracket will undo a sensor specified as a transducer.
Order the mating assembly with the sensor. The sensor terminates in a connector and nothing plugs into it out of the box, so one mating assembly per sensor belongs on the order, in whichever of the available cable lengths reaches your instrument. Worth flagging if you have used our other reaction sensors, several of which ship with a cable included — on this one it is a separate line and easy to overlook until installation day. Choose the length deliberately, and route it so it hangs on the structure rather than on the device you are measuring.
Then choose the instrumentation. An amplifier signal conditioner module gives a voltage or current output for a test system or controller, and a digital display with setpoints suits bench work and machine monitoring where a limit has to act. The Cal-Teds plug and play option adds an IEEE 1451.4 memory to the connector so a smart instrument reads the calibration off the sensor itself — useful on plant where a unit gets unbolted from one machine and fitted to another.
Describe the machine, what it can produce at worst, what else is bolted to it and whether the measurement is steady or transient, and we will help you settle on a capacity and an arrangement. Stock capacities ship quickly, and academic and research purchases are discounted.
STS Series Torque Sensor Applications.
The Transducer Techniques STS Series reaction type dual shaft torque sensors are often installed between devices such as motors, switches, shafts or axles and their mounting plate to provide bi-directional torque output, and serve multiple industries.
- Motor Testing: STS Series sensors measure torque from electric motors and internal combustion engines.
- Switch and Control Mechanisms: These sensors monitor torque in switches and control system mechanisms.
- Automotive Testing: STS Series sensors evaluate steering, suspension and drivetrain components.
- Industrial Machinery: These sensors track torque during mixing, pumping and conveyor operations.
- Material Testing: STS Series sensors characterize the mechanical properties of materials.
- Research and Development: These sensors are used to optimize prototypes and designs.
- Quality Control: STS Series sensors help ensure fasteners and assemblies meet torque specifications.
- Energy and Power Generation: These sensors monitor turbines and generators.
- Calibration Services: STS Series sensors provide reference standards for torque measurement.
- Aerospace: These sensors are used in the assembly of critical aircraft components.
These sensors offer a range of capacity options, making them suitable for various torque measurement needs, with robust construction and environmental protection ensuring reliability across professional fields requiring precise torque measurement.
Frequently Asked Questions
What is a dual shaft reaction torque sensor, and where does it go?
It goes between a device and the plate that device is bolted to. A motor, switch, shaft or axle mounts to one face of the sensor, the mounting plate to the other, and everything the device tries to twist against its mounting passes through the sensor on the way. That is a different position from every other torque sensor in our range, all of which sit in the torque path between a driver and a load. This one sits in the reaction path, underneath, and never touches the drivetrain at all.
What capacities does the STS Series cover?
Six models from 600 to 12,000 in-lb: the STS-600, STS-1.2K, STS-2K, STS-3K, STS-6K and STS-12K. They are built on two body sizes, with the 600, 1,200 and 2,000 in-lb models sharing one footprint and the 3,000, 6,000 and 12,000 in-lb models a larger one. Check which group your capacity falls into before designing the mounting plate and the device adapter, because the dimensions and the drive interface both change at that step.
How does measuring at the mounting give me the device's output torque?
Because a motor cannot push on its load without pushing back equally on itself. Whatever torque leaves the shaft, an equal and opposite torque acts on the casing, and something has to hold that — normally the mounting. Put a sensor in the mounting and you are measuring the same torque from the other side, which is why this technique is described as measuring at the point where torque is transferred into the ground. The practical advantage is that you get the number without cutting into the shaft, disturbing a coupling or changing the alignment of anything downstream.
Does the reading equal shaft torque during acceleration?
At constant speed, yes. During a speed change, not exactly. Reaction measurement is defined as applying where angular acceleration is negligible, and the reason is that an accelerating motor is doing two jobs at once — driving the load and spinning up its own rotor. The reaction at the casing includes both, while the torque arriving at the load does not, so the two numbers separate during the transient and come back together when the speed settles. For steady-state characterisation this is a non-issue. If the transient is the thing you care about, tell us and we will talk through whether a reaction arrangement suits the test.
What is the STS made from and how is it protected?
17-4 PH heat-treated stainless steel with bonded foil gauges sealed against industrial environments — a noticeably more robust specification than the anodized aluminium used across our other reaction torque sensors. That reflects where this configuration ends up: bolted under motors and pumps, built into production machinery, living with dust, coolant mist and everyday moisture rather than sitting on a bench. Sustained wash-down, condensing conditions and corrosive atmospheres are a different requirement and worth describing to us before ordering.
Is the STS calibrated in both rotational directions?
Yes, clockwise and counterclockwise on every model. On a reaction mount that is more than a convenience. A motor under test may be run in either direction, and more importantly the reaction reverses under braking or regeneration — a drive that is holding a load back produces torque in the opposite sense to one that is driving it. Bi-directional certification means the sensor describes both halves of that cycle rather than only the driving half.
How accurate is the STS Series?
Nonlinearity is 0.2% of rated output and hysteresis 0.10%, with zero balance at 1.0% and nonrepeatability published at 0.5%. Worth a conversation with us about the repeatability figure for the model you are considering, because on a reaction mount installation repeatability in service is dominated by how consistently the whole assembly is bolted together — a device removed and refitted between tests is a different mechanical arrangement each time, and that usually matters more to a result than the transducer specification does.
