A load cell turns weight into a small voltage. Load deflects a machined metal body, strain gauges bonded into it change resistance, a Wheatstone bridge reports the imbalance in millivolts, and a scale factor converts that into kilograms.
Anyone who has used a cheap kitchen scale has met the problem this particular design exists to solve. Put the same object in the middle of the platform and then near one corner, and the number changes. That is not a fault in the electronics; it is arithmetic. A plain bending beam responds to moment — force multiplied by the distance from where it is clamped — so moving the load further out bends the beam more and the scale reads high.
A single point design cancels that in the metal. The body is machined with a precisely shaped internal aperture that leaves thin flexure sections at four places, two toward the top and two toward the bottom. Under an off-centre load those flexures deform in a pattern that opposes the tilting moment, so the platform moves downward rather than tipping, and the bridge reports how much weight is present instead of where it happens to be sitting. Manufacturers describe this as moment cancellation, and it is the whole reason a single sensor can carry a platform that a plain beam cannot.
Which is why one sensor replaces four. The obvious way to build a platform scale is to put a load cell under each corner and add the readings. A corner-compensated single point cell does the same job with one sensor, one cable and one set of electronics — less to mount, less to wire, less to go wrong and less to calibrate.
But the compensation has a boundary, and it is a physical one. The geometry cancels the moment produced by a load placed off-centre by a certain amount. Push further out than the design allows and the moment exceeds what those flexures were shaped to absorb, and the position sensitivity comes back. That boundary is expressed as a maximum weighing platform size — and it is worth saying that many manufacturers give only a vague instruction to stay within recommended dimensions, without printing a figure. Ours is published, in inches, on this page.
The platform is therefore a specification, not a detail. On most sensors the fixture is something you design around the part. Here the fixture size is one of the part's stated limits, and it belongs in the design conversation before capacity does.
Sensors of this kind belong in electronic scales of every sort — kitchen, bathroom, postal and industrial — along with retail point of sale systems, parcel and postage weighing, food processing and packaging, pharmaceutical measurement, laboratory balances, custom weighing systems and OEM equipment where a weight reading is one function among several.
The bridge is conventional. What makes this part behave the way it does is a combination of that machined geometry and a compensation network built in behind it, and the resulting specifications are worth reading closely because several of them are unusually good.
Signal, and a detail that catches people out. The bridge is 350 ohms, excited at 10 VDC with 15 VDC the maximum. Rated output is not constant across the range: the 1 kg model gives 1.0 mV/V and the 2, 5 and 10 kg models give 1.5 mV/V. If you develop a product on one capacity and later change to another, that difference has to follow through into your firmware — it is a scale factor change, not a rounding error.
There is a compensation network inside, and it shows. The four-conductor wiring includes an internal temperature compensation and balance network, and the difference that makes is visible in three places on the specification table. The temperature coefficients are a couple of thousandths of a percent per degree Fahrenheit, which is the tightest figure anywhere in our load cell range. The zero balance is expressed as a small percentage of rated output rather than as a large raw voltage, so the unloaded output sits close to zero rather than needing a wide input window to accommodate it. And nonlinearity, hysteresis and nonrepeatability are each two hundredths of one percent — the best accuracy figures we publish on any product at any price.
That combination is unusual and worth understanding. A bare beam gives you whatever the metal and the gauges happen to do; a trimmed and compensated assembly gives you a part that behaves predictably across temperature and between units. It is the difference between a component you have to characterise and one you can design against, and on a weighing product where the sensor may sit inside a sealed housing for years, it is worth a great deal.
Mounting is not a formality. Manufacturer guidance on single point cells is consistent and specific, and every point matters here. Fix the mounting end securely to a rigid structure — a base that flexes puts strain into the beam that has nothing to do with the load. Attach the platform to the loading end only, never bridging both ends, or you short-circuit the flexure and the sensor cannot work. Do not over-tighten the mounting hardware into an aluminium body. Keep the platform within the published size. And route the cable so it is neither pulled nor vibrated. Mounting quality and measured accuracy are directly connected on this design.
