TSW14J57EVM: ADC12J4000EVM

Hi Richard

The reason for the 500 MHz recommendation is that there is a balun on the ADC12J4000EVM that is used to convert the single-ended input to differential for the ADC inputs. The thin-film transmission line type balun (https://cdn.anaren.com/product-documents/Xinger/BalunTransformers/B0430J50100AHF/B0430J50100AHF_DataSheet(Rev_D).pdf)

frequency response is optimized for 400 MHz to 3.0 GHz. Below 400 MHz the insertion loss increases.

With a 250 MHz input signal there will be increased attenuation and somewhat degraded harmonic distortion performance due to worse phase and amplitude balance in the balun outputs.

Lower performance signal generators will have distortion and possibly noise in the output. Even high quality RF signal generators will frequency have some levels of harmonic distortion in the outputs which must be filtered out to prevent degradation in the ADC performance measurements. For that reason we always use either bandpass or lowpass filters to remove these harmonics and also minimize noise in the test signal applied to the ADC.

Can you capture a screen shot of the FFT you are seeing? That will allow us to evaluate the captured signal and provide recommendations and further comment.

The most informative capture configuration would be done using DDC Bypass mode so we can see the full Nyquist zone of the converter. You can save a screen shot from High Speed Data Converter Pro using the 'File>Save Screen shot as' command. 

You can attach this to E2E post by clicking on the 'Insert Code, Attach Files and more....' link at the bottom right of the Reply window. This provides more options for attachments, formatting, etc.

Best regards,

Jim B

Hi Richard

I think the primary issue is that the input signal amplitude is too large and is being clipped at the ADC full scale limits.

Try reducing the output of the signal generator until the minimum time domain values are above code 0 and the maximum values are below code 4095. The FFT should look a lot better once the amplitude is corrected.

Once the signal level is below full scale, fine-tune the signal generator amplitude until the FFT Fund. level is -1dBFS. Then send another FFT snapshot for analysis.

Best regards,

Jim B

And here I have a 300mvpp

Hi Richard

That looks a lot better. For Fin = 240 MHz and -1dBFS the THD and SNR performance of the ADC12J4000 should be better than what you are seeing.

I expect the remaining limitations are due to the signal generator you have available, and the fact that the input frequency is below the optimal range for the on-board balun.

If possible please try testing with a signal source that can provide 400 MHz or higher, and use a bandpass filter at the test frequency, or a lowpass filter slightly above the test frequency to block out noise and harmonics produced by the generator.

Best regards,

Jim B