Strain gauge transducers do not generate anything on their own; they modulate a voltage that something else supplies, and a bridge excitation supply is the source of that voltage. Because a bridge's output is a proportion of its excitation rather than a fixed voltage, the quality of that supply shows up directly in the measurement — which is why a supply intended for transducer work is specified far more tightly than one intended to run electronics.
This one is built to go inside something. Its own description is a chassis mountable power supply package, aimed at dedicated microprocessor systems, instrument design and test equipment. It is flat, it bolts down, its connections are on a recessed barrier strip, and its output is not adjustable.
That last point is the one people query, and it is the wrong way round: for equipment you are building, a fixed output is a feature.
There is nothing to knock out of adjustment. An adjustable supply mounted inside a machine is a setting, and settings get changed. A technician troubleshooting something else turns it to see what happens. A trimmer drifts under vibration. Somebody borrows the unit, adjusts it, and puts it back. A fixed output cannot be knocked off value because there is no value to knock — the excitation becomes a constant of the design rather than a variable in the field.
And it is fixed at the right constant. Ten volts is the standard excitation figure across this industry. It is what the great majority of our load cells and torque sensors are specified at, and what most calibration certificates are written against. Fixing the output there is not a restriction so much as a decision not to make a decision that already has a correct answer — with the practical benefit that your certificate applies directly, with no correction factor for anybody to forget.
The third reason is the one that matters most to anyone building more than one of something. If you produce fifty machines around fifty adjustable supplies, you have fifty slightly different excitation voltages unless somebody sets and verifies every single one, records it, and it stays set. If you produce fifty machines around a fixed supply, you have fifty identical machines. Consistency between units stops being a process you have to maintain and becomes a property of the parts you bought. For an equipment manufacturer that is worth considerably more than the flexibility being given up.
What you gain alongside it is current. Four hundred milliamps is a great deal more than an adjustable bench supply of this kind usually provides, and it changes what one unit can do: a platform on four load cells, a rig with eight channels, a machine carrying a dozen transducers — all from a single supply.
And feeding them all from one source has a quality benefit that is easy to miss. Every sensor sees exactly the same excitation, so if that voltage drifts a little with temperature, every channel drifts together. When you are comparing one channel against another — corner against corner on a platform, one station against another on a rig — a shared drift largely cancels itself, where separate supplies would let the channels wander apart independently. On a multi-channel measurement that is a real advantage, and it is not on the page.
The signal path is short: line in one end, regulated low voltage DC out the other. What is worth knowing about this particular unit is what its numbers permit and where its limits sit.
Four hundred milliamps at ten volts translates into a specific number of sensors, and it is a division worth doing on paper. A standard 350 ohm bridge draws a little under thirty milliamps at ten volts, so around fourteen of them fit within the supply's capability. A 700 ohm bridge draws half that, so twice as many. A 120 ohm bridge is the demanding case at over eighty milliamps each, which brings you down to four. Work out what one of your sensors takes, divide, and then deliberately do not use all of it — a supply running at its ceiling regulates less well than one with room to spare, and leaving margin is what allows somebody to add a channel next year without redesigning the system.
Regulation of five hundredths of one percent is the figure that matters for accuracy. It describes how far the output can move as the line voltage varies or as the load on the supply changes — and because a bridge's output is proportional to its excitation, that same fraction appears in every reading. The load-change half is the one to think about in a multi-sensor system: if channels can be connected and disconnected while the equipment runs, the supply's regulation is what stops the remaining channels shifting when one is unplugged.
Temperature is specified at fifty parts per million for each degree Celsius. Expressed as a percentage that is 0.005 per degree, so a swing of twenty degrees shifts the excitation — and every reading taken through it — by around a tenth of one percent. Inside a piece of equipment that is a real consideration rather than an academic one, because the inside of an enclosure warms up when the equipment is switched on and keeps warming for some time afterward.
Ripple is one millivolt, and it is worth being straight about the comparison. Measured against a ten volt output that is a hundredth of one percent riding on everything you read — small, and for the overwhelming majority of applications entirely irrelevant. It is, however, roughly double the figure of the adjustable supply in this range. That is the trade you are making for the extra current and the fixed output, and it is a sensible trade in most cases. If your measurement is at the very quietest end of what a strain gauge system can do, the other supply is the better choice and we will say so.
