As I said in my first message, I've tired to monitor the instrument with the windows hyperterminal, and there wasn't any additional character.
Then I tried to monitor the instrument with NI Max (using VISA), and the unexpected characters occured.
See the 2 attached files (one for the hyperterminal, and the other for MAX)
HyperTerminal is not a monitoring program. You also do not monitor using MAX. To monitor you need something else, which is why I suggested PortMon, which allows you to see explicitly the bytes that are sent over the serial port. NI-Spy will do this as well, but it won't work with HyperTerminal. You chose to use HDD Software Serial Monitor, which is fine. Of course, your screenshots show no difference between the communication done with HyperTerminal and the communication done with MAX. Both show the same exact response to the *IDN? command.
Something doesn't look right with those captures. And by that I don't mean the received data. I'm referring to the timestamps. In the one for HyperTerminal the response from *RST *IDN? comes about 150 msec after the command. In the MAX case it comes 10.6 seconds later. This makes no sense to me. Additionally, the HyperTerminal capture shows the second set of command being split, with a 12 second gap. The MAX command set was sent 15 seconds after the response from *IDN? and the response coming 2.5 seconds later.
How exactly are you sending these commands in MAX? And how are you sending these commands in HyperTerminal? Are you uploading a file with the command in there?
You said you tried to do this in LabWindows and LabVIEW. Can you upload your actual code?