LMH6401_16 Datasheet, PDF (32/48 Pages) Texas Instruments – LMH6401 DC to 4.5 GHz, Fully-Differential, Digital Variable-Gain Amplifier

LMH6401
SBOS730A – APRIL 2015 – REVISED MAY 2015
www.ti.com
10.2.2.1.1 SNR Considerations
The signal-to-noise ratio (SNR) of the amplifier and filter can be calculated from the amplitude of the signal and
the bandwidth of the filter. The noise from the amplifier is band-limited by the filter with the equivalent brick-wall
filter bandwidth. The amplifier and filter noise can be calculated using Equation 3:
× SNR_AMP+FILTER = 10 log
V^2O
= 20 × log
e^2FILTEROUT
V_O
e_FILTEROUT
where:
• e_FILTEROUT = e_NAMPOUT • √ENB,
• e_NAMPOUT = the output noise density of the LMH6401 (30.4 nV/√Hz) at A_V = 26 dB,
• ENB = the brick-wall equivalent noise bandwidth of the filter, and
• V_O = the amplifier output signal.
(3)
For example, with a first-order (N = 1) band-pass or low-pass filter with a 1000-MHz cutoff, ENB is 1.57 • f_–3dB =
1.57 • 1000 MHz = 1570 MHz. For second-order (N = 2) filters, ENB is 1.22 • f_–3dB. When the filter order
increases, ENB approaches f_–3dB (N = 3 → ENB = 1.15 • f_–3dB; N = 4 → ENB = 1.13 • f_–3dB). Both V_O and
e_FILTEROUT are in RMS voltages. For example, with a 2-V_PP (0.707 V_RMS) output signal and a 1000-MHz first-
order, low-pass filter, the SNR of the amplifier and filter is 55.4 dB with e_FILTEROUT = 30.4 nV/√Hz • √1570 MHz =
1204.5 μV_RMS.
The SNR of the amplifier, filter, and ADC sum in RMS fashion, as shown in Equation 4 (SNR values in dB):
× SNR_SYSTEM = -20 log
-SNR_AMP+FILTER
-SNR_ADC
10 ¹⁰
+ 10 ¹⁰
(4)
This formula shows that if the SNR of the amplifier and filter equals the SNR of the ADC, the combined SNR is
3 dB lower (worse). Thus, for minimal degradation (< 1 dB) on the ADC SNR, the SNR of the amplifier and filter
must be ≥ 10 dB greater than the ADC SNR. The combined SNR calculated in this manner is usually accurate to
within ±1 dB of the actual implementation.
10.2.2.1.2 SFDR Considerations
The SFDR of the amplifier is usually set by the second- or third-harmonic distortion for single-tone inputs, and by
the second-order or third-order intermodulation distortion for two-tone inputs. Harmonics and second-order
intermodulation distortion can be filtered to some degree, but third-order intermodulation spurs cannot be filtered.
The ADC generates the same distortion products as the amplifier, but also generates additional spurs (not
harmonically related to the input signal) as a result of sampling and clock feed through.
When the spurs from the amplifier and filter are known, each individual spur can be directly added to the same
spur from the ADC, as shown in Equation 5, to estimate the combined spur (spur amplitudes in dBc):
-HDx_AMP+FILTER
-HDx_ADC
× HDx_SYSTEM = -20 log 10 ²⁰
+ 10 ²⁰
(5)
This calculation assumes the spurs are in phase, but usually provides a good estimate of the final combined
distortion.
For example, if the spur of the amplifier and filter equals the spur of the ADC, then the combined spur is 6 dB
higher. To minimize the amplifier contribution (< 1 dB) to the overall system distortion, the spur from the amplifier
and filter must be approximately 15 dB lower in amplitude than that of the converter. The combined spur
calculated in this manner is usually accurate to within ±6 dB of the actual implementation; however, higher
variations can be detected as a result of phase shift in the filter, especially in second-order harmonic
performance.
This worst-case spur calculation assumes that the amplifier and filter spur of interest is in phase with the
corresponding spur in the ADC, such that the two spur amplitudes can be added linearly. There are two phase-
shift mechanisms that cause the measured distortion performance of the amplifier-ADC chain to deviate from the
expected performance calculated using Equation 5; one is the common-mode phase shift and other is the
differential phase shift.
32
Submit Documentation Feedback
Product Folder Links: LMH6401
Copyright © 2015, Texas Instruments Incorporated