ADC measurement error study

Documents/2025-02-ADC-Study-HardwareSoftwareInterface.md

@@ -0,0 +1,85 @@
+# Adc-Study-HardwareSoftwareInterface
+
+## Table of Content
+
+1. [Adc Study, hardware software interface](#adc-study-hardwaresoftwareinterface)
+   1. [History](#reference-order)
+2. [Introduction](#introduction)
+    1. [Reference order](#reference-order)
+3. [Hardware Software Interface](#hardware-software-interface)
+   1. [ADC software interface](#hardware-software-interface)
+   2. [ADC software configuration](#adc-software-configuration)
+   3. [Data acquisition loop](#data-acquisition-loop)
+   4. [Data acquisition loop](#data-acquisition-loop)
+4. [Tests to be performed](#tests-to-be-performed)
+    1. [Values at max voltage](#values-at-max-voltage)
+    2. [Values at min voltage](#values-at-min-voltage)
+    3. [Valuees with a known resistor value](#values-with-a-known-resistor-value)
+
+## History
+| Rev | Date       | Author      | Description |
+|-----| --------   | ------------|  -----------|
+| 01  | 23/02/2025 | Ebersold A. | Creation, initial version|
+
+# Introduction
+
+The aim of this document is to study the delta temperature measurement between the real temperature
+and the Condictivity board temperature measures.
+
+## Reference order
+
+This document is referenced as **2025-02-ET-0001-Adc-Study-HardwareSoftwareInterface.md**. It was created to answer
+nehemis request regarding poor temperature measurement results on Conductivity board. The delta between the real temparture and the measure value of the board is 2C°.
+
+# Hardware software interface
+
+The eletronic part is depicted below. It shows the resitor devider and the 
+
+<img src="..\images\elecCTN.png" alt="drawing" width="500" height="200" />
+
+Fig 1.  Electronic block
+
+The TEMP pin is directly connected to pin PA0. Hence it must be configured as an analogue input.
+
+## ADC software configuration
+
+As of today, the ADC is configured to perfom single acquisitions. Hight precision measurement is expected by the ADC. Hence it shall use the 12Bits resolution.
+
+Remarque : There is a possibility to perform a tempature compensatation on VintRef value. This requires a two value aquisition for one measurement. TODO verify if this is required or not.  
+
+## Data acquisition loop
+
+Looking at the firmware, it appears that the temperature acquisition happens after each scheduler loop. A acquisition period should be defined.
+
+Motivation: The temperature changes slowely and it's not necessary to
+aquire a temperature every 10ms or even lesser.
+
+## Data conversion blocks
+
+The temperature computation goes through 3 basic blocks.
+1. The raw value acquisition from the ADC
+2. The conversion of the raw value into a Voltage
+3. The conversion of the voltage into a resistor value
+4. The conversion of the resistor into Kelvin
+
+    This block uses the [Steinhard-hard method](https://www.dxmht.com/fr/article/steinhart-hart-and-ntc-thermistor-equation.html) to convert the resistor value into a temperature.
+5. The conversion from Kelvin to Celsius
+
+
+# Tests to be performed
+
+In order to identify the root cause of the temperature measurement, I would suggest to create readable Open CAN disctionnary registers to be able to readout the ADC raw value, the computed Voltage.
+
+For each test below, compare the voltage computed by the board with the value measure with a voltmeter at TEMP.
+
+## Values at max voltage
+
+Connect CTN_A to 3.3V or VCC and read out the raw value and the voltage.
+
+## Values at min voltage
+
+Connect CTN_A to ground and read out the raw value and the voltage.
+
+## Values with a known Resistor value
+
+Connect a know resistor between CTN_A and CTN_B and read out the raw value and the voltage.