Test rigs that measure what makes your product good — on every part, or every prototype
Two kinds of rig, two different jobs
- Many measurement points, slow sweeps, full curves
- Linearity, hysteresis, repeatability, response
- Engineers change limits and sequences freely
- Output: an engineering report
- The few checks that decide good or bad
- Cycle time measured in seconds
- Operators select a model; limits are locked
- Output: OK / NOT OK, counts and a per-part record
Proportional valves and regulators
A proportional pressure regulator turns an electrical command — 0–10 V or 4–20 mA — into an output pressure. To validate one, the rig steps the command up through the range and back down again, and at each step waits for the output to settle and records it with a reference transmitter. From those two sweeps it works out:
- Linearity — how far the output wanders from the straight line between the end points.
- Hysteresis — the difference between the rising and the falling sweep at the same command.
- Repeatability — how closely a second sweep lands on the first.
- Analog output — whether the device's own feedback signal agrees with the measured pressure.
- Pressure characteristics — how the output holds as supply pressure or flow changes.
Rigs like this go wrong in predictable ways: settle times too short for the output to stop moving, a reference transmitter no more accurate than the part it is judging, and command steps that overshoot and hide the very hysteresis they are meant to measure. The sequence is designed around those three before anything else.
What a rig can measure
| Pressure | Reference transmitters sized to the tolerance; static readings, sweeps and hold tests. |
|---|---|
| Flow | Flow meters on the outlet or supply, for flow-versus-pressure characteristics. |
| Analog signals | Commands generated and feedback measured: 0–10 V and 4–20 mA, in and out. |
| Switching and response | Valve switching, pressure switch set points and response time to a step command. |
| Leak | Pressure decay on the part or across a seat — see the leak testing page. |
| Life cycles | Endurance runs that cycle a valve or cylinder and stop with the cycle count at the first failure. |
What the engineer and the operator get
- A recipe per model: pressures, sweep points, limits and timings.
- Automatic sequence with the measured value and limits shown for every step.
- A pass or fail per step, so a reject says which test it failed.
- Results stored against the part serial; curves exported for validation reports.
- Calibration pages for the reference instruments, with protected settings.
Send photos of the bench and the test sheet your technicians fill in today. That is the specification — we will come back with a concept and a budget range.
Test rig questions
How accurate does the rig's own instrumentation need to be?
Clearly better than the tolerance you are judging. A common rule is that the measuring chain should be at least four times — ideally ten times — more accurate than the tolerance band, or the rig will pass bad parts and reject good ones near the limits. We size the transmitters and flow meters to your tolerances and state the ratio in the concept.
Can one rig test several models?
Yes. Each model is a recipe: test pressures, sweep points, limits and timings. The operator selects the model, or scans it, and the rig runs that model's sequence. Fixture changes are kept to quick-change adaptors where the parts allow it.
Can the rig produce a test report for our customer?
Yes. Every test step's measured value and limits can be stored against the part serial, and printed or exported as a per-part report. For validation rigs, the full curves can be exported for the engineering report.
We test on a manual bench today. Can you automate it?
Usually, and the existing bench is the best specification there is: it shows what is measured, in what order and what a technician looks at to decide. We automate the sequence and the decision, keep the instruments where they are good enough, and replace the ones that are not.
Validation rig or end-of-line tester — can one machine do both?
Sometimes, but they pull in different directions: validation wants many points, slow sweeps and every curve; production wants the few checks that matter, in seconds. Often the best answer is a validation rig first, and an end-of-line tester built from what validation showed actually matters.
Testing for leaks specifically? See leak testing machines.
Have a job that a machine should be doing?
Send the part and the target output. You get a concept, a sequence and a budget range back — free, and with no obligation to build it with us.