How Does a Servo Hydraulic System Work? Closed-Loop Pressure & Flow Explained

If you have ever watched a hydraulic cylinder creep when the directional valve is centered, or watched a clamping pressure sag every time a mold closes, you have already seen what happens when there is no closed loop. A Servo Hydraulic System is the engineering answer to that problem: it replaces the brute-force "oversized pump plus throttling valve" arrangement with a coordinated electrical and hydraulic package that asks the pump for exactly the flow and pressure it needs, then proves it with sensors.

The short version: an electronic Servo Drive commands a servo motor to spin a fixed-displacement hydraulic pump at whatever speed is required to satisfy the current demand, and pressure / flow / position sensors verify that the result matches the setpoint. When the cylinder is idle, the pump idles too. When the cylinder is working, the pump works at exactly the right speed. That continuous adjustment is what makes the system both precise and efficient. Everything else in this article is the longer version of that sentence.

The mental model: hydraulics is a closed-loop system, not a plumbing problem

The most common mistake we see in early-stage machine designs is to treat hydraulics like plumbing — pick a pump, pick a valve, plumb it together, and assume the result will behave. In practice, every hydraulic actuator is part of a control loop whether the designer planned it or not. Because pressure and flow are continuous variables that drift with temperature, oil viscosity, leak rate and external load, a hydraulic system without feedback is fundamentally an open-loop system with unknown error. A servo hydraulic system simply makes that loop explicit: it adds the sensors, controller, and variable-speed pump that let the machine close the loop on purpose, instead of by accident.

That is also why the same machine can feel "soft" when cold and "firm" when warm. Oil viscosity drops by roughly a factor of five between −10 °C and +60 °C, and the leakage coefficient of a typical control valve varies accordingly. The servo controller compensates for this continuously; a fixed-displacement system cannot. If you want a deeper dive on why feedback matters in motion control, the IEEE Control Systems Society and ASME both publish introductory material on closed-loop dynamics that translates well to hydraulics.

Closed-loop pressure control: how the system "holds" a setpoint

The first job of the closed-loop is to keep the working pressure stable under disturbance. Because the controller can read the actual pressure from a transducer and compare it to the commanded pressure, it can correct the pump speed within milliseconds whenever the load changes. This is the same principle a cruise control uses on a car: measure, compare, correct, repeat.

1. Command. The machine controller (PLC / IPC) sends a target pressure to the servo drive, for example 140 bar for clamping.
2. Measure. A pressure transducer (typically 0–250 bar or 0–400 bar, 4–20 mA or IO-Link) on the actuator port reports the actual pressure every cycle.
3. Compare. The servo drive's firmware subtracts measured from setpoint to get the error.
4. Correct. A PID algorithm converts that error into a new motor speed command.
5. Act. The servo motor adjusts pump speed (or in some architectures, swash-plate angle) within a few milliseconds.

Step response — the time it takes to reach and stabilize at the setpoint — is the headline number most engineers compare when they evaluate a servo hydraulic system. In our own test stands we routinely see a pressure step of 0 → 140 bar settle inside 50–80 ms on a 75 kW unit; that figure is governed as much by the pump's volumetric stiffness as by the controller tuning.

Closed-loop flow control: how the system delivers the right amount of oil

Flow is the second half of the closed-loop and the part that controls velocity. Because volumetric flow Q equals pump displacement V_g multiplied by motor speed n, controlling motor speed is, in effect, controlling flow. The relationship is linear enough that, for most industrial purposes, we can think of servo motor RPM as a direct flow command.

This is also where servo hydraulics pulls decisively ahead of proportional-valve throttling. With throttling, you generate the full pressure and then throw away the excess across a valve to meter flow. That excess is heat. With servo pump control, you only generate the flow you need — and the pressure loss is whatever the load actually requires, not whatever the pump is forced to produce. That is the physical origin of the energy savings people quote for servo hydraulic systems.

Three sub-cases matter in practice:

• Acceleration phase. High flow demand — controller ramps pump speed to maximum. Pressure falls to the minimum needed for acceleration, so the motor is not fighting itself.
• Holding phase. Near-zero flow demand — controller drops pump speed to the minimum needed to maintain leakage makeup. Power draw collapses to a small fraction of full-load.
• Deceleration phase. The cylinder is now acting as a pump. Most integrated units use a check valve and accumulator path to recover that energy rather than burn it as heat; the controller simply lets pump speed fall until it matches the natural deceleration curve.

