[Bug]: V14_1HAR_q1 template: wrong analog/digital split — Nano PK Plus (V14.1HAR.q) sends 113 analog + 8 digital words, template defines 112 + 9

[josefstr]

Pre-flight checklist

• I have searched existing issues and confirmed this is not a duplicate.
• I am using a supported and reasonably current version.
• This report contains no secrets, credentials or personal data.

Description

Environment

• Boiler: Hargassner Nano PK Plus, firmware V14.1HAR.q
• Integration: v0.4.0 (also present in v0.3.1)
• Home Assistant 2026.8.3, template V14_1HAR_q1

Symptom

With the stock V14_1HAR_q1 template, all digital words are read one
position too early. Every digital state (heating circuit pumps, mixers,
fault flags, ...) is wrong or nonsensical, and the debug log is flooded
with:

at index 112: invalid literal for int(): '0.00'

Note the message length check does NOT catch this: the telegram has 121
values and the template also expects 121 — only the split is wrong.

Analysis

Our boiler sends 113 analog values followed by 8 digital words
(121 total). The V14_1HAR_q1 template defines 112 analog + 9 digital
— same total, wrong split. As a result the first "digital" word is
actually the last analog value ('0.00' → parse error above), and all
real digital words shift by one position.

We verified this against a raw telnet capture (nc <boiler> 23): the
decimal-place pattern of every channel only lines up with the 113/8
split. The additional analog value sits at position 90 (between
TBs_A and TBs_1); all channels from there shift by +1, and the 9th
digital word (Reserved_8 in the template) does not exist on this
machine.

Fix we are running locally

In the V14_1HAR_q1 template:

1. insert a placeholder analog channel at id 90 (ids of all later
   channels +1),
2. remove digital word id 8 (Reserved_8).

Running stable since 2026-08 on 0.3.1 and now on 0.4.0 (clean logs,
plausible digital states).

Happy to open a PR with the adjusted template. Possibly this warrants a template variant
(similar to the nano2 variant added in v0.4.0 for #17) rather than
changing V14_1HAR_q1 for everyone — we don't know which hardware
combination produces the 112+9 layout, if any.

Steps to reproduce

1. Set up the integration (v0.4.0) against a Hargassner Nano PK Plus running firmware V14.1HAR.q, using the V14_1HAR_q1
   template.
2. Enable debug logging for custom_components.bauergroup_hargassnerintegration.
3. Observe the log message at index 112: invalid literal for int(): '0.00' on every telegram, and compare the digital
   sensors (heating circuit pumps, mixers, fault flags) with the boiler display — they do not match.

Expected behavior

The template splits the 121 telegram values as 113 analog + 8 digital words, so digital states match the boiler and the
telegram parses without errors.

Severity

Medium — degraded, but usable

Affected version / commit

v0.4.0 (also reproducible on v0.3.1)

Environment

• Home Assistant 2026.8.3 (HA OS)
• Integration installed via HACS
• Boiler: Hargassner Nano PK Plus, firmware V14.1HAR.q, connected via Telnet (port 23)

Relevant logs or screenshots

Debug log (repeated for every telegram):

at index 112: invalid literal for int(): '0.00'


Sample telegram line (full 20 s capture attached, 15 lines, each with 121 values — note value #113 (index 112) is `0.00`, i.e. still an analog value where the template expects the first digital word):

  pm 1 1.1 7.5 28.8 0 29.2 30 10 33.8 0 0 62.7 61.2 42.2 92 5 0 0 0 63 0 0 30.0 100 30 30 76 76.3 92 2 0 0 10 3 0 5 0 7 0 0
  6363 0 15063 0.00 0.00 -3 50.2 24107 28.9 117.9 36 -20.0 -20.0 0.0 20.6 20.9 0.0 1 0 -20.0 0 20.0 20.0 0 1 0 28.1 0 20.0
  22.0 0 1 0 28.1 0 20.0 20.5 0 1 0 -20.0 0 20.0 20.0 0 1 0 -20.0 0 62.7 0 0 0 -20.0 0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
  0.00 2 1000 0 0 1 0 0 0