The connections are on a barrier strip with recessed terminals, which is a detail worth appreciating. Recessed means the live parts sit below the surface rather than proud of it, so a dropped screwdriver, a stray strand of wire or a probe that slips cannot bridge across two terminals. In a unit destined to live inside equipment that people open — for maintenance, for fault finding, for adding a channel — that is exactly the right way to terminate a supply. It also makes for a mechanically secure connection that will not work loose under vibration, which a plug and socket may.
Isolation from the line is well above a thousand volts, which is what you want in something being designed into equipment that may need to satisfy a safety standard, and the operating temperature range extends comfortably above what most enclosures reach.
One thing this unit does not offer is sense connections, and that is coherent rather than an omission. Sense wiring exists to correct for voltage lost in a long cable between supply and sensor. This supply is designed to sit inside the same box as the instrumentation, inches from the sensor connector, where that loss is negligible. If your transducer is a long way from wherever the supply will live, that is a different requirement and there is a different unit in the range for it.
Two supplies in this range solve two different problems, and the page for each does not mention the other. Here is the short version, and our applications engineers will confirm it for your case.
Choose this one when the supply is going inside something. A chassis mounting profile, a recessed barrier strip, plenty of current and a fixed standard output are all the characteristics you want in a component built into an instrument, a test system or a machine — particularly if you are building more than one of them and want every unit to behave identically.
Choose the adjustable supply when the voltage has to match something you do not control. Legacy equipment specified at five or twelve volts, a certificate written at an unusual figure, a mixed bench where different transducers want different excitation, or a situation where you need to deliberately reduce excitation to limit self-heating in a small sensor. The adjustable unit also has slightly lower ripple and offers sense connections for a remote transducer.
Count your sensors, then leave margin. Work out the current one bridge draws at ten volts, multiply, and stop well short of the limit. Margin buys you better regulation, a cooler supply and the ability to add a channel later without a redesign. It is the cheapest thing in the whole system to over-specify.
Think about where the heat goes. A supply of this kind converts the surplus voltage into heat rather than switching it away, and the more current you draw the more heat there is. Inside a sealed enclosure that heat has nowhere to go except into the air around your instrumentation, so give it a position with some airflow rather than burying it beneath a cable loom. It matters twice over here: once for the supply's own stability, and once because everything else in that enclosure is now living in a warmer place.
Decide early where the supply sits relative to the sensors. Inside the same enclosure as the instrumentation is the intended arrangement and needs nothing special. A supply in a cabinet feeding transducers out on a machine is a different design and wants either the sense-equipped unit or a deliberate decision to calibrate the system as installed and never change the cabling.
If you are building equipment, write the excitation voltage into your documentation. Ten volts is not merely what this supply produces — it is the number your calibration, your scaling and your instrument setup all assume. Recording it alongside the sensor's sensitivity and the instrument's configuration means that in five years somebody replacing a part has all three numbers and no guesswork.
Tell us the transducers you are powering, how many there are, whether the supply lives with the instrumentation or across the plant from it, and whether you are building one system or fifty, and we will confirm which supply suits.
PSM-F10 Bridge Excitation Supply Applications.
The Transducer Techniques PSM-F10 is a chassis mountable power supply package providing a fixed 10 Vdc at up to 400 mA, suited to dedicated microprocessor systems, instrument design and test equipment, and usable as a supplemental bench or laboratory supply.
- OEM Instrument Design: A chassis mounting supply built into equipment, providing the standard 10 volt excitation as a fixed constant of the design.
- Multi-Sensor Systems: Four hundred milliamps powers a platform, a rig or a machine carrying many transducers from one source.
- Multi-Channel Comparison: Every channel sees the same excitation, so supply drift is common to all of them and largely cancels between channels.
- Production Consistency: A fixed output means every unit of a manufactured system has identical excitation without a setting to verify.
- Test Equipment Manufacture: A compact regulated supply for test systems being designed around bridge transducers.
- Microprocessor and Embedded Systems: Provides a clean regulated rail for dedicated processor systems as well as transducer excitation.
- Bench and Laboratory Support: Serves as a supplemental supply where a stable ten volt source is needed alongside existing equipment.
- Platform and Vessel Weighing: Powers the several load cells of a platform or vessel installation from a single regulated source.
- Panel and Enclosure Installation: A flat profile and recessed barrier strip suit permanent mounting inside a cabinet or instrument case.
- Data Acquisition Support: Supplies excitation where acquisition hardware provides differential inputs but no bridge power.
