Steam-probe Health Before Descaling – Breville & Sage

Breville Dual Boiler / Oracle · Model-specific diagnostics

Steam-probe health before descaling.

Understand what the level probes control and why a normal-looking reading is not a guarantee that a descale will succeed.

7 workshop stagesOriginal BrevTech footageDiagnostic reference

The source compares BES920, BES980 and BES990 examples with different detection thresholds. Do not transfer one model’s AD number, menu or test conditions to another.

The transparent probe demonstration with pump, heat and reference labels.
The transparent probe demonstration with pump, heat and reference labels. 4:10 in the video

Before you start

A diagnostic reference based on the Part 1 workshop investigation. It explains observations and model differences; powered draining, external probe rigs and internal meter work are technician procedures.

Before you work

  • Do not open hot boiler connections or drain a powered machine to reproduce the demonstrations.
  • Do not disconnect sensing or protection to force a diagnostic result.
  • Use qualified service support for abnormal filling, heating or probe behaviour.

Tools and supplies

  • A record of the exact model, symptoms and verified firmware information
  • The correct model’s diagnostic documentation
  • A qualified repairer for any internal, hot-drain or electrical tests

Part 1 is an investigation, not a certified acceptance test. Part 2 explicitly revisits its confidence in new probes and cold measurements; both belong in the same decision.

Part 1 of 3

Assess and prepare

01

Start with the machine’s condition

Record heating, steam, filling and leak symptoms before deciding to descale. The film shows scale at an inlet and element as reasons that maintenance may still be needed.

Resolve a known fault first. Neither “never descale” nor “a descale fixes every fault” follows from the examples.

View this section in the original video0:11

Reference: Video 0:11 · Guide checks

Scale on the opened boiler/element used to illustrate why maintenance matters.
Scale on the opened boiler/element used to illustrate why maintenance matters. 3:10 in the video
Part 2 of 3

Workshop walkthrough

02

Understand high, low and reference sensing

The steam-boiler example has an upper level probe associated with filling, a lower level probe associated with heat permission and a reference/ground path.

The transparent demonstration makes the wet/dry sequence visible. It is an explanatory rig, not a wiring or powered-test setup to reproduce at home.

View this section in the original video3:33

Reference: Video 3:33 · Guide checks

The transparent probe demonstration with pump, heat and reference labels.
The transparent probe demonstration with pump, heat and reference labels. 4:10 in the video
03

Distinguish the three failure patterns

The film separates failure to refill, partial filling with heat and heating an effectively empty boiler. These can have different consequences even when all are described as “no steam” afterwards.

Treat an abnormal fill or heating response as a stop condition. A symptom pattern still needs proper diagnosis before parts are selected.

View this section in the original video5:33

Reference: Video 5:33 · Guide checks

The source’s partial-fill failure illustration.
The source’s partial-fill failure illustration. 6:10 in the video
04

Keep thresholds tied to their source machine

The source compares detection points and AD displays on several models. Those numbers describe the examples and stated firmware, not one universal safe range.

Record the exact machine, firmware if verified, temperature and test state when a technician interprets its readings.

View this section in the original video7:16

Reference: Video 7:16 · Guide checks

The video’s model/firmware detection-range comparison; these are archive-specific observations.
The video’s model/firmware detection-range comparison; these are archive-specific observations. 9:10 in the video
05

Separate a cold bench reading from hot behaviour

The source compares used and new probes under several conditions. Part 2 later shows why a high cold resistance and even newly fitted probes do not guarantee correct behaviour during a descale.

Do not turn the Part 1 results into a promise that soaking, cleaning or replacement makes the cycle safe.

View this section in the original video9:31

Reference: Video 9:31 · Guide checks

Probe designs compared near the end of the Part 1 experiments.
Probe designs compared near the end of the Part 1 experiments. 13:10 in the video
06

Use the correct model’s diagnostic route

The closing section demonstrates different approaches for Oracle Touch, non-touch Oracle and Dual Boiler machines. They are not one shared menu sequence.

The external BES920 self-check guide can help record displayed information. Controlled hot-drain tests and disconnected-probe measurements require a qualified repairer and an appropriate procedure.

View this section in the original video14:47

Reference: Video 14:47 · Guide checks

The Oracle Touch diagnostic display in its model-specific example.
The Oracle Touch diagnostic display in its model-specific example. 16:10 in the video
The separate non-touch Oracle menu example.
The separate non-touch Oracle menu example. 16:40 in the video
Part 3 of 3

Finish and verify

07

Review the later findings before deciding

Read Part 2 alongside this video. Its later observations reduce the confidence that Part 1 places in probe replacement and cold measurements.

If readings are abnormal or the machine cannot reliably fill, stop and arrange service rather than using a numeric margin as a descale pass certificate.

View this section in the original video17:50

Reference: Video 17:50 · Guide checks

The original workshop film

Watch the repair in context

The source contains external rigs, internal electrical work and powered boiler draining. This reference guide explains the findings without supplying those hazardous test recipes.

Final checks

Use this checklist to review the work and the evidence. It does not claim that these checks were completed on your machine.

  • The high/low sensing functions are identified without assuming one colour map for all machines.
  • Readings are tied to the actual model and test conditions.
  • Part 2’s later findings are considered.
  • No bench number, new probe or archive threshold is treated as a guarantee.

Sources and model notes

Written from the original BrevTech footage and its current description. Every screenshot links to the matching video time. The film may omit service detail; model and procedure limits are stated above.

Use the procedure and parts that match the actual machine. Stop if the fault, connections or required safety checks cannot be established. Read the BrevTech repair disclaimer.