Complete Log:

pm 1 1.1 7.5 28.8 0 29.2 30 10 33.8 0 0 62.7 61.2 42.2 92 5 0 0 0 63 0 0 30.0 100 30 30 76 76.3 92 2 0 0 10 3 0 5 0 7 0 0 6363 0 15063 0.00 0.00 -3 50.2 24107 28.9 117.9 36 -20.0 -20.0 0.0 20.6 20.9 0.0 1 0 -20.0 0 20.0 20.0 0 1 0 28.1 0 20.0 22.0 0 1 0 28.1 0 20.0 20.5 0 1 0 -20.0 0 20.0 20.0 0 1 0 -20.0 0 62.7 0 0 0 -20.0 0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00 2 1000 0 0 1 0 0 0 
pm 1 1.1 7.5 28.7 0 29.2 30 10 33.7 0 0 62.7 61.2 42.2 92 5 0 0 0 63 0 0 30.0 100 30 30 76 76.3 92 2 0 0 11 3 0 0 0 7 0 0 6363 0 15063 0.00 0.00 -3 50.2 24108 28.8 119.8 36 -20.0 -20.0 0.0 20.6 20.9 0.0 1 0 -20.0 0 20.0 20.0 0 1 0 28.1 0 20.0 22.0 0 1 0 28.1 0 20.0 20.5 0 1 0 -20.0 0 20.0 20.0 0 1 0 -20.0 0 62.8 0 0 0 -20.0 0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00 2 0 0 0 1 0 0 0 
pm 1 1.1 7.5 28.7 0 29.2 30 10 33.8 0 0 62.7 61.2 42.2 92 5 0 0 0 63 0 0 30.0 100 30 30 76 76.3 92 2 0 0 10 2 0 0 0 7 0 0 6363 0 15063 0.00 0.00 -3 50.2 24107 28.9 119.8 36 -20.0 -20.0 0.0 20.6 20.9 0.0 1 0 -20.0 0 20.0 20.0 0 1 0 28.1 0 20.0 22.0 0 1 0 28.1 0 20.0 20.5 0 1 0 -20.0 0 20.0 20.0 0 1 0 -20.0 0 62.8 0 0 0 -20.0 0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00 2 0 0 0 1 0 0 0 
pm 1 1.1 7.5 28.7 0 29.2 30 10 34.0 0 0 62.7 61.2 42.2 92 5 0 0 0 63 0 0 30.0 100 30 30 76 76.3 92 2 0 0 11 3 0 5 0 7 0 0 6363 0 15063 0.00 0.00 -3 50.2 24107 28.9 117.9 36 -20.0 -20.0 0.0 20.6 20.9 0.0 1 0 -20.0 0 20.0 20.0 0 1 0 28.1 0 20.0 22.0 0 1 0 28.1 0 20.0 20.5 0 1 0 -20.0 0 20.0 20.0 0 1 0 -20.0 0 62.8 0 0 0 -20.0 0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00 2 1000 0 0 1 0 0 0 
pm 1 1.1 7.5 28.7 0 29.2 30 10 33.9 0 0 62.7 61.2 42.2 92 5 0 0 0 63 0 0 30.0 100 30 30 76 76.3 92 2 0 0 10 2 0 3 0 7 0 0 6363 0 15063 0.00 0.00 -3 50.2 24107 28.9 117.9 36 -20.0 -20.0 0.0 20.6 20.9 0.0 1 0 -20.0 0 20.0 20.0 0 1 0 28.1 0 20.0 22.0 0 1 0 28.1 0 20.0 20.5 0 1 0 -20.0 0 20.0 20.0 0 1 0 -20.0 0 62.8 0 0 0 -20.0 0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00 2 0 0 0 1 0 0 0 
pm 1 1.1 7.5 28.7 0 29.2 30 10 34.0 0 0 62.7 61.2 42.2 92 5 0 0 0 63 0 0 30.0 100 30 30 76 76.3 92 2 0 0 11 2 0 0 0 7 0 0 6363 0 15063 0.00 0.00 -3 50.2 24107 28.9 117.9 36 -20.0 -20.0 0.0 20.6 20.9 0.0 1 0 -20.0 0 20.0 20.0 0 1 0 28.2 0 20.0 22.0 0 1 0 28.1 0 20.0 20.5 0 1 0 -20.0 0 20.0 20.0 0 1 0 -20.0 0 62.8 0 0 0 -20.0 0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00 2 0 0 0 1 0 0 0 
pm 1 1.1 7.5 28.7 0 29.2 30 10 34.1 0 0 62.7 61.2 42.2 92 5 0 0 0 63 0 0 30.0 100 30 30 76 76.3 92 2 0 0 11 3 0 0 0 7 0 0 6363 0 15063 0.00 0.00 -3 50.2 24108 28.9 116.0 36 -20.0 -20.0 0.0 20.6 20.9 0.0 1 0 -20.0 0 20.0 20.0 0 1 0 28.2 0 20.0 22.0 0 1 0 28.1 0 20.0 20.5 0 1 0 -20.0 0 20.0 20.0 0 1 0 -20.0 0 62.7 0 0 0 -20.0 0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00 2 1000 0 0 1 0 0 0 
