Field diagnosis

Pressure spikes and burst hoses: where the shock actually comes from

Hoses failing in the same place, filters splitting or a gauge needle that jumps are transient problems. Steady pressure tells you nothing about them - here is what to measure and what to fix.

Check these in order

CheckWhat it looks likeWhat it tells you
Relief valve responseA relief that opens slowly lets the first, steepest part of a pressure rise through. The steady setting still looks correct on a gauge.Check the relief for a stuck or worn pilot stage; a slow relief is a shock generator.
Accumulator prechargeA bladder or piston accumulator with lost precharge cannot absorb a pulse. Gas pressure is usually far below the value on the label.Measure precharge with the system depressurised and compare with the machine specification.
Air in the oilAir compresses and then releases its energy abruptly, turning a normal pressure rise into a spike. Foam in the tank is the giveaway.Check suction clamps, strainer seal, oil level and breather; bleed the circuit after any repair.
Start-up and shutdown sequenceThe worst spike of the day is often at start-up, or when a loaded function is dumped back to tank.Instrument the pressure line with a peak-reading gauge or data logger, then change the sequence if the peak exceeds the rating.
Hose routing and clampingA hose that can move whips when a pulse arrives, and the failure point is usually at a bend or a clamp edge rather than in the middle.Re-route with the minimum bend radius respected, replace clamps, and confirm the hose pressure rating covers the peak, not the average.
Spool switching timeClosing a large spool quickly against a moving load converts kinetic energy into a pressure peak, and a worn or slow return spring makes it worse.Compare opening and closing times in both directions, then service the spool and pilot circuit.

Units and parts this usually leads to

Field notes on this complaint

The notes our own team works to on this symptom. Each one is a principle you can check on the machine, a step from the case-drain and cleanliness procedure we follow, or a figure transcribed from our catalogue - not advice copied from a video or a forum. Anything we could not check against a catalogue, a manufacturer document or a working machine is left out.

  • Repeated bursts are a relief question. Hoses fail at the peak the circuit allows. Check that the relief actually opens at its setting before replacing hoses again. (circuit principle, checkable on the machine)
  • Cold starts and accumulators. Cold oil and a failed accumulator both raise the peaks a hose sees; both are quick to check and cheap to put right. (circuit principle, checkable on the machine)
  • Know which relief you are measuring: override band differs by type A direct-acting relief can crack several hundred psi below the pressure it holds at full flow, while a pilot-operated (balance piston) relief has a very narrow override band, so the two behave differently on a gauge. When you compare a measured setting against the specification, first establish which type is fitted - otherwise you are reading the override band, not the setting, and may adjust a healthy valve. (circuit principle, checkable on the machine)
  • Back pressure in a shared return can keep a relief from opening at its set pressure A conventional spring-loaded relief has to overcome the pressure in its outlet as well as its spring: elevated back pressure from a common return header - or a line one size too small - adds a force that holds the poppet closed, so the valve lifts late and the circuit sees more pressure than the setting says. Where the manufacturer fitted a remote sensing line, the pilot is taken from a steadier point and the effect is removed; that is what the sensing line is for. (circuit principle, checkable on the machine)
  • Two-stage protection: the safety relief sits 20-30% above the main relief Where a circuit has both a main relief and a safety relief, the safety unit is normally set 20-30% above the main one, so it only acts when the main relief is blocked or too slow - it is protection of last resort, not a working setting. On pilot-operated units a field test connection lets you check the set pressure without taking the valve out of service, which is the right way to confirm it before a pressure spike finds a weak hose or seal first. (our own case-drain and cleanliness procedure)

Common questions

Why do my hoses always fail in the same place?

A repeated failure at one point is almost never a hose quality problem. It is a routing problem - a bend radius below the minimum, a clamp edge cutting into the cover, or a pulse that always arrives at the same point of the circuit. Change the routing and the failures move or stop.

Can a pressure spike damage a piston pump?

Yes. The shock loads the drive shaft, bearings and valve plate together, and repeated shocks appear as fatigue rather than as one obvious impact. This is one reason we ask about hose history when a pump fails early.

How do I measure a spike with normal workshop tools?

A glycerine-filled gauge with a peak-hold needle, or a pressure transducer with a data logger, tee'd into the pressure line at the pump outlet. A standard damped gauge cannot show a spike that lasts a few milliseconds.

Why it happens

The mechanism behind this symptom, from our technical articles: reading the principle first usually shortens the checklist.

More in Valve and pilot control: Spools stick or lock when cold · Controls sluggish or dead · Proportional pilot control reacts late, in steps, or not at all · all areas

Measurements above are guidance for organising a diagnosis, not a substitute for the machine specification. Confirm pressures, flows and control settings against the unit nameplate and your service data before ordering parts.