Environmental protection is modest and honest. A moisture-proof sealant protects the gauge area, which suits the indoor, boxed-in installations these sensors normally live in — retail counters, laboratory benches, packaging lines. It is not a sealed or hermetic industrial specification, and equipment facing wash-down, condensation or outdoor exposure should be discussed with us rather than assumed to be covered.
Specifying a single point sensor runs in an order that surprises people the first time: the platform comes before the capacity. Our application engineers deal with these designs regularly and would rather help at the sketch stage than after a housing has been tooled.
Start with the platform, because it is the binding constraint. Work out the largest object your product must weigh and the size of the surface it will sit on, then check that surface against the published maximum. If your platform needs to be larger than the sensor supports, no amount of capacity will fix it — that is a different architecture, and it is far cheaper to establish now than after the industrial design is finished.
Then ask whether you need corner compensation at all. If your product applies force at a defined point — a probe, a plunger, a fixed contact — rather than through an open platform, you may not need it, and a simpler bending beam could be the right answer. Corner compensation is worth paying for precisely when the user decides where to put things, which covers most scales and very few test fixtures.
Then settle the range. Four kilogram steps cover 1 to 10 kg, and metric is usually the language an OEM weighing product is scoped in anyway. Take the smallest step that clears your heaviest object with room to spare, bearing in mind that the accuracy figures are proportions of rated output and that safe overload is 150% — a scale that a user might lean on deserves either headroom or a mechanical stop.
Take advantage of the constant footprint. All four capacities share the same dimensions, so one mechanical design covers the whole range. That is genuinely useful for a product family: build the housing once, and change only the sensor and the scale factor to offer a 2 kg and a 10 kg version of the same product. Remember that the 1 kg model has a different rated output, so it needs its own scale factor.
Design the mounting properly and the rest follows. Rigid fixed end, platform attached to the loading end only, sensible fastener torque into aluminium, and a cable route that nothing tugs on. These are the whole of the mechanical design, and they determine whether you get the accuracy the part is capable of.
Then decide what reads it. Many OEM designs take the bridge directly into their own converter, which suits a sensor whose zero sits close to nominal and whose temperature behaviour is already compensated. If you would rather buy that stage than build it, an amplifier signal conditioner module hands you a voltage or current output ready to use, and a digital display is hard to beat while a design is still on the bench. Cal-Teds plug and play is available with a connector carrying an IEEE 1451.4 TEDS memory, so a compatible instrument can read the sensor's calibration data rather than having it entered by hand — worth considering where units are interchanged between products or test rigs.
Tell us the platform size, the maximum weight, how the load will be placed and what the reading is for, and we will confirm the right capacity and whether this family suits your design. The whole range ships from stock, and there is a discount for universities, schools and research programmes.
LSP Series Load Cell Applications.
The Transducer Techniques LSP Series single point bending beam load cells address various OEM force measurement applications, particularly in the field of weighing and scales.
- Electronic Scales: LSP Series load cells are used in electronic scales of all kinds, including kitchen, bathroom, postal, and industrial scales.
- Weighing Machines: These load cells are used in weighing machines in factories, warehouses, and industrial settings.
- Retail Point of Sale Systems: LSP Series load cells are integrated into checkout counters and cash registers.
- Postal and Shipping Systems: These load cells are used in postage meters and parcel weighing systems.
- Food Processing and Packaging: LSP Series load cells provide product weight measurement in food processing and packaging operations.
- Pharmaceutical Manufacturing: These load cells are used for ingredient measurement and quality control in pharmaceutical manufacturing.
- Laboratory Balances: LSP Series load cells are used in laboratory balances for scientific research and analytical chemistry.
- Material Testing: These load cells are used in compression and tension testers.
- Customized Scales and Weighing Systems: LSP Series load cells are integrated into customized scales and weighing systems for industrial automation.
- Force Measurement in Production Processes: These load cells provide force measurement in manufacturing and assembly processes.
- Quality Control: LSP Series load cells are used for product verification in quality control operations.
- Load Monitoring: These load cells are used for load monitoring in elevators and conveyor systems.
The Transducer Techniques LSP Series single-point bending beam force sensor's combination of low cost, high accuracy, and resistance to eccentric loading makes it an ideal choice for OEMs and various industries requiring precise force and weight measurements.
