 |
|
 |
DAC Bench Measurement examples
ADS5500 14-Bits, 125MSPS
|
|
|
 |
FFT shown as a curve.
Note in the Upper Right corner the results computed from the FFT.
For each harmonic position ia marker is displayed. In this case, user asked for 12 harmonics total (Fundamental + 11 harmonics). |
 |
As
for each dispalyed curve, a zoom can be done from mouse. As about 1 M
points were sampled, it means the FFT has about 500 k points on X axis,
when displayed 'full'. Here, the zoom is around the fundamental of input signal to DUT. The black part of curve is the FFT uncorrected, as produced from mathematical calculus. The corrected (used for any computed results containing Noise) is the red part : -Only one point is retained at Maximum value. -Bottom line is computed from theorical noise floor (according to DUT number of bits) -Width of correction is computed from time windowing applied (here a Blackman Window). Note that this correction is someting not often done when working on FFT, but is capital to enable accurate evaluation of noise level. |
 |
Here the zoom is around the 1st harmonic. Uncorrected and corrected FFT curves are visible as well as marker exactly at computed harmonic position. |
 |
This screen shows same measuremen, but with a
Rectangular time windowing. There are 12 different time windowing
available as standard, more may be added as long as mathematical
definition is available. Note that time windowing would not be necessary to get correct evaluations from time measrement as long as the sampling is coherent (an odd number of period exactly sampled). But due to instruments resolutions, that may not be an available time sampling. In such case, conducting calculus from FFT without windowing would gives eroneous results. |
 |
Here only harmonics values are shown, number of displayed harmonics is from user choice. |
 |
From the measurement in time domain, user may choose : -Time Windowing to apply -Number of harmonics to work with (other would be considered as 'noise' for other calculus) -Suppress or not the DC level User may change these parameters without need to ask for a new measurement, as original time sampling stay the same one. |
 |
Here is "RIF" information tab, it means Result InFormantions. EMA gives user lot of complementary informations about measurement : -Drivers sotware revision -Signal Level in dBm of RF generator for 0 dBFS (if user choice is to settle level refered to Full Scale) -Number of adjustments needed to reach 0 dBFS -Effective Sampling Frequency, and OverSampling Frequency if any -Effective Signal Frequency -Effective number of samples -Effective number of signal periods within time record -DUT Mid Code value -Voltage and Intensity for each Power Supply -Timestamp for Start & End |
 |
Here is the tab used to selct execution of measurement(s). As for any EMA measurement procedure, user may choose from : -Single Shot -Continuous -Sweep (one or two imbricated parameters may be sweeped) |
 |
Here is tab dedicated to AdMeas measurement procedure, and sub-tab dedicated to Sinus signal settling: -User may chose to ask for coherent sampling, with priority to signal or sampling frequency -User may also choose to use 'as it', without correction and possibly for not coherent sampling -User gives input signal level in dBFS -User may choose to adjust input signal level refered to 0dBFS on DUT or not |
 |
Here is tab dedicated to Logic Analyzer setup, sub-tab SETUP is shown, where user may adjust : -Sampling size -Data Threshold -Clock Threshold |
Return to EMA Main Page
BOURBAKY
BP 36 - 13, Rue des Alpes - 07302 TOURNON
Cedex - France
