HW5103 is missing values

[TheEragon]

After recent update 08.hmu.HW5103.csv doesn't contain for exmaple RunDataElectricPowerConsumption,,,,b509,5402005b0d,ign,,IGN:4,,,,value,,EXP,,W,current electrical power consumption (Watt) which is used to.

Is there a reason for this?

[Szpero]

Same here - no data for:
ebusd/hmu/RunDataElectricPowerConsumption

Some logs:

2026-04-26 15:39:41.943 [main notice] ebusd 26.1.26.1 started with broadcast scan on device: ebus-9705ac.local:9999, TCP, enhanced
2026-04-26 15:39:42.322 [device notice] extra info: firmware 1.1[6416].1[6416], jumpers 0x0b
2026-04-26 15:38:10.260 [mqtt error] read message hmu RunDataElectricPowerConsumption not found

2026-04-26 15:40:03.434 [main notice] read scan config file vaillant/08.hmu.HW5103.csv for ID "hmu00", SW0522, HW5103
2026-04-26 15:40:07.494 [main notice] read scan config file vaillant/15.ctlv2.csv for ID "ctlv2", SW0514, HW1104
2026-04-26 15:40:10.564 [main notice] read scan config file vaillant/76.vwzio.csv for ID "vwzio", SW0901, HW5103

[michaelzeipert-cmyk]

Since some days HMU RunDataReturnTemp is missing. It was updated last time on 22-April-2026 in my ioBroker database.

[john30]

this is due to the changes in #481
@burmistrzak why was the block with https://github.com/burmistrzak/ebusd-configuration/blob/ac8ba05b1c91c0164c40324a128eee7a1d6a3f1c/src/vaillant/08.hmu.HW5103.tsp#L55 commented out?
and could it be that more of the hmu are also valid for the HW5103 variant, e.g. RunDataFlowTemp, RunDataReturnTemp @burmistrzak ?

[TheEragon]

@john30 all seems to be valid for HW5103

[burmistrzak]

@john30 IIRC, we disabled some definitions in my fork to establish a baseline of registers that would be supported by all HW5103 revisions and intended to go from there (but admittedly got a bit distracted by all that 720 stuff). I believe a few heat pump units returned 00 for some RunData* registers and we tried working around it because that N/A value always produced a decoding error.

We also now know that some register blocks absolutely do not like to be polled and thus will have to be redefined as passive only.

Lastly, we learned quite a bit about safe registers by observing the NEXT3 internet gateway and the registers block it's regularly polling. Emulating its behavior seems like a good idea, IMHO.

[michaelzeipert-cmyk]

@john30 , thank you for confirming that there may be an issue with a changed file. Is there a way for me to fall back on a version that used to work for me?

[john30]

you can checkout an older commit of https://github.com/eBUS/ebus.github.io and use that one @michaelzeipert-cmyk

[michaelzeipert-cmyk]

you can checkout an older commit of https://github.com/eBUS/ebus.github.io and use that one @michaelzeipert-cmyk

thankyou john30, worked for me (HMU HW5103, SW0905). Now i can read RunDataReturnTemp

[water-escape]

you can checkout an older commit of https://github.com/eBUS/ebus.github.io and use that one @michaelzeipert-cmyk

thankyou john30, worked for me (HMU HW5103, SW0905). Now i can read RunDataReturnTemp

Sorry, what did you do to get it working again? I'm also missing retour and current energy..

[michaelzeipert-cmyk]

you can checkout an older commit of https://github.com/eBUS/ebus.github.io and use that one @michaelzeipert-cmyk

thankyou john30, worked for me (HMU HW5103, SW0905). Now i can read RunDataReturnTemp

Sorry, what did you do to get it working again? I'm also missing retour and current energy..

By default ebusd is looking for the latest configuration files when it is restarted. In the latest files the row for return temperature is missing. I switched the behavior of ebusd to use local configuration files on my host. To achieve that, I changed my compose.yml. After that I copied the latest csv-files to my host and added the missing rows. After a container restart the missing values are showing up.

[water-escape]

Before i go and try that, isnt this going to be solved in a fix?

[Audi678]

@john30 @burmistrzak
The changes from #481 removed many working datapoints on my HW5103 system.
Before the update, these registers worked reliably without decode errors.
After the configuration change, about 27 entities became unavailable.
Could the disabled HMU registers please be reconsidered for HW5103 systems where they actually work correctly?
Maybe:
optional handling,
firmware-specific detection,
or an experimental enable flag
would be better than globally disabling these registers for all HW5103 users.
I can provide logs/testing from a working installation if useful.

[renek74]

same probleme , missing important values. I have openhab for home automation and many rules regarding the values from my heating pump. Need also the old csv file.... Thanks for understanding