An open source firmware for DSO-138 Oscilloscope for STM32F103C8T6 Blue Pill and SPI TFT display.
This is a modification to the DLO-138 for the so called Blue Pill STM32F103C8T6 board and SPI TFT 320x240 display. DSO-138 is an excellent piece of hardware based on ARM Cortex M3 core STM32F103 processor and sufficient for most beginner users. The stock firmware, while quite responsive, can use a few improvements. The main shortcoming which prompted the development of DLO-138 firmware is the inability to get waveform data into a computer for further analysis and the lack of a second channel. Engineers troubleshooting hardware issues need to mark reference points on waveform so having another analog or digital channel can greatly improve analysis. This firmware hopes to improve on these issues.
- Two analog channels
- Two digital logic channels (SWDIO and SWCLK pins (PA13 and PA14) on board)
- Serial port interface for captured waveform data
- Option to use rotary encoder instead of + - and SEL switches
- 2K sample depth
This firmware can NOT be used on stock DSO-138 hardware. We can build the hardware using stock STM32F103C8T6 board and a SPI TFT display.
You need not to have the DSO-138. Instead you can build the hardware on the breadboard or create original PCB. In the case of DLO-138, it is the loss of lowest timebase. Maximum sampling rate in DLO-138 is 20 µs/div instead of 10 µs/div. In the 20 µs/div range, firmware under-samples ADC channels, often reading same data twice. To use the second analog channel, analog front end has to be duplicated. This firmware can be used to provide two digital logic channels.
In the case of DLO-138-SPI, the frontend input circuit is simplified for ease of build. It can't accept negative voltage. Voltage range is now 5v/div to 10mV/div that is calculated from 12bit ADC value. AC mode is implemented with imaginary ground offset. Furthermore, a comparator for trigger detection have been omitted. Instead it uses the analog watchdog function of the ADC for pre-trigger capability. However the interrupt process takes cirtain amount of time at every crossing of the trigger level umtil actual trigger point. So in the scan range faster than 0.1ms/div, it looses some samples at every crossing of the trigger level. This is not the case after the trigger point.
The build environment uses Arduino IDE 1.8.19. For help with setting up IDE visit http://www.stm32duino.com
The board support is STM32F1xx/GD32F1xx boards by stm32duino version 2022.9.26
(additional URL: http://dan.drown.org/stm32duino/package_STM32duino_index.json )
CPU speed is 72MHz.
For graphics output, this project depends on the Adafruit_GFX_AS. However it is old so that it is necessary to add a line
#define textsize textsize_x
around line 62 of the Adafruit_GFX_AS.cpp of the STM32 library.
Circuit diagram is as follows:
Push button in encoder (SEL if using switches) moves focus to next parameter
Left/Right turn in encoder (+/- if using switches) changes the parameter which is in focus
Short press OK to HOLD the waveform and output it on serial port
At wave selection, + button rotates all the combination of A2 D1 D3. A1 is always active.
Long press OK button:
Focus Action
Trigger Level Zero the trigger level to Analog channel 1
Wave X scrollbar Beginning of waveform on screen (at trigger point)
Wave Y cursor Zero the cursor. If Analog CH1 coupling is GND, waveform reference base is set
Other Toggle on screen Analog CH1 statistics display
Press and hold OK button at power up to reset settings to default
DLO-138 - https://github.com/ardyesp/DLO-138
DSO-138 - http://www.jyetech.com/Products/LcdScope/e138.php
STM32Duino - http://www.stm32duino.com
STM32F103 - http://www.st.com/en/microcontrollers/stm32f103.html
Adafruit Graphics Library - https://github.com/adafruit/Adafruit-GFX-Library
Adafruit ILI9341 Arduino Library - https://github.com/adafruit/Adafruit_ILI9341