A load cell is a machined piece of metal that has been persuaded to report on its own deformation. Strain gauges bonded to it change resistance as it flexes, a Wheatstone bridge turns that into millivolts, and a calibration certificate turns the millivolts into pounds.
Buyers spend most of their attention on the numbers that describe that conversion, which is reasonable but incomplete — because a force has to get into the sensor before any of those numbers apply, and how it does so is very often the largest error in the system.
The scale of it surprises people. Force calibration laboratories have measured the contribution made by adapters alone — the fittings that carry load into a sensor — and put the resulting errors anywhere from 0.01% to 5% depending on the device and the arrangement. Set that against a sensor whose published nonlinearity is 0.1%, and the object screwed into the load cell can be worth many times the load cell's entire error budget.
Which is why the shape of a sensor is really a statement about how it wants to be loaded. A disc-shaped low profile cell answers that question differently from anything with a single thread at each end. Force enters through a threaded hole at the centre and leaves through a ring of hold-down bolts around the circumference, so instead of travelling point to point it spreads outward through a symmetric structure.
Symmetry is doing real work there. If a load arrives slightly off-centre or slightly out of square, a symmetric flexure produces strains on opposite sides that partly oppose one another, and a good deal of the error cancels rather than accumulating. No single-threaded geometry can do that, because there is nothing on the far side to cancel against. It is the main reason this family of shapes is what force laboratories build their reference and transfer standards from, and why they are trusted to hold a calibration between visits.
Whatever you bolt it to is now inside the measuring loop. With load reacting through a bolt circle, whatever the sensor is bolted to is carrying that reaction. Published installation guidance for low profile cells is unambiguous that the surface must be flat and rigid enough to resist any appreciable deformation under load. A base that dishes even slightly, or hold-down bolts tightened unevenly, distorts the flexure before a single pound has been applied — and the sensor reports that distortion as though it were force.
Off-axis loading does not add noise; it biases you low. A force arriving at an angle to the measuring axis is reported reduced in proportion to the cosine of that angle. The reading stays perfectly steady and repeatable while being consistently wrong, which is a harder fault to notice than an unstable one.
Sensors of this form suit material and product testing in tension and compression, industrial automation and assembly line verification, hydraulic and pneumatic force monitoring, quality control in automotive and aerospace manufacture, research and prototype evaluation, and laboratory instruction.
The bridge behaves as it does in any strain gauge sensor. What is worth setting out here is the chain either side of it, because on this geometry the details at each end are unusually consequential.
Signal, and why more of it is better. Excitation goes in at 10 VDC and the bridge returns 3 mV/V at rated load — half again as much as the 2 mV/V that is standard across most load cells, including most of ours. Output is expressed per volt of excitation because it is a ratio rather than a fixed voltage. The practical benefit of the larger figure is that everything downstream has more to work with: the same cable noise, amplifier noise and converter resolution represent a smaller share of a bigger signal. It is a specification that rarely gets compared and quietly improves the whole measuring chain.
The centre thread is the one place all the load passes through. Everything the sensor will ever measure arrives through that single threaded hole, so what is fitted into it deserves as much thought as the sensor selection did. For tension, a rod end or self-aligning fitting lets the load path settle onto its own axis instead of forcing the sensor to absorb a disagreement. For compression, a load button introduces the force through a defined, slightly domed contact rather than through a rigid flat interface that transmits every bit of misalignment it is given. Published calibration testing found that using ball-type compression adapters improved reproducibility by 30–40%, which is a very large return on a small component.
Thread engagement and locking matter, and the procedure varies with capacity. Manufacturer guidance for low profile cells distinguishes between smaller and larger units — below roughly twelve and a half thousand pounds, adapters are torqued and locked with a jam nut to documented figures, while above that a common practice is to pre-load the adapter connection by pulling it well past rated capacity before hand-tightening the jam nut, so the joint is properly seated before use. Because the correct procedure depends on both the model and the adapter, ask us for the figures applicable to yours rather than working from a general rule.
Stability is the specification that decides whether a calibration is still true. Zero drift with temperature is quoted at 0.001% of rated output per degree Fahrenheit here — several times better than the figures normally seen on industrial load cells. That is worth more than it looks. Laboratories that track this closely report that drift between calibrations can inflate an overall measurement uncertainty by anywhere from two to ten times, so a sensor whose zero stays where it was put is not simply more convenient; it is the difference between a certificate that still describes the instrument in month eleven and one that does not.
