PLC application in dedicated machine tools
This page explains when a PLC application in dedicated machine tools beats a full CNC controller, and when it does not. It is written for machine builders and process engineers who must pick a control layout for a two-station drilling or tapping machine. Read it to judge whether pulse output, table speed and hole-depth control are enough for your part.

In this article
- 1
- 2
- 3
- 4
- 5
- 6
- 7
Key takeaways
Why a two-station machine needs a PLC and not a CNC
A dedicated machine tool does one job for years. Take a double drilling machine with two stations: the front head drills one part while the inverted head drills a second part from below. Both parts sit on screw slide tables, and neither station interferes with the other. That layout is simple, but the control decision is not, because the two stations must position independently and hold depth.
The traditional answer was hydraulics. A cylinder pushes the head, a solenoid valve reverses it, and a mechanical stop sets the depth. That works until the drawing calls for a depth tolerance tighter than a few hundredths of a millimeter. Hydraulic oil heats up over a shift, the stop wears, and the depth drifts. Setting a feed rate is also awkward: you change a flow control valve and hope the speed stays put.
Servo motors on screw slides fix the depth problem, but they do not fix the control problem. Once you have two servo axes, something has to issue pulses, read the encoder or limit signals, and sequence the stations. A general-purpose CNC can do it, yet the machine only needs point-to-point moves. There is no contour, no interpolation between the two axes, and no tool radius compensation.
That is where programmable logic control earns its place. Many PLC CPUs ship with high-speed pulse outputs, so the same unit that runs the sequence also drives the two axes. The ladder program handles interlocks, clamp confirmation, fault recovery and cycle counting. The motion is only a small part of the code.
- 1Point-to-point onlyEach axis moves to a target and stops. No path blending is required.
- 2Two stations, one cycleThe stations run at the same time but never share a slide or a fixture.
- 3Depth is the critical dimensionHole depth repeatability drives the choice of servo over hydraulic.
Pulse output, table speed and the numbers that decide the layout
Start with the positioning spec. A typical dedicated drilling station calls for a slide table positioning accuracy better than 0.01 mm and a rapid positioning speed of at least 6 m/min. Those two numbers set the pulse frequency you need from the PLC output. The frequency depends on the screw lead and the encoder resolution of the drive, not on the PLC alone.
Work an example. Suppose the ball screw has a 10 mm lead and the servo drive is set to 10,000 pulses per revolution. To move the table at 6 m/min, the screw must turn 600 times per minute, which is 10 revolutions per second. At 10,000 pulses per revolution, the PLC must emit 100,000 pulses per second, or 100 kHz. That figure is the hard limit you check against the CPU datasheet before you buy anything.
A 20 mm lead screw halves the required frequency to about 50 kHz, but it also halves the mechanical resolution for the same encoder count. A finer lead gives better positioning resolution at the cost of a higher pulse rate. Pick the lead and the electronic gear ratio together, not one after the other.
Two axes on one CPU also share the scan cycle. If both heads rapid at the same moment, the output channels must sustain their pulse trains without jitter while the ladder logic keeps scanning. On many compact CPUs this is fine, but the total pulse load has to be added up, including any third axis you might add later for a tool changer.
- 1Frequency firstCalculate kHz from lead, target speed and drive pulse setting before selecting a CPU.
- 2Resolution secondA finer lead improves positioning resolution but raises the pulse demand.
- 3Count every axisAdd future axes to the pulse budget while the panel is still on the bench.
Station interlocks and cycle logic that the ladder must carry
Motion is the visible part of the machine. The invisible part is the interlock list, and it usually takes more ladder rungs than the positioning code. Each station needs a clamp-confirmed signal before the head can feed. If the clamp sensor drops out mid-cut, the axis must stop and retract, not continue to depth.
The two stations must also be prevented from starting a new cycle while the other is still feeding, if they share a hydraulic power pack or a chip conveyor. A simple ready-permissive bit handles that. Add a per-station cycle counter and a fault latch, and the operator can clear one station without touching the other.
Manual mode matters more than builders expect. Setup technicians need jog, single-step and return-to-origin for each axis. Write those as separate subroutines so a fault in automatic mode cannot leave a slide moving with the guard open. Tie the guard switch into the servo enable, not just into a warning lamp.
Keep the human-machine interface thin. Show position, depth offset, cycle count and active alarms. Depth offset is the one value operators adjust daily, so put it on the first screen with a numeric keypad and a hard limit on the range.
- 1Clamp before feedNo feed command until the clamp-confirmed input is true.
- 2Guard into enableWire the guard switch to servo enable, not to an indicator.
- 3One offset fieldExpose depth offset with a bounded range and a reset to zero.
What the drive train contributes to hole depth accuracy
The PLC commands position. The mechanical train decides whether the part comes out right. A ball screw with a small lead, a preloaded nut and a rigid thrust bearing will hold depth far better than a rolled screw with axial play. Check backlash on the slide before you blame the control.
