Reactivepower HMS1800 6,553.5 var

[arittel]

What happened?

The Reactivepower at the HMS-1800 shows the wrong value.
It stay's at 6,553.5 var there for also the Powerfactor gets shown als 0.002

The Value also get's send over the MQQT to ioBroker.

To Reproduce Bug

Just connect to a HMS-1800.

Expected Behavior

Powerfactor should about 1 and the ReactivePower should be around 0.0 - 0.2 var

Install Method

Self-Compiled

What git-hash/version of OpenDTU?

23.5.31

Relevant log/trace output

No response

Anything else?

No response

[tbnobody]

Seems to be the same as https://github.com/tbnobody/OpenDTU/wiki/FAQ#q-with-hms-2000-the-reactive-power-shows-unrealistic-value-65535 and therefor not fixable. What is your inverter firmware version?

[tobiasboecker]

Having the same issue with HMS-1800 and Firmware Version 1.0.14

[tbnobody]

Yes... Thats exactly what the FAQ entry says... Issue in Version 1.0.14... I will extend the FAQ entry to also match the HMS-1800

[tbnobody]

Going to close this issue as there is no chance to fix it as it is related to the firmware on the inverter itself.

[broth-itk]

According to Hoymiles, this bug should have been solved among many other issues.
I upgraded to the latest version 1.0.20 but the values are now changing between

6.553,5 var, 0,0 var and 6.553,4 var

Sometimes a useful value of 0.2 is shown.

Question: Are we having an issue with signed/unsigned integers here?

OpenDTU v23.8.8 is installed.

[stefan123t]

@broth-itk how did you upgrade the firmware from 1.0.14 to 1.0.20 ?
Do you have an Hoymiles DTU too, which model ?

Would you or someone else on the old 1.0.14 firmware still be able to trace the firmware update performed by Hoymiles
or did you receive the firmware images to update them yourself ?
We are very interested in firmware images and detailled traces of the update procedure / protocol 😉

[broth-itk]

@stefan123t The Upgrade was done with DTU-Pro-S after Hoymiles Support has initiated the upgrade process.

I would have been very happy to help and capture the communication, but somehow I remembered that the project is only meant to collect data ;-)

Anyway, with the original DTU we can still capture a trace when a new firmware version is released.
For now we can capture the method to set a grid profile or any other communication.

My OpenDTU has to be set to "listen only" and dump all to serial.

According to Hoymiles, the overflow is present in 1.0.14 for HMS-2000-4T (or maybe others).
1.0.14 on my HMS-350-1T is behaving fine as far as I can tell.

Other issues noticed with 1.0.14 on HMS-2000-4T were: MPPT inconsistencies or a very long waiting time (20 minutes and more) until providing power to the grid. I will need to check my grafana diagrams to confirm that those were solved.

In order not to mix up topics, maybe stick to the signed/unsigned 16 bit integer issue.

[stefan123t]

For the Power Factor I have seen one Power Factor 0.95 with Leading being selected prior to it in the AT TOR-A Grid Profile PDF.
See here https://gist.github.com/noone2k/0b3a116a6f35286abef7199b62a0777a?permalink_comment_id=4770954#gistcomment-4770954

position: 96 	: 005f 	 95 	 0.95	[]		[Power Factor (PF)] -> in CRC 6c58
vs.
position: 96 	: ffa1 	 65441 	 654.41	[]		[Power Factor (PF)] -> in CRC 9c27

@broth-itk this Lagging or negative flag seems to set the 0xff in front of 0xa1 in this case.
Leading being positive does have 0x00 in front of 0x5f.
BTW: I have done some guess work:
0x0000-0x01 = 0xffff or -1 -> Lagging PF -0.01
0xffff-0x5e = 0xffa1 or 95 -> Lagging PF 0.95 (actually -0.95)
0xffff-0x63 = 0xff9c or 100 -> Lagging PF 1.00 (acutally -1.00)

See here for the corrected documentation quote: https://gist.github.com/noone2k/0b3a116a6f35286abef7199b62a0777a?permalink_comment_id=4785215#gistcomment-4785215

We also have another oddity with regards to the RVHF divisor in the grid profile for some versions Hoymiles has changed that obviously.

[github-actions]

This issue has been automatically locked since there has not been any recent activity after it was closed. Please open a new discussion or issue for related concerns.