Frequently Asked Questions
What does "single point" mean, and why does it matter?
It means the sensor is built to give the same reading wherever on its platform the load is placed. A plain bending beam cannot do that, because it responds to bending moment — force multiplied by distance — so an object set near the edge reads heavier than the same object in the middle. A single point body is machined with a shaped internal aperture that leaves thin flexure sections above and below, and those flexures cancel the tilting moment an off-centre load produces. The practical result is that one sensor can carry a whole weighing platform instead of needing four, one under each corner.
What capacities does the LSP Series cover?
Four, specified in kilograms: LSP-1 at 1 kg (2.2 lb), LSP-2 at 2 kg (4.4 lb), LSP-5 at 5 kg (11 lb) and LSP-10 at 10 kg (22 lb). All four share identical dimensions, which is genuinely useful for an OEM — one housing design covers the whole range, and you change only the sensor and the scale factor to offer several versions of the same product. Note that the 1 kg model has a different rated output from the other three, so it needs its own scale factor rather than a shared one.
What is the maximum weighing platform size?
7.87 inches square — 200 mm each way. That figure is the practical boundary of the corner compensation, and it is one of the most useful numbers on this page. Beyond it, the moment an off-centre load applies exceeds what the machined flexures were shaped to cancel, and the reading starts to depend on where the object sits again. Many manufacturers say only that you should stay within recommended dimensions without printing a number; check your platform against ours before committing to an industrial design, because this constraint is harder to work around than capacity.
How accurate is the LSP Series?
Nonlinearity, hysteresis and nonrepeatability are each 0.02% of rated output. Those are the tightest figures we publish on any load cell in the catalogue, at any capacity and any price, and on a low cost OEM part that is a genuinely unusual result. It comes from the combination of the machined single point geometry and the compensation network built into the wiring, which together produce a part that behaves predictably rather than one you have to characterise yourself.
What is the rated output, and why does it differ between models?
The 1 kg model is 1.0 mV/V nominal; the 2, 5 and 10 kg models are 1.5 mV/V. The difference reflects how the flexure is proportioned at the lowest capacity. For a designer the consequence is straightforward but easy to miss: firmware written around the 1.5 mV/V models will read half again high if a 1 kg sensor is fitted without changing the scale factor. If your product family spans capacities, make the scale factor a configurable value rather than a constant.
How stable is the LSP over temperature?
Compensated from 14° to 104°F with a safe range to 140°F, and with temperature coefficients of roughly two thousandths of one percent per °F on both zero and output — the best figures in our range by a wide margin. The compensated band reaching well below freezing suits cold rooms, vehicles and unheated spaces, and the tight coefficient means the reading holds up across a working day without constant re-zeroing. That stability comes from the internal temperature compensation and balance network rather than from the beam alone.
How is the LSP wired?
Four conductors, with an internal temperature compensation and balance network built in behind them. The network is why the zero balance specification is a small percentage of rated output rather than a large raw offset, which in turn means your front end does not need an unusually wide input window before gain — a helpful difference if you are taking the bridge straight into a converter on your own board. Excitation is 10 VDC with 15 VDC the maximum.
How should an LSP be mounted?
Fix the mounting end firmly to a rigid, flat structure, and attach the weighing platform to the loading end only. That second point is the one that gets missed: anything bridging both ends of the sensor bypasses the flexure and the cell simply will not read properly. Beyond that, avoid over-tightening fasteners into an aluminium body, keep the platform within the published size, and route the cable so nothing pulls or vibrates it. Manufacturer guidance on single point cells is consistent that mounting quality and achieved accuracy are directly connected.
What environmental protection does the LSP have?
A moisture-proof sealant over the gauge area, which suits the enclosed indoor installations these sensors normally live in — retail counters, laboratory benches, packaging equipment, postal machines. It is a lesser claim than a sealed or hermetic industrial specification and deliberately so, since the cost of the part reflects it. Wash-down, condensing environments, outdoor exposure or corrosive atmospheres should be described to us before ordering rather than assumed to fall inside what this family covers.
What is the TEDS option and when is it worth having?
