Serial Port Communication Protocol

Question

In serial communication with devices such as a digital Multimeter (ex. BK Precision 2831E), why do I need to send a query command once but read the output twice? For instance, I sent a query command for the voltage measured, and received an echo but no value of voltage. I then sent the query command twice which returned the echo and the measured voltage. In essence, to read out the voltage measured, I had to send the same query command in succession twice. I do not understand this concept. Can anyone kindly help me out with this reasoning.

I have attached a sample code here below:

def readoutmm(portnumber_multimeter):
    import serial
    import time
    ser2 = serial.Serial(
    port="com"+str(portnumber_multimeter),
    baudrate=9600,
    bytesize=serial.EIGHTBITS,  
    parity=serial.PARITY_NONE,
    stopbits=serial.STOPBITS_ONE
    )
    ser2.write(b'fetc?\n') # Query command
    voltage= ser2.readline() # Returns echo
    voltage=ser2.readline() # Returns measured voltage
    voltage=float(voltage)
    ser2.close()
    packet=[voltage]
    return packet

score 0 · Accepted Answer · answered Apr 21 '22 at 11:55

This is actually quite common with devices based on RS232/RS485 protocols.

From the manual of the machine you mentioned, I quote:

The character received by the multimeter will be sent back to the controller again. The controller will not send the next character until the last returned character is received correctly from the meter. If the controller fails to receive the character sent back from the meter, the possible reasons are listed as follows:

The serial interface is not connected correctly.

Check if the same baud rate is selected for both the meter and the controller.

When the meter is busy with executing a bus command, it will not accept any character from the serial interface at the same time. So the character sent by controller will be ignored.

In order to make sure the whole command is sent and received correctly, the character without a return character should be sent again by the controller.

On a lot of devices this is actually a setting which you can turn on and off.

Now, as for your question:

why do I need to send a query command once but read the output twice?

You are supposed to read every character back before sending a new one to validate if the character was received correctly. But in your code are actually sending all the character before reading a single one of them.

In scenario's where you have a reliable connection, you method will work as well, but as a consequence you'll need to read twice; once to validate if the command was received and the second time to retrieve the actual data.

Do keep in mind that read buffers might be limited to a certain amount. If you are experiencing unexpected behavior while querying large amount of data and sending a lot of commands, it might be due to the fact these buffers are full.

That makes this concept very clear. Thanks for the help and taking out time — helsinki, Apr 21 '22 at 12:45
To further clarify on this topic of discussion: As you said, the read buffers might have limited storage, is there any way I can exception handle this to prevent my algorithm from crashing? Or any other solution? — helsinki, Apr 21 '22 at 13:21
As long as the buffers are read, you're good and a crash is unlikely. You can always uses pythons [try-except](https://www.codingem.com/try-catch-in-python/) construction.But this will stop execution of the algorithm. I would approach this in a pragmatic way: the amount of data returned will be within specified boundaries. As long as you send the command and directly read the result, it should not cause any issues. — Stefan, Apr 21 '22 at 13:34

Serial Port Communication Protocol

1 Answers1