Thermal growth is the second factor. A head that runs for six hours will extend, and a 300 mm steel column grows roughly 0.003 mm per degree Celsius. If the depth tolerance is ±0.01 mm, a 5 °C rise eats half of it. Reference the axis at the start of each shift and let the control compensate from a known point.
Servo tuning affects the stopping point. A drive with a soft position loop will overshoot and settle slowly, which shows up as depth scatter. Set the position gain so the axis settles within a few milliseconds after the in-position window opens, and keep the in-position width tight enough to catch a real stop.
For parts where the PLC-driven station is the first operation, leave stock for a later finishing pass. A dedicated drilling machine holds a good depth window, but it is not a jig grinder. Match the process to the tolerance, then measure the first article before the run goes long.
- 1Check backlash firstMeasure axial play on the slide before tuning the control.
- 2Re-reference each shiftThermal growth of a few degrees can consume the depth budget.
- 3Tighten in-positionA wide in-position window hides real stopping error.
Where a PLC layout stops making sense
A PLC layout stops making sense once the part needs a path. If the head must follow a curve, interpolate two axes at once, or apply cutter compensation, move to a CNC. Point-to-point logic cannot fake a contour, and the workarounds cost more than the controller.
It also stops making sense when the axis count climbs. Three or four coordinated axes, a rotary table, and a tool changer push a compact CPU past its pulse budget and past what a ladder program can keep readable. At that point the maintenance technician is debugging motion in ladder instead of reading a program.
Very high pulse rates are another limit. If the required frequency runs above what the chosen CPU can emit per channel, the drive will lose steps and the depth will scatter. Check the channel rating, not the total, because a two-channel output shares nothing when both run at once.
Finally, keep the PLC when the machine is one of many identical units. A compact ladder program with a small parameter set is easy to clone, easy to service and easy to hand to a local electrician. Complex CNC programs spread across a fleet are harder to keep in revision.
- 1Move to CNC for contoursAny interpolated path belongs on a motion controller.
- 2Watch channel limitsPer-channel pulse rate matters, not the combined total.
- 3Cloning favors PLCA small parameter set is easier to duplicate across a fleet.
Hydraulic, CNC and PLC control compared for a dedicated station
Read the row that matches your tolerance and axis count.
| Criterion | Hydraulic cylinder | General-purpose CNC | PLC with pulse output |
|---|---|---|---|
| Depth control | Mechanical stop, drifts with heat | Servo positioning, tight | Servo positioning, tight |
| Speed setting | Flow valve, indirect | Feed rate in program | Parameter in ladder or HMI |
| Axes supported | One motion per valve | Two or more, interpolated | Two independent pulse axes |
| Programming | None, hard wired | G-code with positioning blocks | Ladder plus HMI parameters |
| Simultaneous moves | Limited by one power pack | Path control, more complex | Two stations run in parallel |
| Best fit | Loose depth, low cost | Contours and many axes | Point-to-point, two stations |
Pick the controller that matches the motion, not the catalog
For two independent point-to-point axes with a depth window near ±0.01 mm, a PLC with high-speed pulse output is the cheaper and simpler choice. Once the part needs interpolation, cutter compensation or three or more coordinated axes, buy a CNC.
Questions engineers ask before the build
How many pulses per second does a 6 m/min table need?
It depends on the screw lead and the drive pulse setting. With a 10 mm lead and 10,000 pulses per revolution, the answer is about 100 kHz.
A 20 mm lead brings that down to about 50 kHz. Confirm the PLC channel rating at the higher of the two figures if you plan to change the lead later.
Can one PLC run both stations at the same time?
Yes, provided the CPU has two independent high-speed pulse channels. The stations do not share a slide, so the axes never need to be coordinated.
Add up the pulse load on both channels before you order. If a later tool changer adds a third axis, the budget changes.
Why not just keep the hydraulic cylinder and add a scale?
A scale tells you where the slide is, but it does not remove the valve dead band or the oil temperature drift. The control loop becomes a fight with the hydraulics.
Once you need a closed position loop, a servo and screw is the shorter path to a repeatable depth.
What positioning accuracy is realistic on a screw slide?
A preloaded ball screw with a rigid thrust bearing and a good servo can hold better than 0.01 mm at the slide, if the machine is referenced each shift.
Backlash and thermal growth usually dominate the error budget, not the pulse resolution of the PLC.
Where does the depth offset belong, in the ladder or on the HMI?
Put the value on the HMI and keep the ladder reading it as a register. Operators adjust depth daily, and a ladder edit for every tweak is a maintenance risk.
Bound the range in the HMI so a mistyped value cannot send the head into the fixture.
Does the PLC need a separate safety relay?
Yes. The PLC handles sequence and interlocks, not personnel safety. Guard switches and emergency stops should cut servo enable through a dedicated safety circuit.
Keep the PLC as a monitor that reports the state, not as the only device standing between an operator and a moving slide.
Send the drawing and the control spec together
We review the part, the depth tolerance and the axis layout, then quote the machining and the fixture work that supports the station. Quotation and free DFM analysis within 12 hours.
12-hour quote100% inspectionNo minimum order quantityNDA on request