AC1-5.0-E-G2 Illegal Data Address exception response #715

markheppell · 2024-12-27T17:03:16Z

markheppell
Dec 27, 2024

I have just had my inverter (AC1-5.0-E-G2) firmware updated and I am still getting an Illegal Data Address exception response

If I do a Home Assistant Developer action of Read registers with a count of 45 I get the following

values:
"31000": 10
"31001": 0
"31002": 0
"31003": 10
"31004": 0
"31005": 0
"31006": 2497
"31007": 26
"31008": 661
"31009": 5002
"31010": 2508
"31011": 10
"31012": 661
"31013": 5003
"31014": 65535
"31015": 4
"31016": 670
"31017": 0
"31018": 283
"31019": 380
"31020": 1956
"31021": 35
"31022": 686
"31023": 263
"31024": 33
"31025": 200
"31026": 400
"31027": 3
"31028": 1
"31029": 1
"31030": 0
"31031": 0
"31032": 0
"31033": 0
"31034": 0
"31035": 0
"31036": 0
"31037": 0
"31038": 0
"31039": 0
"31040": 0
"31041": 0
"31042": 0
"31043": 0
"31044": 0

So it looks like the Elfin is passing data, but if I increase the count to 46 I get

error: >-
Error reading registers. Type: RegisterType.INPUT; start: 31000; count: 46;
slave: 247 from tcp://192.168.0.227:502: Exception Response(132, 4,
IllegalAddress)

Home Assistant debug logs are returning

2024-12-27 16:24:45.452 DEBUG (MainThread) [custom_components.foxess_modbus.modbus_controller] Reading addresses on tcp://192.168.0.227:502 247: (10016, 3)
2024-12-27 16:24:45.542 DEBUG (MainThread) [custom_components.foxess_modbus.modbus_controller] Modbus error when polling tcp://192.168.0.227:502 247: Exception Response(132, 4, IllegalAddress)

I've kind of reached my knowledge limit and would love to receive some help or guidance as to what to look at next.

Many Thanks
Mark

Answered by canton7

Jan 31, 2025

#750

View full answer

FozzieUK · 2024-12-28T09:26:12Z

FozzieUK
Dec 28, 2024

The integration is failing to read the Master, Slave, Manager firmware versions at 10016/17/18 - so it looks like the AC1-G2 may have changed the register map from the old AC1 (G1) input register maps.

In HA if you look in Devices & Services, find the Foxess_Modbus integration and click '1 Device' - what (if any) sensors do you see populated.

To start off can you use Developer action to firstly prove the H1/AC1 input range is invalid

Start Address: 10016, Count: 3, Input - and see what the response is (will likely fail)

Then try

Start Address: 30016, Count: 3, Holding - and see what the response is, paste it here if you get a valid response

If you could also try

Start Address: 10000, Count: 8, Input - and see what the response is

Start Address: 30000, Count: 16, Holding - and see what the response is

Again if you could list the responses here please.

0 replies

markheppell · 2024-12-28T09:44:37Z

markheppell
Dec 28, 2024
Author

I have 1 device, 89 entities, 13 disabled, 75 unavailable, and 1 that is showing something, which is the Connection Status which reads:-

Error reading registers. Type: RegisterType.INPUT; start: 10016; count: 3; slave: 247 from tcp://192.168.0.227:502: Exception Response(132, 4, IllegalAddress)

0 replies

FozzieUK · 2024-12-28T10:24:32Z

FozzieUK
Dec 28, 2024

ok thanks, clearly not responding to some of the input registers but ok on the holding registers.

Inverter Model is in hex 41,43,31,2D,35,2E,30,2D,45,2D,47,32 = 'AC1-5.0-E-G2'

Could do with a quick test to see whether the register map is more akin to the H1_G1 using holding regs or the H1_G2

Can you try these and record the responses please

BMS Connect State
Start Address: 31029, Count: 1, Holding
or
Start Address: 31028, Count: 1, Holding

Inverter State
Start Address: 31027, Count: 1, Holding
or
Start Address: 31028, Count: 1, Holding
or
Start Address: 39063, Count: 1, Holding
Start Address: 39065, Count: 1, Holding

0 replies

markheppell · 2024-12-28T10:28:36Z

markheppell
Dec 28, 2024
Author

Start Address: 31029, Count: 1, Holding
values:
"31029": 1
Start Address: 31028, Count: 1, Holding
values:
"31028": 1
Start Address: 31027, Count: 1, Holding
values:
"31027": 3
Start Address: 31028, Count: 1, Holding
values:
"31028": 1
Start Address: 39063, Count: 1, Holding
values:
"39063": 4
Start Address: 39065, Count: 1, Holding
values:
"39065": 0

1 reply

FozzieUK Dec 28, 2024

Hmmmm bit of both...

Ok as a quick test can you open file editor and open this file /homeassistant/custom_components/foxess_modbus/inverter_profiles.py

go to line 221and edit it change RegisterType from INPUT to HOLDING

i.e.