Range without a change of character. Fourteen capacities from 100 lb to 50,000 lb are covered by the same geometry in four body sizes, all in heat-treated 17-4 PH stainless with bonded foil gauges sealed against most industrial environments. Diameter grows from three inches to six, and height only from one inch to one and four-fifths — fifty thousand pounds in a disc under two inches thick.
With fourteen capacities and a single well-understood form, choosing an LPU is mostly a matter of landing on the right range and then getting the interfaces right. The second half is where the money is either earned or wasted, so it is worth a conversation with our application engineers rather than a quick pass down the table.
Land close to your working force. The range is stepped finely for a reason. Every accuracy figure is a proportion of rated capacity, so a sensor chosen well above what you actually apply spends its resolution on force you never produce. Fourteen steps exist precisely so you can sit near your working load while keeping sensible margin above your credible peak.
Specify the adapters at the same time as the sensor. This is the recommendation most often ignored and most often regretted. Decide now what goes into the centre hole — a load button for compression, a rod end or threaded adapter for tension — and treat it as part of the measurement rather than as hardware to source later. Where results have to withstand scrutiny, having the sensor calibrated with its adapters fitted removes a variable that is otherwise very difficult to quantify after the fact.
Design the mounting before you order. Confirm you have a flat, rigid surface with the right bolt pattern, and check the bolt circle and hole size for your specific capacity rather than assuming they carry across the range. Diameter and bolt circle both step up as capacity increases, and a fixture drilled for one body size will not accept another.
Decide the direction and the certificate. The sensor is universal, reading tension and compression through the same centre thread. A compression certificate comes with the cell as standard, and tension certification can be added — at one price up to 20,000 lb and a higher one above it. Which certificate you need is a question about the results you have to defend, not about what the hardware can do.
Confirm the environment. Sealed bonded foil gauges in stainless handle most industrial surroundings, dust, and incidental moisture. Wash-down, chemical exposure, sustained outdoor service or immersion are a different specification, and we would rather size that correctly at the enquiry than have you discover the boundary in service.
Then settle the readout. Three millivolts per volt is a strong signal but still a small one, so it needs conditioning — an amplifier signal conditioner module for a usable voltage or current, or a digital display where you want a local reading with setpoints, analog output or logging. Make sure whatever you choose is configured for the higher output rather than assuming a two millivolt per volt default, and consider Cal-Teds plug and play where sensors move between fixtures.
Tell us the forces, the directions, the mounting you have and what the results are for, and we will help you settle on a capacity and the adapters to go with it. Standard ranges are held in stock, and teaching and research buyers should ask about discounted pricing.
LPU Series Load Cell Applications.
The Transducer Techniques LPU Series low-profile, universal tension or compression "pancake" type load cells are versatile force sensors used in a wide range of applications that require precise force measurement in both tension and compression.
- Industrial Automation: LPU Series load cells are integrated into manufacturing and assembly systems for force measurement and quality control.
- Material Testing: These load cells are used in laboratory testing, including tensile, compressive, and fatigue analyses.
- Product Testing: LPU Series load cells are used for load testing, performance evaluation, and durability assessment.
- Quality Control: These load cells verify force specifications in automotive and aerospace manufacturing.
- Assembly Line Testing: LPU Series load cells verify components and assemblies during production.
- Research and Development: Engineers use these load cells for prototyping and structural integrity evaluation.
- Custom Machinery: LPU Series load cells are integrated into specialized testing and manufacturing equipment.
- Compression Force Measurement: These load cells are ideal when used with ALB load buttons for accurate compression measurement.
- Hydraulic and Pneumatic Systems: LPU Series load cells provide force monitoring and control in hydraulic and pneumatic systems.
- Educational Laboratories: Educational institutions use these load cells for mechanics and materials science instruction and experimentation.
The Transducer Techniques LPU Series load cells offer a versatile and reliable solution for precise force measurement in a wide range of industrial and scientific applications. Their durability and accuracy contribute to improved product quality, safety, and process efficiency.
Frequently Asked Questions
What is a "pancake" load cell and how is the LPU mounted?
