Pymodbus read holding registers

Question

I was assigned to perform the task without any documentation. I have a problem with reading data from MODBUS. This is the script that I was able to create:

from pymodbus.constants import Endian
from pymodbus.payload import BinaryPayloadDecoder
from pymodbus.payload import BinaryPayloadBuilder
from pymodbus.client.sync import ModbusTcpClient

client = ModbusTcpClient('X.X.X.X')
connection = client.connect()

request = client.read_holding_registers(12606,2)
result = request.registers
decoder = BinaryPayloadDecoder.fromRegisters(result, Endian.Big, wordorder=Endian.Little)
print "Counter1: %0.2f" % decoder.decode_32bit_float()

request = client.read_holding_registers(12482,2)
result = request.registers
decoder = BinaryPayloadDecoder.fromRegisters(result, Endian.Big, wordorder=Endian.Little)
print "Counter2: %0.2f" % decoder.decode_32bit_float()

client.close()

Everything looks fine, But the data on the counter is different from those in the script for example:

Value on the counter : 39558853.30 (value is decimal)
Value from the script: 58853.30
(value is decimal)

Read input registers (HEX): E54D 4765

And this is how the address documentation looks like"

P   12458       Q2  4\DW12458 = 1\ND20_Q2\P(F)
Q   12462       Q2  4\DW12462 = 1\ND20_Q2\Q(F)
S   12466       Q2  4\DW12466 = 1\ND20_Q2\S(F)
I   12470       Q2  4\DW12470 = 1\ND20_Q2\I(F)
U   12474       Q2  4\DW12474 = 1\ND20_Q2\U(F)
f   12478       Q2  4\DW12478 = 1\ND20_Q2\f(F)
EP_POB  12482       Q2  4\DW12482 = 1\ND20_Q2\EP_POB(F)
EP_ODD  12486       Q2  4\DW12486 = 1\ND20_Q2\EP_ODD(F)
EQ_IND  12490       Q2  4\DW12490 = 1\ND20_Q2\EQ_IND(F)
EQ_POJ  12494       Q2  4\DW12494 = 1\ND20_Q2\EQ_POJ(F)
THDVL1  12498       Q2  4\DW12498 = 1\ND20_Q2\THDVL1(F)
THDVL2  12502       Q2  4\DW12502 = 1\ND20_Q2\THDVL2(F)
THDVL3  12506       Q2  4\DW12506 = 1\ND20_Q2\THDVL3(F)
THDIL1  12510       Q2  4\DW12510 = 1\ND20_Q2\THDIL1(F)
THDIL2  12514       Q2  4\DW12514 = 1\ND20_Q2\THDIL2(F)
THDIL3  12518       Q2  4\DW12518 = 1\ND20_Q2\THDIL3(F)
UL1 12522       Q2  4\DW12522 = 1\ND20_Q2\UL1(F)
UL2 12526       Q2  4\DW12526 = 1\ND20_Q2\UL2(F)
UL3 12530       Q2  4\DW12530 = 1\ND20_Q2\UL3(F)
IL1 12534       Q2  4\DW12534 = 1\ND20_Q2\IL1(F)
IL2 12538       Q2  4\DW12538 = 1\ND20_Q2\IL2(F)
IL3 12542       Q2  4\DW12542 = 1\ND20_Q2\IL3(F)
PL1 12546       Q2  4\DW12546 = 1\ND20_Q2\PL1(F)
PL2 12550       Q2  4\DW12550 = 1\ND20_Q2\PL2(F)
PL3 12554       Q2  4\DW12554 = 1\ND20_Q2\PL3(F)
QL1 12558       Q2  4\DW12558 = 1\ND20_Q2\QL1(F)
QL2 12562       Q2  4\DW12562 = 1\ND20_Q2\QL2(F)
QL3 12566       Q2  4\DW12566 = 1\ND20_Q2\QL3(F)
S1  12570       Q2  4\DW12570 = 1\ND20_Q2\S1(F)
S2  12574       Q2  4\DW12574 = 1\ND20_Q2\S2(F)
S3  12578       Q2  4\DW12578 = 1\ND20_Q2\S3(F)

Answer 1

I improved your code as follows:

from pymodbus.constants import Endian
from pymodbus.payload import BinaryPayloadDecoder
from pymodbus.client.sync import ModbusTcpClient

def validator(instance):
    if not instance.isError():
        '''.isError() implemented in pymodbus 1.4.0 and above.'''
        decoder = BinaryPayloadDecoder.fromRegisters(
            instance.registers,
            byteorder=Endian.Big, wordorder=Endian.Little
        )   
        return float('{0:.2f}'.format(decoder.decode_32bit_float()))

    else:
        # Error handling.
        print("The register does not exist, Try again.")
        return None


client = ModbusTcpClient('X.X.X.X', port=502)  # Specify the port.
connection = client.connect()

if connection:
    request = client.read_holding_registers(12606, 2, unit=1)  # Specify the unit.
    data = validator(request)
    print(data)

    request = client.read_holding_registers(12482, 2, unit=1)  # Specify the unit.
    data = validator(request)
    print(data)

    client.close()

else:
    print('Connection lost, Try again')

---

[NOTE]:

Are you sure about the desired float32 decoding?

1. float AB CD == byteorder=Endian.Big, wordorder=Endian.Big
2. float CD AB == byteorder=Endian.Big, wordorder=Endian.Little
3. float BA DC == byteorder=Endian.Little, wordorder=Endian.Big
4. float DC BA == byteorder=Endian.Little, wordorder=Endian.Little

Set the unit_id:

• In many cases unit is 1 as default.

---

[UPDATE]:

Maybe you need to read and decode the result as double/float64 value for the 12482 register address, because I think when the respective register in the doc is 12482 and the next register is 12846, you need to read 4regs — float64/double:

request = client.read_holding_registers(12482, 4, unit=1)

And

return float('{0:.2f}'.format(decoder.decode_64bit_float()))

Answer 2

I'm pretty sure the value 39558853.30 is to large to store in a IEEE single precision float. There are 7.22 digits of precision, that number requires 9 digits. I did some experimentation with assigning the value to float and double values in C# confirms this.

That leads me believe:

1) Like Benyamin Jafari suggested, you need to read four registers. However, that number (as a double) in hex is 0x4182dcf62a666666 which doesn't seem to correspond any of the data that you are reading.

OR

2) It is also possible that it is returned as a UINT32 which must be scaled by (1/100) to give you what the counter is showing. 0xE54D4765 = 3847047013 => scaled by 1/100.0 = 38470470.13 which is close to what you are seeing on the counter. In my experience, this is a common practice in Modbus.

OR

3) They are using some other (non-standard) format to represent the data.

Can you give us the product name, model, etc?