Skip to content

docs: record field confirmations in the errata - #42

Merged
kristoflemmens merged 2 commits into
mainfrom
docs/field-confirmations
Oct 25, 2026
Merged

kristoflemmens merged 2 commits into
mainfrom
docs/field-confirmations

Conversation

@kristoflemmens

@kristoflemmens kristoflemmens commented Oct 25, 2026 •

Copy link
Copy Markdown
Contributor

What and why

Closes #32

Field confirmations from other Genesis implementations and the reference pump, recorded in docs/errata.md.

  • Input 83–84 "Electric meter kWh total": the row now cites the field report. Users of CJNE/thermiagenesis saw wrong kWh totals while it read MSB first (#292; a Calibra, firmware 16.00 per a later comment), and pythermiagenesis switched to LSB first at scale 10 (chore: support working in git worktrees #3). No correct reading after that switch has been reported, so the row keeps to be verified on a pump with an electric meter (17.1).
  • Holding 118 "Outdoor temperature sensor": it reads 20000 until a BMS writes it (reference pump: 117 = 0, 118 = 20000). A written value takes effect only when holding 117 selects the BMS (200°C Target Temperature infradom/thermia-genesis-calibra-modbus-hints#1). Per its own description (chapters 3 and 4), the pump falls back to its own sensor when 118 isn't refreshed within 12 hours.

Already handled elsewhere, from the same issue:

Not added: the confirmation that the operating hours are MSB first (input 48–51, checked against the myThermia export and the display). That is neither a contradiction nor undocumented behaviour, and the errata only list those. The map and the fixture already pin the order.

Docs only; no code changes.

Checklist

  • The test was written first and failed for the right reason before the implementation (n/a: docs only)
  • Coverage stays at 100%
  • docs/design.md / CONTEXT.md updated if a design choice or domain term changed (errata only)
  • PR title follows Conventional Commits (feat:, fix:, docs:, ...)

🤖 Generated with Claude Code

The LSB-first order of the electric meter kWh total (input 83-84) is
confirmed on a Calibra with an EM340 meter, so the row drops "to be
verified". Holding 118 reads 20000 until a BMS writes it, and a written
value takes effect only when holding 117 selects the BMS.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@kristoflemmens kristoflemmens added the documentation Improvements or additions to documentation label Oct 25, 2026

@kristoflemmens kristoflemmens left a comment

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Reviewed c76eb00 (only docs/errata.md; 61 files, the same as main) against the sources in #32.

Holding 118: correct. infradom/thermia-genesis-calibra-modbus-hints#1 (2025-11-27) says: "If the value has not been written, it shows 200. If you write the value, it will not be active until you enable the BMS temperature flag also." The 12-hour fallback is in the descriptions of holding 118 in chapters 3 and 4.

Leaving out the operating hours confirmation: justified. MSB first for 48–51 is what the document says, so there is no contradiction and no undocumented behaviour. That falls outside the scope of the errata as the intro defines it. The order is already pinned by the map, the decoding test and the fixture. And as the comment on #32 itself says, with MSB halves of 0 it mainly rules out a swap.

One concrete finding:

83–84: "confirmed on real hardware" claims more than the sources show
I read CJNE/thermiagenesis#292 and CJNE/pythermiagenesis#3 / ff20ce2:

  • The symptom (more than 16,000,000 kWh in 24 h) comes from the reporter (petterkatt), and that report gives no firmware.
  • The analysis "LSB first … and a scale of 10" (2025-03-29) by cr0i is derived from the documentation: "All other Values with MSB and LSB are in the order first MSB … But the Electric Meter has LSB first on adress 83". It is not a reading of corrected values.
  • ff20ce2 (2025-06-10) switches to TYPE_LONG_LE and scale 10. Right after, cr0i writes that the electric meter is no longer visible after the update ("after Update to 0.0.12, I can´t find the electric meter anymore"). Nobody reports a correct total after the fix.
  • Firmware 16.00 and the meter come from cr0i on 2025-12-29, and the meter with a reservation: "Carlo Gavazzi EM340, I think".

So what the sources support: the symptom fits LSB first and scale 10 read the wrong way round, and another implementation adopted that order. A confirmed correct reading on a meter does not exist. Suggestion: write it like that, for example "supported by a field report: a Calibra (firmware 16.00) showed an absurd total with MSB first, and CJNE/pythermiagenesis read it LSB first at scale 10 since ff20ce2; no corrected reading has been reported". Keep "to be verified on a pump with an electric meter", possibly as "on 17.1". The Resolution "LSB first, as documented" can stay.

…fied

No correct kWh reading after the switch to LSB first has been reported, so
the order stays to be verified on a pump with an electric meter.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

@kristoflemmens kristoflemmens left a comment

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Re-check of 4f60d59: the finding is resolved. The 83–84 row now describes the evidence as a field report (wrong totals with MSB first, the switch to LSB first at scale 10 in pythermiagenesis, firmware 16.00 from a later comment). It states that no correct reading after the switch has been reported, and keeps "To be verified on a pump with an electric meter (17.1)". The PR description matches. Only errata.md changed; the tree has 61 files. No new findings.

@kristoflemmens
kristoflemmens merged commit f275a4a into main Oct 25, 2026
12 checks passed
@kristoflemmens
kristoflemmens deleted the docs/field-confirmations branch October 25, 2026 13:15
kristoflemmens pushed a commit that referenced this pull request Oct 25, 2026
🤖 I have created a release *beep* *boop*
---


## 0.1.0 (2026-10-25)


### Features

* add 17.1 register map with consistency test
([#5](#5))
([0d33c96](0d33c96))
* add the thermia-genesis dump command
([#29](#29))
([dfb224b](dfb224b))
* add the ThermiaGenesis API with async_update
([b24031c](b24031c))
* detect the optional parts and read only what the pump has
([#27](#27))
([7504e25](7504e25))
* expose the firmware version
([#28](#28))
([abc38cc](abc38cc))
* map the cooling registers of firmware 17.1
([#22](#22))
([0809c97](0809c97))
* map the electric meter registers of firmware 17.1
([#24](#24))
([bf929ae](bf929ae))
* map the heat pump unit registers of firmware 17.1
([#18](#18))
([ffd9923](ffd9923))
* map the heat registers of firmware 17.1
([#20](#20))
([6c24ba5](6c24ba5))
* map the pool registers of firmware 17.1
([#23](#23))
([f3a1503](f3a1503))
* map the status registers and complete the 17.1 map
([#25](#25))
([9abd506](9abd506))
* map the tap water registers of firmware 17.1
([#21](#21))
([badbbf5](badbbf5))


### Bug Fixes

* read an absent room sensor as no value
([#40](#40))
([2d0dbee](2d0dbee))
* read the cooling mix valve opening at scale 100
([#41](#41))
([f3afa14](f3afa14))


### Documentation

* cite the 17.0 and Mega 17.1 documents in two errata rows
([#38](#38))
([065d06e](065d06e))
* detect optional subsystems from their enable coils only
([#17](#17))
([d306b8d](d306b8d))
* record field confirmations in the errata
([#42](#42))
([f275a4a](f275a4a))
* restore the negative number examples of chapter 2
([#39](#39))
([8d6b655](8d6b655))

---
This PR was generated with [Release
Please](https://github.com/googleapis/release-please). See
[documentation](https://github.com/googleapis/release-please#release-please).

Co-authored-by: klconsultancy-release-bot[bot] <345732020+klconsultancy-release-bot[bot]@users.noreply.github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

documentation Improvements or additions to documentation

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Record field confirmations in the errata (meter 83–84, holding 118, input 121)

1 participant