It supplies the sensor with a connector carrying a TEDS memory to IEEE 1451.4, holding the unit's calibration data on the sensor itself so a compatible instrument can read it automatically instead of someone entering constants by hand. It is worth having where sensors are interchanged — a development bench running several capacities, a service operation swapping units in the field, or a production test rig where the wrong scale factor would go unnoticed. In a single fixed installation it earns less.
Questions From The Field
My scale reads differently depending on where I put the object. Is the sensor faulty?
Check the platform size first. Corner compensation works within the dimensions the geometry was shaped for, and ours is published as 7.87 inches square — a platform larger than that applies more moment at the edge than the flexures can cancel, and position sensitivity returns. If the platform is within size, look at the mounting: a platform that touches or is fastened to the mounting end as well as the loading end defeats the compensation entirely, and a base that flexes under load produces the same symptom. Faulty sensors are the last thing on this list, not the first.
The sensor barely responds to load at all.
Nine times out of ten something is bridging the two ends. A single point cell works because one end is fixed and the other moves; if the platform, a bracket, a screw or even a stiff cable connects across both, the load bypasses the flexure and very little strain reaches the gauges. Look for anything that touches both halves of the sensor, including at full deflection rather than only at rest. The other common cause is a platform fouling the housing somewhere, so part of the weight is being carried by the enclosure instead of the sensor.
Can I use a bigger platform than the specification allows if I accept some error?
You can, but understand what you are trading. The error is not random noise you can average away — it is a systematic function of where the object sits, so a user placing something in a corner gets a consistently wrong number rather than a slightly noisy one. That is a poor experience on a scale, because it looks like the product cannot make up its mind. If you genuinely need a larger platform, tell us the size and the accuracy you need and we will talk through the alternatives, which usually means a different architecture rather than a different sensor.
Does my product still need a tare and a span calibration?
A tare, yes — every weighing design needs one, and it also removes the weight of your own platform, which is generally larger than any sensor offset. Span calibration is a judgement call. The zero balance and temperature specifications on this part are tight enough that a design needing only a repeatable indication can often work from the nominal figures. Where your product displays a weight a user acts on, calibrate each assembled unit against a known mass anyway: it captures your platform, your mounting and your electronics as well as the sensor, and that combination is what your customer actually experiences.
How do I choose between the LSP and a plain bending beam?
Ask who decides where the load lands. If the user puts objects on an open platform — a scale, a checkout, a postal machine — you need corner compensation and the single point design is the right answer. If your product applies force at a fixed point through a probe, a plunger or a defined contact, the position is controlled by your mechanism and a simpler beam may serve for less. Tell us which describes your design and we will point you at the right family rather than the dearer one.
My readings drift for the first few minutes after switch-on.
Usually thermal settling rather than a sensor problem. Electronics inside a closed housing warm the air around them, and although the temperature coefficients on this part are very good, a warming enclosure combined with a platform and a fixture that all expand slightly will move the zero a little while everything equalises. The straightforward answer is to take the tare after the product has settled rather than immediately at power-up, and to add a slow auto-zero routine if the product is left running unattended. If the drift continues well beyond warm-up, look for something in the mechanism creeping under load instead.
Can I mount the sensor upside down or on its side?
The measuring principle does not care about orientation, but two practical things do. The weight of your own platform and fixture acts differently depending on which way the sensor faces, so the tare changes accordingly and needs to be taken in the final orientation. And the compensation is designed around load arriving perpendicular to the mounting face; a sideways installation where gravity acts along the beam rather than across it is a different problem and worth discussing with us before you build it. Vertical, platform on top, is what the design assumes.
What is the first thing to check if a unit stops reading correctly?
Measure the raw bridge output with no load and compare it against the zero balance specification, then measure the resistance across the excitation pair and compare it against the 350 ohm nominal. An open circuit or a wildly different resistance points at a damaged gauge or a broken lead — and on a product with a fine pigtail into a moving assembly, the lead is a likely suspect. If both are healthy, check the mechanics before the electronics: something fouling the platform, a fastener that has loosened, or debris under the platform accounts for far more field failures than the sensor itself does.