QAM

[Sharmi]

Hi,

I am working on a project to transmit a QAM signal and demodulate the same at the receiver using PXI.

I am using PXI 5673 and 5663 for transmission and reception respectively.I found a problem, which I do not know how to solve.

In the example MT RFSG Generate QAM vi, if I set the 'QAM M-ary' parameter as 4,16 or 64 and the 'Samples per Symbol' as 2,4,6 respectively, it works fine. However, if I set  'QAM M-ary' parameter as 8 or 32 and  the 'Samples per Symbol' as 3 or 5 respectively, I get the following error-

Property Node (arg 5) in MT RFSG Generate QAM.vi <ERR>
The AWG reported the following error:

Invalid value for parameter or property.

Property: IQ Rate
Requested Value: Inf
Valid Values Begin with: 6.103600e3
Valid Values End with: 200.0e6

Status Code: -200077

If I set the number of stages/symbols as 8 (or 32), should I not have 3 (or 5) samples per symbol?

Could someone please tell me why I am getting this error?

Thanks so much,

Sharmi

[Peter-E]

Hi Sharmi!

The Help File for the function that accepts the Samples per Symbol, "MT Generate QAM System Parameters.vi", states that the Samples per Symbol input needs to be "an even, positive number of samples dedicated to each symbol". 

This would be the reason that this function throws the error whenever you specify Samples per Symbol as odd. I tried the 'QAM M-ary' parameter as 4, 16, and 64 with odd numbers of Samples per Symbol and still receive the error.

Please let me know if you have any additional questions! Have a great day!

[Sharmi]

Hi Peter,

Thanks so much for the reply.
Then how does it work with the number of samples per symbol and the number of symbols? Is it still following the equation below for cases like 8 and 32 symbols?

No of samples per symbol = log₂Number of symbols 

Thanks so much,

Sharmi

[Peter-E]

Hi Sharmi!

I completely understand the confusion. The 'number of symbols' in modulation techniques correspond to the different mappings that the I/Q data can correspond to. The following equation, similar to the one in your last response, relates the number of symbols to the 'M'-QAM:

Number of symbols = log₂(M) 

The 'samples per symbol' parameter for generation just offers more samples to represent the IQ data (essentially "oversampling"), and therefore reduces the error during demodulation. This does not change the number of bits/symbol or the mapping, but simply the amount of data generated to represent those symbols.

I hope this helps!