The PSM-F10 works with our full range of load cells, torque sensors and pressure transducers, providing fixed 10 Vdc excitation to full Wheatstone bridge transducers.
Frequently Asked Questions
What is the PSM-F10 designed to go into?
Equipment rather than a bench, although it works perfectly well on one. Its own description is a chassis mountable power supply package for dedicated microprocessor systems, instrument design and test equipment, and everything about it reflects that: a flat profile that bolts down inside a case, a recessed barrier strip for permanent wiring, generous current, and a fixed output. If you are designing an instrument, a test system or a machine that has to excite transducers, this is the component form of that requirement. It will also serve as a supplemental laboratory supply where a stable ten volt source is wanted.
Why is the output fixed rather than adjustable?
Because in built-in equipment that is an advantage rather than a limitation. An adjustable supply inside a machine is a setting, and settings get moved — by a technician investigating something else, by a trimmer drifting under vibration, by anyone who borrows the unit. A fixed output cannot be knocked off value because there is nothing to knock. It also removes a step from your build and your service documentation: there is no voltage to set at manufacture, verify at test, or check during a repair. The excitation becomes a property of the parts rather than a process you have to maintain.
Why ten volts specifically?
Because it is the standard, and standardizing on the standard has real value. Ten volts is what the great majority of our load cells and torque sensors are specified at, and what most calibration certificates are written against. Fixing the output there means your certificate applies directly with no correction factor, your instrument scaling assumes a familiar number, and a replacement sensor ordered in five years will expect exactly what your equipment provides. It is a decision not to make a decision that already has a correct answer.
How many sensors can 400 milliamps power?
Divide the available current by what one bridge draws at ten volts. A 350 ohm bridge takes a little under thirty milliamps, so roughly fourteen of them fit. A 700 ohm bridge draws half as much and doubles that figure. A 120 ohm bridge is the demanding case at over eighty milliamps each, which brings the count down to about four. Deliberately do not use the whole budget: a supply working at its ceiling regulates less well than one with headroom, runs hotter, and leaves you nowhere to go when somebody wants an extra channel.
What does "chassis mountable" mean in practice?
That it is designed to be fastened down inside something rather than sat on a surface. The package is deliberately flat — well under an inch and a half tall — so it fits under a chassis plate, behind a panel or in a shallow enclosure where a taller unit would not. That profile is the main physical difference between this and a bench supply, and it is usually the thing that decides which one fits your design. Allow for access to the barrier strip when you position it, since somebody will eventually need to get a screwdriver to it.
What is a recessed barrier strip and why does it matter?
A terminal block whose connections sit below the surface rather than standing proud of it. The benefit is straightforward: a dropped screwdriver, a stray strand of wire or a probe that slips cannot bridge across two terminals, which is exactly the protection you want in a unit living inside equipment that people open for maintenance and fault finding. It also gives a clamped connection that will not work loose under vibration in the way a plug and socket can. Small detail, and the right one for a supply intended to be installed once and left alone.
How does it compare with the adjustable supply in the range?
They solve different problems. This one gives you a fixed standard voltage, considerably more current, a chassis mounting profile and a recessed terminal strip — the characteristics of a component built into equipment. The adjustable unit gives you a selectable voltage from a few volts to fifteen, slightly lower ripple, and sense connections for a transducer some distance away — the characteristics of a bench instrument. Choose this one for equipment you are building, especially if you are building several. Choose the other when the voltage has to match something you do not control, or when the sensor is remote.
Is it only for load cells?
No. It suits any full Wheatstone bridge transducer — load cells, torque sensors and pressure transducers alike — and its own description also names dedicated microprocessor systems and instrument design, so it will serve as a clean regulated rail for electronics as well as for excitation. If you are powering both from the same unit, keep the arithmetic honest by adding up everything drawing from it, and think about whether digital circuitry sharing a supply with your excitation is a good idea in your particular design.
How stable is the output?
Line and load regulation within five hundredths of one percent, temperature coefficient of fifty parts per million per degree Celsius, and ripple of one millivolt. Translated into what they do to a measurement: the regulation figure is how much your reading can move as line voltage or loading changes, the temperature figure is about a tenth of a percent across a twenty degree swing, and the ripple is a hundredth of one percent riding on everything. All three are comfortably below the drift of the transducers being powered, which is where a supply belongs.
What does it need electrically and physically?