Inside an integrated electro-hydraulic servo unit

An integrated electro-hydraulic servo unit is the modern packaging of everything described above. Instead of a motor bolted to a tank, with hoses running to a valve manifold and pressure switches, the entire loop is one frame:

• Servo motor. Permanent-magnet synchronous, encoder-equipped, typically IE4 / IE5 efficiency class. Spins the pump directly, no coupling alignment needed.
• Servo pump. Internal gear or axial piston, fixed displacement. Volumetric efficiency above 95% at rated conditions; that is what makes the loop fast.
• Reservoir & filtration. Built into the frame, sized for the application. Return filtration in the 10–25 μm absolute band is typical.
• Drive electronics. The servo drive sits on top of the motor or in an adjacent compartment, accepting fieldbus commands and running the inner current / speed / pressure loop in firmware.
• Sensors. Pressure transducers on the pump port and on the actuator line; flow meter or speed-derived flow; optional temperature and level switches in the tank.
• Cooling path. Air-cooled for small frames, water-cooled (or oil-cooled through the housing) for higher continuous power. Because the system only generates heat when it works, peak cooling demand is also lower.

The practical consequence is short plumbing. Because the energy source sits inside the same steel frame as the sensors and the drive, the hydraulic lines carrying high-pressure oil are short, the electrical cables carrying control signals are short, and the closed loop is fast. That is the single biggest reason integrated units have displaced discrete component stacks in modern plastic-injection-molding machines.

Open-loop vs closed-loop: what you actually give up

The honest way to think about this is to ask what you lose without feedback. With a fixed-displacement pump and proportional valve, you cannot independently control pressure and flow, you cannot compensate for viscosity drift, and you cannot correct for leak-down on a held cylinder. Because the servo system reads the real world every few milliseconds, it can do all three at once, and that combination is what produces repeatable motion.

Those differences show up as energy bills, cycle times, and scrap rates — not as abstract numbers. In injection molding, energy savings in the 50–80% band vs. a traditional proportional-valve machine are commonly reported in industry surveys; cycle time reductions of 5–15% are typical because acceleration and deceleration phases no longer wait for valves to open.

Tuning a servo hydraulic loop in practice

The closed loop is only as good as its tuning. Three rules we apply when commissioning a new machine on the test stand at our Ningbo facility:

• Start with the inner loop, then the outer loop. Current loop first (firmware-tuned at the factory), speed loop next (drive-side auto-tune usually gets within 80%), pressure loop last (requires the actuator attached).
• Tune for the worst case, not the average case. The pressure loop is most stressed during deceleration and holding. If you only tune on acceleration, you will chase oscillations forever.
• Verify with a step response, not a sine sweep. A 10–90% step test tells you settling time, overshoot and steady-state error in one number. Sine sweeps are diagnostic, not acceptance.

Our application team can also run a remote commissioning session with your machine builder if you want a second pair of eyes on the loop. Reach out via the VICKS engineering team page to set one up.

Where servo hydraulic systems make the most sense (and where they don't)

Because a servo hydraulic system is more expensive per kilowatt than a fixed-displacement system, it is not the right answer everywhere. The economics turn in its favor when one or more of these is true:

• The duty cycle is cyclic with significant idle time. Injection molding, die casting, press lines, and test rigs all qualify.
• Energy cost is material. Continuous-running hydraulic units in 24/7 plants add up fast; the servo unit pays back the incremental capex.
• Process quality depends on pressure repeatability. Closed-loop pressure holds ± 0.5% in the band where proportional valves drift ± 5%.
• Noise or oil cleanliness matters. Because the pump idles, integrated units run noticeably quieter and keep oil cleaner than discrete component stacks.

Where they do not make sense: continuous full-flow duty (a hydraulic press running flat-out all day), applications where the cost of an encoder failure is unacceptable without a redundant path, and ultra-low-power applications under ~1 kW where a small inverter-driven gear pump is cheaper. If your use case is on the borderline, talk to us — the right answer depends as much on the duty cycle as on the nameplate power.

FAQ