Single Measurement¶

Main Menu¶

Assuming all devices are connected, a two-port system is modeled as follows:

../_images/setup_type_ltf.png — A 2-Port System with Unknown DUT.¶

The Main Menu summarizes the test-bench as shown below:

../_images/main_menu_view.png — The 2-port System Dashboard Representation¶

The Main Menu consists of two-primary elements:

The Sidebar

The Instrument Manager

Sidebar¶

The Sidebar is a multi-tab window that can be customized based on the desired application. The Main Menu defaults to the Main tab.

Start by configuring the global System Settings in the sidebar by specifying:

Frequency Sweep

Settings().f_0: The fundamental carrier frequency.

Settings().num_harmonics: The number of harmonics frequencies.

Time Sweep

Settings().t_{stop}: the measurement period.

Settings().t_{step}: the measurement time-step.

Trigger

Settings().trigger_device: the trigger output device type.

Settings().trigger_port: the trigger output device port.

Datagroup

Settings().datagroup: The name of the database file (default Single.h5).

Settings().trigger_port: The name of a node in a database file (default Default).

You can specify additional global variables:

In the sknrf.yml config file before runtime.

In the Settings Menu during runtime.

Instrument Manager¶

The Instrument Manager consists of:

Settings().num_ports + 1 ports

Port 0 is a reserved calibration reference port.

Settings().num_duts DUTs

The aux devices consisting of non-compatible instruments:

Power Meter

Spectrum Analyzer

Vector Network Analyzer

Ports¶

The number of ports is configured in sknrf.yml and is available at runtime in Settings().num_ports. Each measurement port is represented by a column in the Instrument Manager.

A Port PortModel a collection of devices that can be connected to a measurement port:

port.lfsource, a low-frequency signal source that sets $v(t, f)$ .

port.lfreceiver, a low-frequency signal receiver that gets $v(t, f)$ and $i(t, f)$ .

port.lfztuner, a low-frequency impedance controller that sets and gets $z_p(t, f)$

port.rfsource, a high-frequency signal source that sets $a_p(t, f)$ .

port.rfreceiver, a high-frequency signal receiver that gets $a_p(t, f)$ and $b_p(t, f)$ .

port.rfztuner, a high-frequency impedance controller that sets and gets $\gamma_p(t, f)$ .

Each measurement port provides a Thevenin Equivalent at Low-Frequency (LF) and High-Frequency (HF). Thus the following equations fully describe the inputs and outputs of a port:

VIZ->BAG

$\begin{eqnarray} B_p & = & \frac{1}{2 \sqrt{ \Re Z_p }} \left( V - Z_p^*I \right) \\ A_p & = & \frac{1}{2 \sqrt{ \Re Z_p }} \left( V + Z_pI \right) \\ \Gamma_p & = & \frac{Z_p - Z_0}{Z_p + Z_0} \\ \end{eqnarray}$

BAG->VIZ

$\begin{eqnarray} V & = & \frac{1}{\sqrt{ \Re Z_p }} * \left( Z_p^*a + Z_pb \right) \\ I & = & \frac{1}{\sqrt{ \Re Z_p }} * \left( a - b \right) \\ Z_p & = & z0 \frac{1 + \Gamma_p}{-\Gamma_p + 1} \\ \end{eqnarray}$ where,

$\begin{eqnarray} V & = & \mathcal{F} \left( v(t, f) \right) \text{w.r.t } t\\ I & = & \mathcal{F} \left( i(t, f) \right) \text{w.r.t } t\\ Z_p & = & \mathcal{F} \left( z_p(t, f) \right) \text{w.r.t } t\\ B_p & = & \mathcal{F} \left( v(t, f) \right) \text{w.r.t } t\\ A_p & = & \mathcal{F} \left( i(t, f) \right) \text{w.r.t } t\\ \Gamma_p & = & \mathcal{F} \left( \gamma_p(t, f) \right) \text{w.r.t } t\\ \end{eqnarray}$

These raw waveforms have the following meaning:

$v(t, f)$ : The input voltage (default $0.0$ ).

$i(t, f)$ : The output current (default $0.0$ ).

$z(t, f)$ : The port termination impedance (default $z_0 = 50.0$ )

$b_p(t, f)$ : The output (reflected) power-wave (default $0.0$ ).

$a_p(t, f)$ : The input (incident) power-wave (default $0.0$ ).

$\gamma_p(t, f)$ : The port reflection coefficient (default $0.0$ ).

Since each of these waveforms has a default value, we need only connect an instrument when we know that the DUT does not meet these assumptions:

Devices with direct access to power and ground do not require a port.lfztuner.

Devices matched to 50 Ohm do not require a port.rfztuner.

Devices matched to 50 Ohm do not require a port.rfreceiver that can measure $a_p(t, f)$ .

LF circuits do not require port.rfsource, port.rfreceiver, port.rfztuner.

RF circuits do not require port.lfsource, port.lfreceiver, port.lfztuner.

Supported Instruments¶

`LF Source`

DC Supply

AWG

DAC

`LF Receiver`

Multimeter

Oscilloscope

ADC

`LF ZTuner`

Open/Short Circuit

Varactor

Active Load

`RF Source`

Signal Generator

VSG

AWG

Pulse Generator

SRD

`RF Receiver`

VNA

VSA

Power-Meter

Oscilloscope

Sampling Scope

`RF ZTuner`

50 Ohms

Passive Load

Active Load

The choice of port.rfreceiver determines the trade-off between:

speed vs. accuracy

digital vs. analog

affordable vs. expensive

This architecture manage the trade-off between digital and analog verification:

Oscilloscope: Low-Cost, digitally sampled receiver in an embedded-device.

(Sub)-Sampling Oscilloscope: Mixed-Signal with a slow digital sample-rate and fast analog sample-rate.

Network Analyzer: High-Cost, high accuracy measurement for analog circuits in an R&D lab.

Auxiliary Port¶

The following instruments (found in most labs) provide turn-key design verification, but result in significant challenges due to low inter-operability. These instruments (and suggested alternatives) are:

Power Meter (Anything Else)

Calibrated to NIST standards, but are limited to scalar measurements.

Requires manual range adjustment to achieve accuracy

Spectrum Analyzer (Signal Analyzer):

Scalar frequency swept measurements do not translate to the envelope domain.

A Signal Analyzer is a good alternative.

3-Receiver VNA (4-Receiver VNA):

3-Receiver VNAs use 12-Term error correction to extract S-Parameter models.

4-Receiver VNAs use 8-Term (or 12-term error correction) to measrure reflectometry (o extract S-Parameter models).

Since these instruments are still common, the Aux Port provides a way to log these measurements: