RX parameters questions

Discussion in 'X2 & X4 SoC' started by Vince Tseng, Aug 3, 2016.

1. Hi,

I tried to determine the Radar Cross Section mathematically and verify it by experiment. Need you help to clarify some points in the X2 specification.

In section 2.5 RX parameters,

Q1: Input port match, at 3.1-10.6GHz range, S11= -7.5dB. Does it mean that I will encounter 10^(-7.5/20) = 42.16% signal lost?

Q2: Receiver gain = 9 dB. Is it the gain of LNA on Figure 4.1?

Q3: "Input RF voltage level at 1 dB gain compression point (peak-to-peak) = 160 mV." What does it mean? Is it the maximum peak-to-peak voltage at RFIN of Figure 4.1?

Q4: If I set DACmax = 8191, DACmin = 0, what is the voltage for a single step (000000000000 to 000000000001)?

Q5: What is the equivalent voltage of DACmax and DACmin at RFIN?

Q6: What is the DAC value of zero RFIN voltgae, is it 4096?

Q7: When I measured signal using RadarScope3, I noticed there was a low frequency carrying wave. Why is it there? How can I remove it? If by high pass filter, then what is the cut off frequency, 1/10 of sampling frequency?

Q7: In RadarScope3, there is a filtered view. How was it derived?

Thank you,

Vince Tseng

2. Olav LisethAdministratorStaff Member

Hi Vince,

Thanks for good questions!

A1: -7.5 dB is the single worst matching value for measured over the entire frequency range for the samples in the characterization pool. I.e, from 3.1-10-6GHz there are no values measured on the samples used for characterization which goes above -7.5dB and any given frequency point between 3.1-10.6GHz. A S11 of -7.5dB means that 17.8% ( 10^(-7.5/10) ) of the energy is reflected back to the source at this worst case sample at this worst case frequency. For the X2M200 frequency band, S11 is typically -10.1dB or 9.8% of the energy is reflected.

A2: The given gain is for the complete signal chain.

A3: Input referred 1dB compression point gives the signal swing at RFIN which gives a 1dB compression at the sampled signal.

A4: The DAC range is from 0 to 1.04 Volt, thus VLSB is 0.127 mV.

A5: Not applicable. The 1dB compression point is the interesting one, in addition the chip maximum ratings of course.

A6: Yes, it will vary, but be around 2^13/2.

A7: Yes, that’s an out of band oscillation. You should filter out irrelevant energy, e.g. band pass filter your signal according to the pulse generator setting.

A8: Several filters are available, but I guess you are referring to the adaptive clutter map. Here’s a python example:

Code:
def __init__(self):
self.cluttermap = None
self.w=0.01

def filter(self, frame):
if self.cluttermap==None:
self.cluttermap = frame.copy()

clutter = frame - self.cluttermap
self.cluttermap = (1-self.w)*self.cluttermap + self.w * frame
return clutter

3. Hi, Olav:

Thank you for clarification. A few more questions about the 9dB overall gain, there are three possible scenarios of the gain composition.

Scenario 1:

RFIN => S11 loss (~17.8% to 9.8%) + Amplifier + ADC
-------------------------------------------------------------
9dB gain

Scenario 2:

RFIN => S11 loss (~17.8% to 9.8%) + Amplifier + ADC
---------------------------------------------------
9dB gain

Scenario 3:

RFIN => S11 loss (~17.8% to 9.8%) + Amplifier + ADC
------------
9dB gain

For each scenario, the RFIN is

Scenario 1: max. RFIN = 1.04 / (9dB / gain of ADC)
Scenario 2: max. RFIN = 1.04 / 9dB
Scenario 3: max. RFIN = 1.04 / 9dB / (1-17.8% or 9.8%)

Which one is correct?

Among those scenarios, the minimum of max. RFIN = 1.04/2.8184 = 369mV.
369mV > 160mV (P1dB compression point), it means at this voltage (369mV), the gain shall be less than 9dB. If it is, what is maximum RFIN that will make ADC output 1FFF readout. Or the 9dB is overall gain which is irrelevant to gain compression.

Thank you and best regards,

Vince

4. Mats Risopatron KnutsenNew MemberStaff Member

Hi Vince,

Thanks for more good questions.

We're in compression with less signal swing that the full DAC range, i.e. the 9dB overall gain in irrelevant to gain compression.