search

StephanJoubert / home_assistant_solarman

Home Assistant component for Solarman collectors used with a variety of inverters.
Apache License 2.0
550 stars 199 forks source link

Request to copy your code #96

Open AdyRock opened 2 years ago

AdyRock commented 2 years ago

Hi,

Would you mind if I copy your code and convert it to Node.js for use with Homey?

And to be really cheeky, could you provide a typical request packet for a dummy sn, etc, so I can compare the output from my code to ensure I have got things right?

StephanJoubert commented 2 years ago

Hi @AdyRock , no problem. I wanted to do Node.js as well, but too much stuff on my plate as usual. Will see if I can get hold of some sequences for you.

AdyRock commented 2 years ago

Thank you, much appreciated. How do you get the MODBUS register information for various inverters? I have a Sofar HYD 3600-ES so not sure if the description files in your repo will work for me.

AdyRock commented 2 years ago

I have had some success. My inverter has a different register layout, so I have made a new description file and I now get valid data back most of the time. However, I find occasionally the data is crazy and it seem that the response has the wrong number of bytes. The Solarman V5 packets are perfectly formed with the correct check sum, etc, but the MODBUS payload is fragmented within it, so the first V5 data packet returns the start of the MODBUS registers and then, after the next request, I get a second V5 packet with the remaining MODUS registers. As the code doesn't verify the MODBUS CRC, the result is it decodes the values incorrectly.

Edit: also, I don't receive the second packet until I send a new request, so do you know if there is meant to be any acknowledgment that a packet has been received that might trigger the remaining data to be sent?

Have you seen anything like this?

StephanJoubert commented 2 years ago

I have not seen that behavior exactly, but found that when the registers has a gap, it works better if the request is split into multiple requests, with each one requesting contiguous registers.

AdyRock commented 2 years ago

Yep, I have done that. Maybe I will reduce the block size as my first block goes from 512 to 547, which, in theory, all contain valid data.

igieon commented 2 years ago

@AdyRock i think modbus protocol (registers) is same as in my inverter only that is single phase so your interval is not same as me https://github.com/StephanJoubert/home_assistant_solarman/blob/main/custom_components/solarman/inverter_definitions/sofar_g3hyd.yaml. And when i look into your range 0x512-0x547 its outside of values so probably look into this file and mask which is correct register adress. I think file is correct https://www.facebook.com/groups/2477195449252168/permalink/2895353217436387/.

AdyRock commented 2 years ago

Ah, sorry, the register numbers I gave are in decimal (JSON can't cope with hex), so the a <html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:x="urn:schemas-microsoft-com:office:excel" xmlns="http://www.w3.org/TR/REC-html40">

Name | Decimal reg | Hex reg -- | -- | -- Running state | 512 | 200 Fault message 1 | 513 | 201 Fault message 2 | 514 | 202 Fault message 3 | 515 | 203 Fault message 4 | 516 | 204 Fault message 5 | 517 | 205 Grid A Voltage | 518 | 206 Grid A Current | 519 | 207 Grid B Voltage | 520 | 208 Grid B Current | 521 | 209 Grid C Voltage | 522 | 20A Grid C Current | 523 | 20B Grid Frequency | 524 | 20C Charge/Discharge power | 525 | 20D Battery Voltage | 526 | 20E Battery Charge/Discharge current | 527 | 20F The residual capacity of battery | 528 | 210 Battery Temperature | 529 | 211 Feed in/out power | 530 | 212 The power of the load | 531 | 213 Input/Output power | 532 | 214 The power of generation | 533 | 215 EPS output voltage | 534 | 216 EPS output power | 535 | 217 Generation of one day | 536 | 218 The power sell to grid of today | 537 | 219 The power buy from grid of toda | 538 | 21A Today consumption of the load | 539 | 21B Total generation H Byte | 540 | 21C Total generation L Byte | 541 | 21D Total power sell to grid H Byte | 542 | 21E Total power sell to grid L Byte | 543 | 21F Total energy buy from grid H Byte | 544 | 220 Total energy buy from grid L Byte | 545 | 221 Total Consumption H Nyte | 546 | 222 Total Consumption L Byte | 547 | 223 Reseved | 548 | 224 Reseved | 549 | 225 Reseved | 550 | 226 Reseved | 551 | 227 Reseved | 552 | 228 Reseved | 553 | 229 The countdown time | 554 | 22A Inverter alert message | 555 | 22B Battery Cycle times | 556 | 22C INV bus voltage | 557 | 22D LLC bus voltage | 558 | 22E Buck current | 559 | 22F Grid R Voltage | 560 | 230 Grid R Current | 561 | 231 Grid S Voltage | 562 | 232 Grid S Current | 563 | 233 Grid T Voltage | 564 | 234 Grid T Current | 565 | 235 Generation current | 566 | 236 Battery power | 567 | 237 Inner temperature | 568 | 238 Heat sink temperature | 569 | 239 Country | 570 | 23A Current dc component | 571 | 23B Voltage dc component | 572 | 23C Battery fault message 1 | 573 | 23D Battery fault message 2 | 574 | 23E Battery fault message 3 | 575 | 23F Battery fault message 4 | 576 | 240 Battery fault message 5 | 577 | 241 Communication board inner message | 578 | 242 Today generation time | 579 | 243 Total generation time High-Byte | 580 | 244 Total generation time,Low-Byte | 581 | 245

AdyRock commented 2 years ago

After splitting the ranges a bit more, I'm still getting the occasional split MODBUS packet. The data is all good, it's just the first V5 packet is missing the data from the last few registers of the MODBUS payload and the second V5 packet contains the remaining register data from the first request plus the data from the second request.