Alfen Eve Single S-line (NG910) zeigt seltsames Modbus Verhalten

[rdorsch]

Hallo,

ich vesuche gerade von der Alfen per Modbus TCP die "Socket measurement registers" mit pymodbus auszulesen.

Dabei produziert

            # A modbus message must not exceed 256 bytes. Since this limits messages to a
            # somewhat lower length then 128 (125 worked) due to header bytes. To be on the
            # safe side, split the transaction into two reads.
            second_read_length=16*4
            first_read_length=126-second_read_length
            rr = self.modbus_client.read_holding_registers(300,
                count=first_read_length,  # count: number of 16 bit values
                slave=1)
            
            decoder = pymodbus.payload.BinaryPayloadDecoder.fromRegisters(rr.registers,
                byteorder=pymodbus.constants.Endian.Big,
                wordorder=pymodbus.constants.Endian.Big)

            rr = self.modbus_client.read_holding_registers(300+first_read_length,
                count=second_read_length,  # count: number of 16 bit values
                slave=1)
            decoder2 = pymodbus.payload.BinaryPayloadDecoder.fromRegisters(rr.registers,
                byteorder=pymodbus.constants.Endian.Big,
                wordorder=pymodbus.constants.Endian.Big)

            self.energy_measurements = {
                'meter_state': decoder.decode_16bit_uint(),
                'meter_last_value_timestamp': decoder.decode_64bit_uint(),
                'meter_type': decoder.decode_16bit_uint(),
                'V_L1_N': decoder.decode_32bit_float(),             # unit: Volt
                'V_L2_N': decoder.decode_32bit_float(),
                'V_L3_N': decoder.decode_32bit_float(),
                'V_L1_L2': decoder.decode_32bit_float(),
                'V_L2_L3': decoder.decode_32bit_float(),
                'V_l3_L1': decoder.decode_32bit_float(),
                'I_N': decoder.decode_32bit_float(),
                'I_L1': decoder.decode_32bit_float(),
                'I_L2': decoder.decode_32bit_float(),
                'I_L3': decoder.decode_32bit_float(),
                'I_SUM': decoder.decode_32bit_float(),
                'P_FACTOR_PH_L1': decoder.decode_32bit_float(),
                'P_FACTOR_PH_L2': decoder.decode_32bit_float(),
                'P_FACTOR_PH_L3': decoder.decode_32bit_float(),
                'P_FACTOR_PH_SUM': decoder.decode_32bit_float(),
                'FREQ': decoder.decode_32bit_float(),               # unit: Hz
                'REAL_P_PH_L1': decoder.decode_32bit_float(),
                'REAL_P_PH_L2': decoder.decode_32bit_float(),
                'REAL_P_PH_L3': decoder.decode_32bit_float(),
                'REAL_P_SUM': decoder.decode_32bit_float(),
                'APPARENT_P_PH_L1': decoder.decode_32bit_float(),
                'APPARENT_P_PH_L2': decoder.decode_32bit_float(),
                'APPARENT_P_PH_L3': decoder.decode_32bit_float(),
                'APPARENT_P_SUM': decoder.decode_32bit_float(),
                'REACTIVE_P_PH_L1': decoder.decode_32bit_float(),
                'REACTIVE_P_PH_L2': decoder.decode_32bit_float(),
                'REACTIVE_P_PH_L3': decoder.decode_32bit_float(),
                'REACTIVE_P_SUM': decoder.decode_32bit_float(),     # unit: Watt
                'REAL_E_DELIVERED_L1': decoder2.decode_64bit_float(),
                'REAL_E_DELIVERED_L2': decoder2.decode_64bit_float(),
                'REAL_E_DELIVERED_L3': decoder2.decode_64bit_float(),
                'REAL_E_DELIVERED_SUM': decoder2.decode_64bit_float(),
                'REAL_E_CONSUMED_L1': decoder2.decode_64bit_float(),
                'REAL_E_CONSUMED_L2': decoder2.decode_64bit_float(),
                'REAL_E_CONSUMED_L3': decoder2.decode_64bit_float(),
                'REAL_E_CONSUMED_SUM': decoder2.decode_64bit_float(),
                'APPARENT_E_L1': decoder2.decode_64bit_float(),
                'APPARENT_E_L2': decoder2.decode_64bit_float(),
                'APPARENT_E_L3': decoder2.decode_64bit_float(),
                'APPARENT_E_SUM': decoder2.decode_64bit_float(),
                'REACTIVE_E_L1': decoder2.decode_64bit_float(),
                'REACTIVE_E_L2': decoder2.decode_64bit_float(),
                'REACTIVE_E_L3': decoder2.decode_64bit_float(),
                'REACTIVE_E_SUM': decoder2.decode_64bit_float(),
            }​