pm 1 1.1 7.5 28.8 0 29.2 30 10 34.2 0 0 62.7 61.2 42.2 92 5 0 0 0 63 0 0 30.0 100 30 30 76 76.3 92 2 0 0 10 2 0 4 0 7 0 0 6363 0 15063 0.00 0.00 -3 50.2 24107 28.9 119.8 36 -20.0 -20.0 0.0 20.6 20.9 0.0 1 0 -20.0 0 20.0 20.0 0 1 0 28.2 0 20.0 22.0 0 1 0 28.1 0 20.0 20.5 0 1 0 -20.0 0 20.0 20.0 0 1 0 -20.0 0 62.7 0 0 0 -20.0 0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00 2 1000 0 0 1 0 0 0 
pm 1 1.1 7.5 28.8 0 29.2 30 10 34.1 0 0 62.7 61.2 42.2 92 5 0 0 0 63 0 0 30.0 100 30 30 76 76.3 92 2 0 0 10 4 0 4 0 7 0 0 6363 0 15063 0.00 0.00 -3 50.2 24107 28.9 117.9 36 -20.0 -20.0 0.0 20.6 20.9 0.0 1 0 -20.0 0 20.0 20.0 0 1 0 28.2 0 20.0 22.0 0 1 0 28.1 0 20.0 20.5 0 1 0 -20.0 0 20.0 20.0 0 1 0 -20.0 0 62.7 0 0 0 -20.0 0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00 2 1000 0 0 1 0 0 0 
pm 1 1.1 7.5 28.8 0 29.2 30 10 34.2 0 0 62.7 61.2 42.2 92 5 0 0 0 63 0 0 30.0 100 30 30 76 76.3 92 2 0 0 11 1 0 0 0 7 0 0 6363 0 15063 0.00 0.00 -3 50.2 24108 28.9 119.8 36 -20.0 -20.0 0.0 20.6 20.9 0.0 1 0 -20.0 0 20.0 20.0 0 1 0 28.2 0 20.0 22.0 0 1 0 28.1 0 20.0 20.5 0 1 0 -20.0 0 20.0 20.0 0 1 0 -20.0 0 62.7 0 0 0 -20.0 0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00 2 0 0 0 1 0 0 0 
pm 1 1.1 7.5 28.8 0 29.2 30 10 34.0 0 0 62.7 61.2 42.1 92 5 0 0 0 63 0 0 30.0 100 30 30 76 76.3 92 2 0 0 11 2 0 0 0 7 0 0 6363 0 15063 0.00 0.00 -3 50.2 24108 28.9 119.8 36 -20.0 -20.0 0.0 20.6 20.9 0.0 1 0 -20.0 0 20.0 20.0 0 1 0 28.2 0 20.0 22.0 0 1 0 28.1 0 20.0 20.5 0 1 0 -20.0 0 20.0 20.0 0 1 0 -20.0 0 62.8 0 0 0 -20.0 0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00 2 0 0 0 1 0 0 0 
pm 1 1.1 7.5 28.8 0 29.2 30 10 34.1 0 0 62.7 61.2 42.1 92 5 0 0 0 63 0 0 30.0 100 30 30 76 76.3 92 2 0 0 10 3 0 4 0 7 0 0 6363 0 15063 0.00 0.00 -3 50.2 24108 28.8 116.0 36 -20.0 -20.0 0.0 20.6 20.9 0.0 1 0 -20.0 0 20.0 20.0 0 1 0 28.2 0 20.0 22.0 0 1 0 28.1 0 20.0 20.5 0 1 0 -20.0 0 20.0 20.0 0 1 0 -20.0 0 62.8 0 0 0 -20.0 0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00 2 1000 0 0 1 0 0 0 
pm 1 1.1 7.5 28.7 0 29.2 30 10 34.2 0 0 62.7 61.2 42.1 92 5 0 0 0 63 0 0 30.0 100 30 30 76 76.3 92 2 0 0 11 1 0 4 0 7 0 0 6363 0 15063 0.00 0.00 -3 50.2 24107 28.9 114.2 36 -20.0 -20.0 0.0 20.6 20.9 0.0 1 0 -20.0 0 20.0 20.0 0 1 0 28.2 0 20.0 22.0 0 1 0 28.1 0 20.0 20.5 0 1 0 -20.0 0 20.0 20.0 0 1 0 -20.0 0 62.8 0 0 0 -20.0 0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00 2 1000 0 0 1 0 0 0 
pm 1 1.1 7.5 28.8 0 29.2 30 10 33.8 0 0 62.7 61.2 42.1 92 5 0 0 0 63 0 0 30.0 100 30 30 76 76.3 92 2 0 0 11 4 0 1 0 7 0 0 6363 0 15063 0.00 0.00 -3 50.2 24107 28.9 116.0 36 -20.0 -20.0 0.0 20.6 20.9 0.0 1 0 -20.0 0 20.0 20.0 0 1 0 28.2 0 20.0 22.0 0 1 0 28.1 0 20.0 20.5 0 1 0 -20.0 0 20.0 20.0 0 1 0 -20.0 0 62.8 0 0 0 -20.0 0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00 2 0 0 0 1 0 0 0 
pm 1 1.1 7.5 28.8 0 29.2 30 10 34.2 0 0 62.7 61.2 42.1 92 5 0 0 0 63 0 0 30.0 100 30 30 76 76.3 92 2 0 0 10 2 0 0 0 7 0 0 6363 0 15063 0.00 0.00 -3 50.2 24107 28.9 116.0 36 -20.0 -20.0 0.0 20.6 20.9 0.0 1 0 -20.0 0 20.0 20.0 0 1 0 28.2 0 20.0 22.0 0 1 0 28.1 0 20.0 20.5 0 1 0 -20.0 0 20.0 20.0 0 1 0 -20.0 0 62.8 0 0 0 -20.0 0 0.0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00 2 0 0 0 1 0 0 0 