        InverterModelProfile(InverterModel.AC1, r"^AC1-([\d\.]+)", capacity_parser=CapacityParser.H1)
        .add_connection_type(
            Inv.H1_G1,
            ConnectionType.AUX,
            RegisterType.INPUT,
            special_registers=H1_AC1_REGISTERS,
        )

becomes

        InverterModelProfile(InverterModel.AC1, r"^AC1-([\d\.]+)", capacity_parser=CapacityParser.H1)
        .add_connection_type(
            Inv.H1_G1,
            ConnectionType.AUX,
            RegisterType.HOLDING,
            special_registers=H1_AC1_REGISTERS,
        )

And then restart home assistant.

When it starts can you check in devices & services to see what entities are showing, and also report any foxess_modbus log entries

markheppell · 2024-12-28T10:54:39Z

markheppell
Dec 28, 2024
Author

home-assistant_foxess_modbus_2024-12-28T10-48-58.661Z.log

There is now a lot of Not Provided entries, Connection Status has changed to Connected, Max Charge Current, Max Discharge current, Min SOC and Work Mode are showing values, all the rest are either Not Provided or Unknown

1 reply

FozzieUK Dec 28, 2024

ok, it's a start as the AC1 register range is being restricted.

Can you screen shot the foxess_modbus device page so I can see what entities are showing.

markheppell · 2024-12-28T11:28:04Z

markheppell
Dec 28, 2024
Author

0 replies

FozzieUK · 2024-12-28T11:31:11Z

FozzieUK
Dec 28, 2024

Ok, right a lot of restrictions because of the old AC1's input registers -

Again can you open file editor, open the file /homeassistant/custom_components/foxess_modbus/inverter_profiles.py

go to line 222 and edit it change special_registers to H1_G2_REGISTERS

i.e.

        InverterModelProfile(InverterModel.AC1, r"^AC1-([\d\.]+)", capacity_parser=CapacityParser.H1)
        .add_connection_type(
            Inv.H1_G1,
            ConnectionType.AUX,
            RegisterType.HOLDING,
            special_registers=H1_AC1_REGISTERS,
        )

becomes

        InverterModelProfile(InverterModel.AC1, r"^AC1-([\d\.]+)", capacity_parser=CapacityParser.H1)
        .add_connection_type(
            Inv.H1_G1,
            ConnectionType.AUX,
            RegisterType.HOLDING,
            special_registers=H1_G2_REGISTERS,
        )

Restart home assistant and when it restarts have a look at the devices & services page, and attach any modbus logs

0 replies

markheppell · 2024-12-28T11:43:46Z

markheppell
Dec 28, 2024
Author

Now we're looking a bit healthier.

home-assistant_foxess_modbus_2024-12-28T11-38-18.554Z.log

15 replies

markheppell Dec 29, 2024
Author

Yes 2 EP5 batteries and the fox app says 10.36 kWh

FozzieUK Dec 29, 2024

ok that's good enough for me.

If you don't mind another quick test could you - as before open the file /homeassistant/custom_components/foxess_modbus/inverter_profiles.py

go to line 219 and edit it change the type to Inv.H1_G2

i.e.

        InverterModelProfile(InverterModel.AC1, r"^AC1-([\d\.]+)", capacity_parser=CapacityParser.H1)
        .add_connection_type(
            Inv.H1_G1,
            ConnectionType.AUX,
            RegisterType.HOLDING,
            special_registers=H1_G2_REGISTERS,
        )

so it becomes

        InverterModelProfile(InverterModel.AC1, r"^AC1-([\d\.]+)", capacity_parser=CapacityParser.H1)
        .add_connection_type(
            Inv.H1_G2,
            ConnectionType.AUX,
            RegisterType.HOLDING,
            special_registers=H1_G2_REGISTERS,
        )

Restart home assistant and when it restarts have a look at the devices & services page - hopefully inverter state will be correct 'on-grid' and you will have a battery kwh remaining sensor.
Would you mind screen shotting the entities page again so I can see what is and isn't showing correctly please.

markheppell Dec 29, 2024
Author

FozzieUK Dec 29, 2024

ok, thanks - looking better.

I'll take a good look through all the sensors in the morning to cross check.

FozzieUK Jan 12, 2025

adding AC1 G2 profile type in this PR #730

markheppell · 2025-01-12T15:18:15Z

markheppell
Jan 12, 2025
Author

Excellent, let me know if you want any testing or confirmation. I did your last update and reapplied our manual changes and it stayed working.