Standard line voltage at either mains frequency, and a flat area roughly four inches by three to bolt it to, under an inch and a half tall. Isolation from the line is well above a thousand volts, which is what you want in something being designed into equipment that may need to satisfy a safety standard. The operating temperature range extends comfortably above what most enclosures reach. The practical requirement beyond that is airflow, because a supply of this type turns its surplus voltage into heat rather than switching it away.
Questions From The Field
I am building fifty machines. Which supply should I standardize on?
This one, and the reason is consistency rather than cost. Fifty adjustable supplies means fifty excitation voltages that somebody has to set, verify, record and keep set — and any unit that comes back for repair has to have that value re-established before it goes out again. Fifty fixed supplies means fifty identical machines with no setting to get wrong, no step in your test procedure and nothing to drift. Write the excitation voltage into your documentation anyway, alongside the sensor sensitivity and the instrument scaling, so that whoever services machine number thirty-seven in eight years has all three numbers.
Can I adjust the output if I need a different voltage?
No — this unit's output is fixed, which is the point of it. If your application genuinely needs a different excitation voltage, the adjustable supply in this range covers roughly four to fifteen volts and is the right part. Before you conclude that you need it, though, check why: if the reason is a calibration certificate written at a different figure, the sensitivity in millivolts per volt is largely independent of excitation, so it may simply be a matter of scaling your instrument for ten volts instead. Send us the certificate and the instrument and we will tell you which situation you are in.
My sensor is a long way from where the supply will sit.
Then think carefully, because this unit has no sense connections and cable resistance between a supply and a bridge makes the sensor receive less than the supply produces — which makes readings proportionally low in a way that looks like a calibration error. Three options. Move the supply nearer the sensor, which is the intended arrangement and the best answer. Use the sense-equipped supply in this range instead. Or calibrate the system exactly as installed and never change the cabling, which works but is fragile. Tell us the distance and the cable and we will advise.
Can I power the sensors and the rest of my electronics from the same unit?
Electrically it will do it, and the description names microprocessor systems as an intended use, so it is a designed-for arrangement. Two things to weigh. First the arithmetic: add up everything drawing current, including the bridges, and stay well inside the limit. Second, and more subtly, digital circuitry draws current in bursts as it switches, and those bursts appear on the supply rail as small disturbances that your excitation then carries into the bridge. In a careful design that is manageable with decoupling and layout; in a marginal one it becomes noise on your measurement. If your measurement is demanding, a separate supply for the excitation is the safer architecture.
My readings drift for the first half hour after switch-on.
Almost always thermal, and inside an enclosure it can come from three places at once: the supply itself warming, the transducer's bridge self-heating, and the enclosure's internal air rising as everything in it dissipates power. All three move the reading in the same direction and none of them is a fault. The practical answer is to allow a warm-up period before recording data you intend to trust, and to zero after it rather than before. If you are designing the equipment, position the supply where its heat can escape rather than where it will bake the instrumentation, and expect the enclosure to stabilize considerably later than the electronics do.
I need more current than 400 milliamps.
Tell us the arrangement, because there is usually a better answer than a bigger supply. Splitting the sensors across two supplies works but gives up the advantage of every channel sharing one reference, so if channel-to-channel comparison matters, say so. Higher resistance transducers draw proportionally less current and may solve the problem at the sensor end. And on a system with a great many channels, individual conditioners that each generate their own excitation are often the cleaner architecture than one large supply feeding everything. Send us the sensor count and resistance and what the system has to do.
Does the supply need fusing or protection?
Follow your own standards for the line side, which will generally specify protection appropriate to the wiring rather than to the device, and remember that a unit built into equipment falls under whatever approval regime that equipment has to meet. On the output side, the thing worth thinking about in a multi-sensor installation is what happens if one sensor cable is damaged and shorts — whether the rest of the channels keep working or the whole system goes down with it. If that distinction matters to your design, raise it with us early and we will go through the options.
Which of the two supplies is quieter?
The adjustable one, by roughly a factor of two on ripple. Put that in proportion before it decides anything: one millivolt against a ten volt output is a hundredth of one percent riding on your reading, which is far below the uncertainty of any transducer you could connect and irrelevant to the overwhelming majority of applications. It becomes worth caring about only at the very quietest end of strain gauge measurement, where every contribution is being counted. If you are working there you probably already know it — and if you are not sure, describe the measurement and we will tell you whether the difference is one you would ever see.