A pancake load cell is a flat disc rather than a beam, a block or a link. The LPU has a ring of hold-down bolt holes around its face and a threaded hole at the centre, so it bolts flat to a rigid surface and takes the force through the middle. Load enters at the centre and reacts outward through the bolt circle. That arrangement is what gives the format its two defining strengths — a very short height for the capacity, and a symmetric structure that tolerates imperfect loading better than point-to-point designs do.
What capacities does the LPU Series cover and what do they cost?
Fourteen ranges, the widest span we offer: 100, 250, 500 and 1,000 lb at $880; 2,000, 3,000, 4,000 and 5,000 lb at $930; 7,500 lb at $1,050; 10,000 and 15,000 lb at $1,100; and 20,000, 30,000 and 50,000 lb at $1,275. Because price is flat within each group, there is nothing to save by choosing the smallest capacity in a band — but there is accuracy to gain by picking the one closest to your working force, since all the accuracy figures are proportions of rated capacity.
Why is the LPU rated at 3 mV/V when most load cells are 2?
It produces half again as much signal for the same excitation, which is a genuine and under-appreciated advantage. Everything after the sensor — cable, amplifier, converter — contributes a roughly fixed amount of noise and a fixed resolution, so a larger signal means all of that represents a smaller proportion of your reading. The one practical caution is to make sure your instrument is configured for 3 mV/V rather than defaulting to 2, or your readings will be out by half as much again.
What is the LPU's temperature stability?
The compensated range is 60° to 160°F with safe limits from −65° to 200°F. Output drift is 0.005% of load per °F, and zero drift is 0.001% of rated output per °F — the latter is several times better than the figures typical of industrial load cells, including most of our own range. Zero stability is worth more than it appears on a datasheet: it is largely what determines whether a calibration certificate still describes the sensor accurately months after it was issued.
How accurate is the LPU Series?
Nonlinearity and hysteresis are each 0.1% of rated output, nonrepeatability 0.05%, and zero balance 1.0%. Worth keeping in perspective, though: on this geometry the adapters and the mounting frequently contribute more error than the sensor does. Published calibration work puts adapter-related errors anywhere between 0.01% and 5% depending on the arrangement, so a 0.1% sensor fitted carelessly will not deliver 0.1% results.
What are the LPU's dimensions and bolt patterns?
Four body sizes across the range. The 100 to 1,000 lb models are 3.0" diameter and 1.0" thick with six hold-down holes on a 2.25" bolt circle and a 3/8-24 UNF centre thread. The 2,000 to 5,000 lb models are 3.5" by 1.0" with six holes on a 2.63" circle and a 1/2-20 UNF thread. The 7,500 to 15,000 lb models are 5.5" by 1.8" with eight holes on a 4.50" circle and a 1-14 UNS thread. The 20,000 to 50,000 lb models are 6.0" by 1.8" with eight holes on a 4.88" circle and a 1 1/2-12 UNF thread. Design the fixture around your capacity band, since a plate drilled for one will not accept another.
What are ALB load buttons and do I need one?
A load button screws into the centre thread and gives compression a defined, slightly domed contact surface to press against instead of a rigid flat interface. It matters more than its size suggests, because a flat-to-flat contact transmits any angular misalignment straight into the sensing element while a domed one lets the contact find its own centre. Published calibration testing found ball-type compression adapters improved reproducibility by 30 to 40%. For any serious compression work it is worth specifying one with the load cell rather than treating it as an afterthought.
Can the LPU measure tension as well as compression?
Yes — that is what "universal" refers to. The same centre thread accepts a tension fitting, so one sensor covers both directions. Compression calibration is included as standard, with tension calibration available as an option priced at $125 up to 20,000 lb and $250 above that. The hardware reads both ways regardless; what you are buying is a certificate covering the direction your results have to stand on.
What is the LPU made from and how well is it protected?
Heat-treated 17-4 PH stainless steel across every capacity, with bonded foil strain gauges sealed for protection against most industrial environments. That covers workshops, laboratories, production areas, dust and everyday moisture. It is not a hermetic specification, so regular wash-down, chemical exposure, marine air or sustained outdoor service should be described to us before you order rather than assumed to be within scope.
What is the safe overload, and what should I allow for?