…

________________________________ From: Dave Foster ***@***.***> Sent: Sunday, January 12, 2025 2:46:14 PM To: nathanmarlor/foxess_modbus ***@***.***> Cc: markheppell ***@***.***>; Author ***@***.***> Subject: Re: [nathanmarlor/foxess_modbus] AC1-5.0-E-G2 Illegal Data Address exception response (Discussion #715) adding AC1 G2 profile type in this PR #728<#728> — Reply to this email directly, view it on GitHub<#715 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AR2D4CAQOIY6Y6NO3HS76AD2KJ53NAVCNFSM6AAAAABUI3UN5OVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTCOBRGIZTANA>. You are receiving this because you authored the thread.Message ID: ***@***.***>

0 replies

zeclab · 2025-01-29T23:16:40Z

zeclab
Jan 29, 2025

Hi,
I've just got the AC1 G2 and have the same issue. I tried version v1.13.1b2 as I noticed your fix was in there but still didn't work for me.
I think there is issue with the inverter name as had to comment out 274-280 and change line 267 from:
InverterModel.AC1, r"^AC1-([\d\.]+)-E-G2", capacity_parser=CapacityParser.H1
To:
InverterModel.AC1, r"^AC1-([\d\.]+)", capacity_parser=CapacityParser.H1
Then all the sensors populated after a restart.
The integration shows the device model as "AC1-5.0-E-G2 - AUX".
Let me know if you like me to test anything.

13 replies

FozzieUK Jan 31, 2025

I'm scratching my head a bit, can you enable debugging logging, then remove the device and re-add it again - you should get a log entry when it starts up that shows the autodetected inverter model and extra info - can you post that please.

canton7 Jan 31, 2025
Collaborator

I'm really perplexed by this as well -- it defies all logic.

I think there is issue with the inverter name as had to comment out 274-280 and change line 267 from:
InverterModel.AC1, r"^AC1-([\d.]+)-E-G2", capacity_parser=CapacityParser.H1
To:
InverterModel.AC1, r"^AC1-([\d.]+)", capacity_parser=CapacityParser.H1
Then all the sensors populated after a restart.

Can you post your full file after you edited it, and after confirming that your edits fixed foxess_modbus to the point where all sensors were updated correctly and there were no errors or warnings in the log?

canton7 Jan 31, 2025
Collaborator

Argh, think I've got it.

canton7 Jan 31, 2025
Collaborator

#750

Answer selected by canton7

zeclab · 2025-01-30T11:21:07Z

zeclab
Jan 30, 2025

Thanks for the response. Here is the output you requested: values: "30000": 16707 "30001": 12589 "30002": 13614 "30003": 12333 "30004": 17709 "30005": 18226 "30006": 8224 "30007": 8224 "30008": 0 "30009": 0 "30010": 0 "30011": 0 "30012": 0 "30013": 0 "30014": 0 "30015": 0

…

On Thu, 30 Jan 2025, 09:01 Dave Foster, ***@***.***> wrote: Thanks, looks like another 'subversion' of H1 G2 to test for - If you could go to Developer Tools, Actions and select the action foxess modbus: read registers then see what the response is - Start Address: 30000, Count: 16, Holding And if you could post the response here. — Reply to this email directly, view it on GitHub <#715 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AIWM4426GYKXSLSVIIYMFU32NHTANAVCNFSM6AAAAABUI3UN5OVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTEMBQGQZTOMY> . You are receiving this because you commented.Message ID: ***@***.*** com>

1 reply

canton7 Jan 31, 2025
Collaborator

Can you confirm that this was definitely reading holding registers, not input registers?

zeclab · 2025-01-31T17:19:25Z

zeclab
Jan 31, 2025

Yup that fixed it for me. Thanks both!

…

On Fri, Jan 31, 2025 at 11:35 AM Antony Male ***@***.***> wrote: #750 <#750> — Reply to this email directly, view it on GitHub <#715 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AIWM443TNPSMDWLI2NBY3FT2NNNYBAVCNFSM6AAAAABUI3UN5OVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTEMBRGY4TGNA> . You are receiving this because you were mentioned.Message ID: ***@***.*** com>

1 reply

canton7 Jan 31, 2025
Collaborator

Thanks for confirming. I'll get that released soon.

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

AC1-5.0-E-G2 Illegal Data Address exception response #715

{{title}}

Replies: 12 comments 32 replies

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

Select a reply

AC1-5.0-E-G2 Illegal Data Address exception response #715

Replies: 12 comments · 32 replies

markheppell Dec 28, 2024 Author

markheppell Dec 28, 2024 Author

markheppell Dec 28, 2024 Author

markheppell Dec 28, 2024 Author

markheppell Dec 28, 2024 Author

markheppell Dec 29, 2024 Author

markheppell Dec 29, 2024 Author

markheppell Jan 12, 2025 Author

canton7 Jan 31, 2025 Collaborator

canton7 Jan 31, 2025 Collaborator

canton7 Jan 31, 2025 Collaborator

canton7 Jan 31, 2025 Collaborator

canton7 Jan 31, 2025 Collaborator

Replies: 12 comments 32 replies

markheppell
Dec 28, 2024
Author

markheppell
Dec 28, 2024
Author

markheppell
Dec 28, 2024
Author

markheppell
Dec 28, 2024
Author

markheppell
Dec 28, 2024
Author

markheppell Dec 29, 2024
Author

markheppell Dec 29, 2024
Author

markheppell
Jan 12, 2025
Author

canton7 Jan 31, 2025
Collaborator

canton7 Jan 31, 2025
Collaborator

canton7 Jan 31, 2025
Collaborator

canton7 Jan 31, 2025
Collaborator

canton7 Jan 31, 2025
Collaborator