[github-actions]

AI Analysis

Type: issue
Title: [Bug]: V14_1HAR_q1 template: wrong analog/digital split — Nano PK Plus (V14.1HAR.q) sends 113 analog + 8 digital words, template defines 112 + 9
Author: @josefstr

This issue requires review. Please check the full description for details.

---

This summary was automatically generated by AI to help with triage and may not be 100% accurate.

[karlspace]

Thanks for the capture — fifteen lines with the operating states noted is what turned this from a guess into something provable, and it ended up uncovering more than the issue you reported.

Your two structural conclusions were both right

113 analog + 8 digital words, and Reserved_8 does not exist. Both confirmed.

What I did not expect is why. This repository ships a recording from the manufacturer at docs/private_firmware_samples/DAQ00001.DAQ, and its own header says:

KT=Nano.2 32
SW=V14.1HAR.q1

So it is Hargassner's channel definition for exactly the firmware V14_1HAR_q1 claims to describe — and it declares 112 analog + 8 digital = 120 values. The template we shipped declared 121 and disagreed on channel order in five further places:

Defect	shipped until v0.4.0	manufacturer
TRA_A, TRA_1, TRA_2, TRA_B	before TVL_x	third in the block, after TVLs_x
TB1 / TBs_1	swapped	TB1 before TBs_1
Reserved_8	ninth digital word	does not exist — 8 words
Reserved_5	declared	does not exist

The TRA_x misplacement meant every heating circuit's room temperature was sitting in its flow-temperature entity and vice versa, for everyone on that template. Your issue is what surfaced it.

On the extra value at position 90

Your instinct pointed at the right region, and the explanation is a little different from an insertion. In the factory order the block reads TBA, TBs_A, TB1, TBs_1 — the shipped template had TB1 and TBs_1 swapped, so "between TBs_A and TBs_1" in the old numbering is precisely where TB1 belongs. Measured against the manufacturer's list, your boiler actually:

• omits BLDC_ES ist and BLDC_ES soll (−2)
• adds Programm HKM1 in the prefix (+1)
• adds BoiZustand_1 and one placeholder after TBs_1 (+2)

which nets to the 113 you counted. One inserted placeholder at 90 would have fixed the length and the digital split, but left the TRA_x and TB1 mappings wrong — so it is just as well this went the long way round.