What do I need to order alongside the sensor?
A mating assembly, one per sensor, available in several cable lengths. It is required rather than optional, since the sensor terminates in a connector. This is worth calling out if you have used our other reaction torque sensors, several of which include a cable in the box — on this series it is a separate item and it is the thing most likely to be missing on installation day. Choose the length to suit the route rather than defaulting to the shortest, and order it at the same time as the sensor.
How does temperature affect an STS reading?
The compensated range is 60° to 160°F with safe limits from −65° to 200°F. Output drift is 0.005% per °F and zero drift 0.01% of rated output per °F. On this configuration there is usually a larger temperature effect than the sensor's own: the device you are measuring warms as it runs, its mounting and any attached pipework expand, and cables stiffen or relax. Zero the system once the machine has reached its normal running temperature rather than from cold, and expect the first minutes of a run to be less trustworthy than the rest.
Can the STS be used with something other than a motor?
Yes — anything whose torque is ultimately reacted by a mounting works the same way. Switches and control mechanisms, valve actuators, gearboxes, pumps, mixers, winches and steering assemblies have all been measured this way. The test is simple: if the device is bolted down and produces or resists torque about the axis of its mounting, the mounting has to hold that torque and a sensor in the mounting can read it. Tell us what the device is and how it is fixed and we will confirm whether the arrangement suits.
Questions From The Field
My reading is lower than the machine should be producing.
Something is almost certainly carrying part of the reaction around the sensor. On a reaction mount the sensor only measures the torque that actually passes through it, so every other solid connection to the device is a candidate: a rigid conduit, a stiff power cable, hard-piped coolant or air, a bracket added for support, a foot resting against the frame. Walk round the device and list everything touching it, then make each one flexible or remove it. The tell is that the shortfall usually changes with temperature and after anything is disturbed, because the stiffness of those parallel paths is not constant.
Readings look right at speed but wrong during start-up.
That is the inertial term rather than a fault. While the motor is accelerating it is both driving the load and spinning up its own rotor, and the reaction at the casing includes both contributions while the torque reaching the load does not. As soon as the speed stops changing the two converge. Two practical responses: take steady-state readings after the speed has settled, and if the acceleration phase is genuinely what you need to characterise, discuss it with us, because that is a different measurement problem from the one this configuration is best at.
Does the motor shaft need clearance where it passes the sensor?
Yes, and it is worth checking properly. Published installation guidance for reaction sensors used with motors is explicit that the shaft must pass through without touching the inner diameter, because any contact transmits force directly and corrupts the measurement. Two things to allow for beyond the static gap: a shaft deflects slightly under load, and both shaft and housing grow with temperature. A clearance that looks generous on a cold bench can close under a hot, loaded running condition, and the symptom is a reading that drifts or sticks rather than one that fails obviously.
Can the STS stay installed on a machine that runs continuously?
It is well suited to it — stainless construction with sealed gauges and no moving parts is exactly the specification for permanent installation, and industrial machinery monitoring is a listed application. Three things deserve a schedule rather than being left. Check the unloaded reading periodically against the zero balance figure, since a growing offset is the earliest sign anything has changed. Inspect the bolted joints at both faces, because fasteners loosen and a loose plate changes the reaction path. And look at the cable and connector, which on machinery that vibrates is usually the shortest-lived part of the assembly.
What capacity do I need for a motor — running torque or stall torque?
Size on what the motor can produce, not what it usually does. Locked rotor or stall torque on an electric motor is commonly several times the rated running figure, and a reaction mount sees all of it — there is nothing between the casing and the sensor to slip or absorb. Direct-on-line starting, a jammed load, a reversal or a braking event can all take you well past the nameplate. Work out the credible worst case, keep it inside the safe overload figure rather than the rated figure, and if the machine is capable of far more than the test needs, consider a mechanical stop in the arrangement.
Can I measure a gearbox or a pump this way rather than a bare motor?
You can, and it is often more useful, but be clear about which torque you are getting. A sensor under a gearbox measures the reaction the gearbox casing applies to its mounting, which corresponds to the difference between what goes in and what comes out rather than either one on its own. On a pump or mixer it measures what the whole unit is reacting against its base. Both are real, meaningful numbers — they are just not always the number people assume. Describe the machine and what you want to know and we will confirm what a mount-mounted sensor will actually tell you.
The sensor is bolted down properly but readings drift as the machine warms.
Look at the surrounding structure before the sensor. The device heats as it runs, its mounting plate and frame expand, and anything rigidly attached across that growing assembly develops force it did not have when cold — pipework and conduit are the usual culprits, and both apply torque to the casing as they are stretched or compressed. The sensor's own drift figures are small next to that. Re-zero once the machine has reached working temperature, and if the drift persists after everything has stabilised, work through the attached services looking for anything rigid.
Do I have to disconnect the machine's drivetrain to install one of these?
No — that is the point of the configuration. The device keeps its own output shaft, coupling and alignment, and the installation happens at the mounting instead. What you do have to do is create space for the sensor and the two plates between the device and its existing mounting surface, which changes the height and sometimes the alignment of the whole assembly relative to whatever it drives. That is the design work worth doing carefully: the drivetrain is undisturbed only if the device ends up back where it started. Send us the arrangement and we will look at the stack-up with you.