un-sufficient AM modulation index under large signal

[welljustin]

location of modulation index in signal

Hallo,

I have a problem of the un-sufficient AM modulation index under large signal generated by RF signal generator.
The test setup is as follows:
I use HP33120A arbitrary waveform generator to create the digital baseband signal(2KHz) for the ASK modulation, and connect the output to the modulation input of the signal generator, HP8647A. And then I set the RF output as **MHz/100%AM. And when I try to measure the maximum power that the receiver can handle, the problem shows up. From the HP8591E spectrum analyzer, it seems that when the RF output power is higher than a threshold, like -20dBm, the AM modulation index will never be 100% again. The higher the power is, the less the modulation index is. And when the signal is weak, everything is ok.
I have checked the datasheet of the signal generator, unfortunately, I can not find the answer, and who can help me a little bit here ?
Thanks a lot in advance!

 

[FvM]

am modulation index

Sounds like wrong setup of signal generator or spectrum analyzer. You may want to check the output signal in time domain with an oscilloscope first. I don't know 8647A operation details, but 8648A has an AM pulse mode, that should be used for ASK in my opinion.

 

[welljustin]

modulation index in am

I do not believe that this is because of the wrong setup of the signal generator or FSA. Actually, I think this maybe caused by the voltage of the modulation input. And I am trying now. You know, the RF signal is actually generated by some kind of mixer inside, not the same as the pulse modulation. Unfortunately, the pulse modulation is not available with HP8647A.
Thanks .

Added after 14 minutes:

Now I can get the correct signal. I did the following:
1. Because there is always no any problem with the signal internally generated by the waveform generator, and the signal are all minus and plus symmetrical. So I reprogram my own signal to the one with minus and plus value.
2. Adjust the output level to a certain voltage, then it works.

The problem, I guess, could be the following:
1. maybe it is because of the AC coupling ?
2. maybe the dc offset for the modulation input will bring some deviation to the RF signal ?
But I still have no answer for this.
Who can explain this a little bit from the issues I described.