How the layout was pinned

Not by plausibility. A boiler's channel list is a subsequence of the manufacturer's full list — a machine without circuits 3 and 4 omits those channels and keeps the order — so the layout is the strictly increasing map from your 113 positions onto the reference list that minimises dop violations (dop='0' never prints a decimal, dop='2' always prints two). That minimum is 0 violations across all 15 lines, against 300 for the stock template.

Zero-violation alignments are not unique, so the rest was settled structurally: circuits come in complete blocks of seven, and both of your uninstalled circuits read -20.0 0 20.0 20.0 0 1 0 — the manufacturer's own signature for a circuit that is not fitted. Boilers A and B read the absent-DHW signature -20.0 0, boiler 1 reads 62.7 °C, and Wasserdruck is the last analog channel and the only two-decimal value at the tail.

What shipped

• V14_1HAR_q1 corrected to the manufacturer's definition — 634a6d6. This is a breaking change; the superseded layout is kept as V14_1HAR_q1_legacy for anyone who needs the old mapping back.
• V14_1HAR_q1_nanopkplus — your boiler's layout, 113 + 8 — 7395333, with four regression tests built from seven of your lines (the subset that still reproduces every distinct token at every position that varies).

Switch to V14_1HAR_q1_nanopkplus under Settings → Devices & Services → Hargassner → Configure. The connection sensor's expected_length and last_message_length attributes should both read 121, and at index 112: invalid literal for int(): '0.00' should be gone.

One incidental confirmation from your data: position 47 reads 24107 — that is U Netzteil, the 24 V supply rail in mV. Read two positions early, as the V14_0HAR_q template does, that rail lands in BRT and appears as a burner temperature of about 24000 °C, which is the symptom in #22. Your capture helped settle that one too.

On your PR offer — thank you, and it is already in, derived from your data rather than around it. The commit credits you. I have reopened this until you have confirmed it reads correctly on the actual boiler; if anything is still off, the capture route is the same as before.

[karlspace]

Correction to my previous comment, and a heads-up before you update.

Install 0.5.1, not 0.5.0. The part that concerns you is unchanged: V14_1HAR_q1_nanopkplus exists under that exact name, with the same 113 + 8 layout derived from your capture, and the same four regression tests. Nothing about your boiler's template changed.

What I got wrong was the other half of that comment. I also rewrote V14_1HAR_q1 itself to match the manufacturer's DAQ file, and that was a mistake — not because the DAQ file is fake, but because I drew the wrong conclusion from it. An HG-PK32 in the field reads plausible heating-circuit values with the original order (flow 32.1 °C, unfitted room sensor 0.0 °C) and implausible ones with the manufacturer's order (flow 0.0 °C against a setpoint of 32.1 °C). So two channel orders genuinely exist under the same V14.1HAR.q1 firmware string, and DAQ00001.DAQ documents one machine rather than the firmware as such.

0.5.1 restores V14_1HAR_q1 byte-identical to 0.4.0 and moves the manufacturer's order to its own key, V14_1HAR_q1_nano2_32. Redefining a key that installations already use moves their values between entities, and the new layout being defensible would not have made that acceptable.

So one of my claims to you needs withdrawing: I wrote that the stock template "was wrong in six places, not one". That is not established. The differences are real, but they are differences between two layouts that both occur in the field — not defects. Your own two conclusions are unaffected and still confirmed: 113 analog + 8 digital words, and no Reserved_8 on your machine.

Still reopened until you confirm V14_1HAR_q1_nanopkplus reads correctly on the actual boiler.

[josefstr]

Confirmed fixed with v0.5.1 and the new V14_1HAR_q1_nanopkplus template.

Switched our Nano PK Plus (V14.1HAR.q) to the new template today: no more invalid literal ... at index 112 messages, message length matches (113 analog + 8 digital), and the values are plausible (HK1 flow temperature with the HK1 pump running, buffer temperatures, pellet counter, outdoor temperature matches our ventilation unit within 0.5 K).

For the record: our local workaround only had the 113/8 split right. Your template also corrects the channel order between index 37 and 92 (no BLDC channels on this unit, but Programm HKM1, BoiZustand_1 and a dummy), which our patch had wrong by up to two positions. Thanks for the quick fix, closing this