Capacitive Torch Height Control Can Track Oxy-Fuel Plate

A capacitive torch height controller can be a useful way to manage oxy-fuel standoff when the plate surface varies, but the deciding rule is whether its sensor tracks the actual plate reliably across the machine’s cutting paths. In the described setup, slag under a plate can raise a corner by at least 1/4 inch, while the operator currently corrects height manually with the up and down buttons. Validate sensing and motion on the real plate, including cutouts, before relying on automatic height correction.

Reading the height-control problem

Separate surface-following from process control. A torch height controller (THC) moves the torch vertically to maintain a sensed distance or height reference. It does not remove slag, flatten a plate, or set the correct oxy-fuel cutting parameters.

The identified problem is mechanical: a slag chunk under a 4 × 8 or 5 × 10 plate can lift a corner by 1/4 inch or more. That changes torch-to-plate distance as the machine crosses the raised area. Whether that change demands automatic correction depends on the torch’s required standoff and the machine’s ability to follow the surface; read the torch/process documentation for the target rather than choosing a value by guesswork.

Manual up/down correction can accommodate a slow, visible change, but it depends on the operator noticing and responding. Automatic correction is useful only when the sensor sees the plate and the Z axis can follow without excessive delay, oscillation, or collision risk.

Capacitive sensing and its limits

A capacitive sensor detects a change in electrical capacitance as the distance and geometry between its sensing element and a conductive workpiece change. The controller uses that signal as a height reference and commands the torch’s vertical axis. Unlike plasma THC systems that use arc voltage as a height signal, oxy-fuel cutting does not provide that same arc-voltage feedback; a capacitive system is therefore a separate sensing approach.

Capacitance depends on more than torch-to-plate distance: the sensing element’s position, nearby conductive objects, workpiece geometry, wiring, grounding, and electrical noise can affect the signal. An edge or opening changes the conductive area beneath the sensor. A sensor may continue to work around a cutout under some configurations, but that does not establish reliable tracking for every sensor mounting, path, or opening size. The controller’s signal and vertical motion must be checked on the intended geometry.

Surface following also has a mechanical limit. The sensor cannot correct a sudden step if the axis cannot respond in time or travel far enough. Confirm the machine’s available Z travel and response behavior from its documentation and commissioning checks; no universal response time or correction range applies to this installation.

Checks before installing automatic correction

1. Record the process reference. Identify the torch maker’s specified standoff and operating setup. Expected result: a documented target height and a repeatable manual setup, not an assumed number.
2. Inspect the plate and support. Remove loose slag where practical, then measure or observe the lift at the raised corner. Expected result: the plate’s high and low regions are known before sensor performance is judged.
3. Confirm sensor compatibility. Check the controller documentation for workpiece material, sensor mounting, grounding, cable routing, and permissible operating conditions. Expected result: the installation matches the specified sensing arrangement.
4. Test the input while stationary. Move the sensor or plate through a small, controlled height change and monitor the controller’s height indication. Expected result: a stable signal that changes consistently with distance and returns to baseline, without unexplained jumps.
5. Check the Z axis separately. Use the controller’s documented manual or test mode to command small up/down movements. Expected result: correct direction, smooth motion, and no binding; stop if commanded motion does not match the control indication.

Commissioning the controller on the machine

1. Install and wire the sensor according to its manufacturer’s instructions. Keep sensing conductors separated from sources of electrical interference as specified in that documentation, and verify the required workpiece reference and grounding.
2. With cutting disabled, establish the sensor’s baseline over solid plate. Use the controller’s documented calibration or sensitivity procedure. Expected result: repeatable height indication at the same physical gap.
3. Run a controlled dry traverse over flat plate and then over the known raised region. Observe the height signal and Z motion. Expected result: the sensor detects the change and the axis follows smoothly without hunting or approaching a collision.
4. Repeat the dry test near an edge, around a cutout, and across the intended contour transitions. Expected result: no false height command or loss of tracking that drives the torch toward the plate or away from the process reference.
5. Make a supervised test cut only after dry tracking passes. Compare the resulting cut and torch behavior against the process target and manual baseline. Adjust only through documented controller settings, changing one setting at a time.

Verification checks and recurring pitfalls

Verify performance in the same plate-support condition that creates the problem. A flat-sheet test alone does not validate tracking over a corner lifted by slag. Check the sensor reading and Z response together: a changing indication with no corresponding motion points to an axis/control-path issue; motion without a stable, repeatable indication points to sensing, wiring, or setup.

1. Solid plate: expected reading is a stable baseline at a fixed gap.
2. Raised corner: expected reading changes consistently as the surface rises, with smooth Z correction and no collision.
3. Cutout or hole: expected behavior is continuous tracking without a spurious command. If the indication drops or jumps, disable automatic correction for that path until the sensor arrangement or control strategy is corrected.
4. Repeated traverse: expected behavior is the same indication and motion for the same geometry and setup. A drifting baseline requires investigation before production use.

Do not treat a positive report that a capacitive unit can cut around holes as proof that every machine will track every opening. Do not let the sensor mask poor plate support or accumulated slag, and do not tune sensitivity to compensate for an unverified wiring or grounding problem. Retain manual control as the fallback until the complete test path passes.

Frequently asked questions