Analog circuits

A project log for Microcontroller based curve tracer

Electronic Curve tracer for measure and compare components on circuit boards

SUFSUF 08/10/2018 at 13:250 Comments

I'm heavily working on the next version of the curve tracer. Today I almost finished the schematics of analog modules of the device. All of drawings are uploaded into the files section. Actually everything reside in a private repo on the GitLab. This will be publicly opened when I order the first revision PCBs. Until this time the whole circuit is so shaky that I don't want anyone to build it, because the design will probably heavily changed and there are incompatibilities exists between the schematic an PCB designs.

About the circuitts

Power supply module:

The power supply use a enclosure mounted toroid transformer with two 14V secondary windings (probably a 100VA model). The circuit itself provide +-15V from linear regulators mainly for the analog circuits and the power amplifier, a +5V and +3.3V from buck converters for the logic circuits (plus MCU, Display, etc.). And a four channel buffered rail splitter (virtual ground) used for the level conversion of the ADC and DAC (probably two will be used, but it is easy to build four from a 4 channel opamp)

Power Amplifier module:

It converts the 0.65V-2.65V swing sine wave from the MCUs DAC module to +-10V (20Vpp) gound referenced load capable signal.

In the input I put a unity gain differential amplifier. This will convert the input to -+1V by substracting the 1.65V reference DC from the power supply.

The next stage is a 1.28x gain non-inverting amplifier creating 2.56Vpp signal. This signal attenuated by a 256 step digipot. This give us 10mV step control of the power amplifier. The power amp has a 10x gain. So the result: we can control the output voltage of the amp 0-20V in 100mV steps.

Attenuator module:

This module control the current limiting resistors on the output of the Power Amplifier. As requirement we can't use conventional mechanical relays mostly because of the noise it makes. Therefore I choose three different kind of solutions to set the current limiting resistor value (we need AC capable devices):

1. Low current version: ADG451 (or ADG441) and ADG1411 analog switches (capable around 100mA)

2. Medium current version: Vishay VO14642 photovoltaic AC MOSFET SSR (capable 1A AC)

3. Load switching: Si8751 isolated MOSFET drivers + dual MOSFETs (capable 4.5A AC)

Also this circuit provide the TVS protected output for the DUT and the measurement point for the output current and voltage. The output current measured across the whole resistor network and not on an additional current sense resistor. This solution probably generate less noise, but on the other side more sensitive to the temperature drift of the resistor (we will see, how this works in practice)

Measurement Amplifier:

This is a two channel amplifier one for the current, and one for the voltage. The input signals are buffered with two unity gain amplifiers. The voltage signal is coming from the voltage across the DUT and amplified by a "PGA" capable of creating x0.1, x1, x10, x100 and x1000 gain respectively. The current signal is comming from the voltage across the current limiting resistor. As this is not ground referenced I used a differential amplifier configuration with x0.1, x1 and x10 gain.

The outputs of this circuit can be directly connected to the oscilloscope X and Y inputs and will be also connected to the MCUs 12bit ADC through a level shifter (not implemented on this board) 

This is all for now. More to come.

If you have any question, suggestion on this design, please let me know in the comments