viele NaN-Werte:

Energy measurements:
{
"meter_state": 3,
"meter_last_value_timestamp": 171,
"meter_type": 0,
"V_L1_N": 231.12998962402344,
"V_L2_N": 231.61000061035156,
"V_L3_N": 233.89999389648438,
"V_L1_L2": NaN,
"V_L2_L3": NaN,
"V_l3_L1": NaN,
"I_N": NaN,
"I_L1": 0.0,
"I_L2": 0.0,
"I_L3": 0.0,
"I_SUM": NaN,
"P_FACTOR_PH_L1": NaN,
"P_FACTOR_PH_L2": NaN,
"P_FACTOR_PH_L3": NaN,
"P_FACTOR_PH_SUM": 0.0,
"FREQ": 49.97500228881836,
"REAL_P_PH_L1": NaN,
"REAL_P_PH_L2": NaN,
"REAL_P_PH_L3": NaN,
"REAL_P_SUM": 0.0,
"APPARENT_P_PH_L1": NaN,
"APPARENT_P_PH_L2": NaN,
"APPARENT_P_PH_L3": NaN,
"APPARENT_P_SUM": NaN,
"REACTIVE_P_PH_L1": NaN,
"REACTIVE_P_PH_L2": NaN,
"REACTIVE_P_PH_L3": NaN,
"REACTIVE_P_SUM": NaN,
"REAL_E_DELIVERED_L1": NaN,
"REAL_E_DELIVERED_L2": NaN,
"REAL_E_DELIVERED_L3": NaN,
"REAL_E_DELIVERED_SUM": 2485612.0,
"REAL_E_CONSUMED_L1": NaN,
"REAL_E_CONSUMED_L2": NaN,
"REAL_E_CONSUMED_L3": NaN,
"REAL_E_CONSUMED_SUM": NaN,
"APPARENT_E_L1": NaN,
"APPARENT_E_L2": NaN,
"APPARENT_E_L3": NaN,
"APPARENT_E_SUM": NaN,
"REACTIVE_E_L1": NaN,
"REACTIVE_E_L2": NaN,
"REACTIVE_E_L3": NaN,
"REACTIVE_E_SUM": NaN
}​

Das Encoding habe ich dem Alfen Dokument

Modbus Slave TCP/IP
Implementation of Modbus Slave
TCP/IP for Alfen NG9xx platform
Author: T. Nederlof
Version: 2.3
30-10-2020

entnommen. Da einige Werte durchaus plausible Werte haben, nehme ich an, dass die Vorgehensweise nicht fundamental falsch ist.

Weiß jemand, ob das Encoding vielleicht falsch in dem Dokument angegeben ist oder seht Ihr die gleiche Ausgabe?

Danke und Gruß
Rainer

[rdorsch]

Wahrscheinlich ist das durch den satz im Alfen Dokument abgedeckt:

• Whenever a register is reserved or not available, the register reply is filled with Not a Number (NaN), which is set to 0xFFFF for a 16 bit register.