Parts on Detail
Diamond Detector Secondary Shower Acquisiton System
ICECAL_V3 Mezzanine > ADS8138: Linearity, Noise and Dynamic Performance
1. The Set-Up and parts tested
The different tests performed here only concern the ADC ADS8138 contained on the ICECAL_V3 Mezzanine board. Since no ADC Drivers are present on this board, a SMA cable containing a wideband Coilcraft 1:1 transformer has been included and AC coupled to the scope with a 100nC capacitor in order to provide a differential signal for the dynamic characterization, the bandwidth of this transormed is specified to be between 0.4-600Mhz.

2. ADC Dynamic Parameters
*
Input Signal 400Khz,
Input Signal 1Mhz,
Input Signal 2.5Mhz,
Input Signal 4Mhz,
Input Signal 10Mhz*
ADC Dynamic parameters for 2.5Mhz Sinusoidal Input:
2.1. Graphical Summary of ADC Dynamic Parameters
3. Noise Distribution
The noise measurements on this ADC were done two times, the first one only considering the ADC inputs with a pull-up resistor to VCM, the second one connecting the wideband transformer and the SMA cable to the scope (wihle the scope was off) a window of 2.5ms is taken per channel (100000 points per channel @ 40MSPS) for the first measurement is taken, wheres for the second configuration the window was 250us (10000 points @ 40Mhz). The noise distribution of the ADC in both cases is checked by histograming the raw data.
Measurement 1 : Wideband Transformer & SMA Cable
disconnected from ADC inputs.
Measurement 2 : Wideband Transformer & SMA Cable
connected to ADC inputs.
Considering that the ADC Vdiff = 2V and LSB = 1.22e-4 V
- No Cable in ADC Inputs: Noise sigma ~ 0.13 mV
- Cable in ADC Inputs: Noise sigma ~ 0.23 mV
4. Linearity Check with F.Generator
4.1 Test 1: Using a ramp signal and fitting
A triangular waveform is introduced on each chanel independently with the F.Generator Tektronix AFG3252. The ramp signal from 0 to 16384 corresponds to an amplitude of 2V in the scpoe, only one of the rising ramps has been taken as reference to be fitted per channel. The residuals of the fit in ADC show the approximated linearity of the ADC, bellow is the residuals are show as relative deviation(top) and absolute counts (bottom).
On both plots there is a
dominance on the noise level Vs possible non linearities, therefore it is
not possible to distinguish from here any possible deviations from the desired ADC response.
Relative Linearity Error:
Absolute Linearity Error:
Topic revision: r6 - 2016-07-04
- unknown