150% of rated output, which describes what the sensor can experience without permanent damage rather than a working allowance. When sizing, think about the largest force your arrangement can credibly produce rather than the one you intend to apply — a hydraulic system that can develop more than your test needs, a fixture that could be driven past its stop, an operator applying a wrench. Sizing to sit close to your working load is the right instinct, but not so close that a foreseeable excursion goes past the overload figure.
Questions From The Field
Do I really need a load button, or can I press directly on the load cell?
You can press directly on it, and the reading will be worse than the sensor is capable of. Two flat faces meeting are almost never perfectly parallel, so contact happens at an edge or a high spot, the force arrives eccentrically, and the result varies with how the fixture happens to be sitting that day. A domed button turns that into a contact that finds its own centre, which is why calibration laboratories use ball adapters and why published testing credits them with a 30 to 40% improvement in reproducibility. If the measurement matters, fit one.
My readings changed after I remounted the load cell on a different plate.
That is expected behaviour on this geometry rather than a fault. The load reacts through the bolt circle into whatever the sensor is fixed to, so the mounting surface is part of the measuring structure — and a different plate means a different stiffness, a different flatness and a different bolt torque pattern. Check that the new surface is flat and rigid enough not to deform appreciably under load, tighten the hold-down bolts evenly rather than in sequence around the circle, and re-zero after mounting. If precision matters, keep the sensor with its plate.
The reading is steady and repeatable but consistently lower than I expect.
Suspect alignment before you suspect calibration. A force arriving at an angle to the measuring axis is reported reduced by the cosine of that angle, and because the geometry is unchanged from one test to the next the error is perfectly repeatable — which is what makes it so easy to miss. Check that the force acts perpendicular to the mounting face and passes through the centre of the sensor, and that nothing in the fixture is carrying a share of the load in parallel with the cell.
How tight should the centre adapter be?
Tight enough to be properly seated and positively locked, and the correct procedure depends on the capacity. Manufacturer guidance for low profile cells distinguishes smaller units, where the adapter is torqued and secured with a jam nut to documented figures, from larger ones, where a common practice is to pre-load the joint by pulling the adapter well beyond rated capacity before hand-tightening the jam nut so the connection is fully seated. Because the right figures depend on your model and your adapter, ask us rather than applying a general rule — and once set, leave it alone, since changing the adapter arrangement changes the measurement.
Which capacity should I choose if my forces vary a lot?
Size for the peak, but understand what that costs at the bottom end. Every accuracy figure is a percentage of rated output, so on a 50,000 lb cell a 0.1% specification is 50 lb whether you are measuring 50,000 lb or 500. If your working range genuinely spans a wide band, the fine capacity steps in this series exist to help you land as low as your peak safely allows, and the high output and low zero drift help keep the bottom of the range readable. Where a single sensor cannot cover the span sensibly, two cells is a legitimate answer and often the cheaper one.
Can I use an LPU as a reference to check other load cells?
The geometry is the right family — symmetric low profile designs are what force laboratories build reference and transfer standards on, precisely because they hold a calibration well and tolerate imperfect loading. Whether a given unit is suitable as your reference depends on the accuracy class you need to demonstrate, how it is calibrated and what traceability your quality system requires, so it is a question to put to us with those requirements stated. What is true generally is that a reference is only as good as the adapters and the loading arrangement used with it, and that stability between calibrations matters more than the headline accuracy figure.
How often should an LPU be recalibrated?
Annually is the usual starting point, adjusted for how hard the sensor works and how much rests on the result. The specification that really drives the interval is stability rather than accuracy: laboratories that track this find that drift between calibrations can inflate an overall measurement uncertainty several-fold, which is why a low zero drift figure is worth paying attention to. Any suspected overload, any impact, and any change to the adapters or mounting are reasons to recalibrate regardless of the schedule.
Does it matter which way up the LPU is mounted?
The sensing works the same either way, but the practical considerations differ. Mounted with the threaded face upward, debris and liquid can collect around the centre hole, which is worth avoiding in a dirty environment. Mounted underneath, the cable exit and its route need thinking about so the cable is not resting against the mounting face or being pulled by the fixture. In either case the load must still arrive perpendicular to the face and through the centre, and the mounting surface still has to be flat and rigid. If the orientation is unusual or the sensor is being mounted on something other than a plain plate, send us the arrangement.