H750-based oscilloscope

Implementation of Signal Conditioning Circuit:
We will first briefly introduce the analog input and output channels on the AFE03 board, and then explain the design and calculation methods in detail later.
Analog Input Channel Introduction:
This includes signal conditioning implemented with resistor voltage dividers and operational amplifiers, and a square wave output implemented with a comparator (for triggering and frequency measurement).
INA, INB: The input terminals of the oscilloscope. The pocket instrument sends an analog signal to this point. Here, a 1MΩ input impedance is achieved through a series resistor voltage divider, generating two signals for selection: one is a direct input, and the other is attenuated to 1/20.
gain: A selection switch that selects either the direct signal or the signal attenuated to 1/20 to enter the first-stage non-inverting amplifier. The non-inverting amplifier performs two functions: first, it amplifies the input signal at the non-inverting terminal by a factor of two; second, it shifts the amplified signal by 1.65V, calculated as Vo = 1.65 + 2*Vi. Therefore, the overall gain of the corresponding circuit is either 2 times or 1/10 times.
AnalogA and AnalogB: The analog signals, amplified and shifted by the inverting amplifier, are connected to the STM32H750 development board and enter the H750's ADC.
TrigerA and TrigerB: The square wave signal generated by AnalogA, AnalogB, and the DC reference level (generated by one of the H750's DACs) after passing through a comparator enters the STM32H750's timer for frequency measurement.
DAC_OUT2: The DC reference level is output through the STM32H750's internal DAC2 configuration.
 
Analog output channel description:
Includes signal conditioning implemented by resistor dividers and operational amplifiers, and a square wave output implemented by a comparator (for triggering and frequency measurement).
The STM32H750's DAC1 output has a waveform range of 0-3.3V.
A second-order RC filter implements a low-pass filter function.
A resistor divider and buffer convert the 5V input to a low-impedance 2V output, which is then amplified by -5 times for shifting the output signal.
The output amplifier performs two functions: first, it amplifies the non-inverting input by 6 times; second, it shifts the amplified signal by -10V before outputting it. The calculation formula is Vo = -10 + 6 * Vi.
The voltage divider network is used to achieve good results when outputting small signals by utilizing analog circuit voltage division.
 
 
Below, we will introduce the calculation of the analog circuit in detail:
Analog Input Channel Calculation:
This section requires some basic knowledge of operational amplifiers.
First, we need to understand that when the VREF of the STM32H750 is powered by 3.3V, the input range of the STM32's ADC is 0-3.3V, while our maximum input signal is ±15V. Therefore, we need to solve the problem of large signal input not saturating. Let's solve an equation:
15 * a + b = 3.3
- 15 * a + b = 0,
which gives a = 0.11 and b = 1.65.
This means we need to attenuate the input signal to at least 0.11 times (approximately 1/9), and then add 1.65V DC to meet the full-scale input of the ADC. Therefore, we use the following operational amplifier circuit. The resistor network achieves 1/20 attenuation, and the operational amplifier performs a 2x amplification and a 1.65V voltage shift, thus achieving 0.1INB + 1.65V.
We can use the superposition theorem to analyze this circuit. First, when analyzing the contribution of the input signal INB to the output Vo, we ground the other voltage source in the circuit, -1.65V. This way, the input signal is divided to 1/20 after passing through R14 and R18, and then amplified by a factor of 2 by the non-inverting amplifier circuit. The overall gain of the input signal is 1/10. When analyzing the contribution of -1.65V to the output, we ground the input signal AIN, and the amplification factor of -1.65V is -1. Therefore, we obtain the output Vo = -1.65V * (-1) + AIN/10 = 1.65V + AIN/10.
 
The -1.65V above is obtained from -12V through a resistor divider and buffer.
This circuit solves the problem of matching the ±15V input to the ADC's 0-3.3V input range. We also need to consider accurate sampling even with small input signals. For example, a 10mV signal will attenuate to 1mV after passing through this circuit. To maximize the signal-to-noise ratio of the input signal, we add a switching mode to the analog front-end. When acquiring small signals, the switch selects the direct input of INB to the non-inverting input of the op-amp, instead of selecting the attenuated signal from INB. This ensures the signal entering the ADC is as large as possible. Combined with a 16-bit ADC, this ensures accurate and reliable sampling results.
As shown in the diagram, we add a signal switch (relay or manual switch) after the 1M ohm input voltage divider resistor to select whether INB enters the op-amp's non-inverting input directly or after being divided by 1/20. Both methods result in a 1M ohm input impedance for INB. When we need to acquire small signals, we can toggle the switch to use the direct input for more accurate measurement results.
We can calculate that when the direct input is selected, Vo = 2AIN + 1.65, and when the attenuation input is selected, Vo = AIN/10 + 1.65.
 
Analog output channel calculation:
When the VREF of the STM32H750 is powered by 3.3V, the output range of the internal DAC is 0-3.3V. To achieve the ±10V output required in the problem, we need to solve the following equations:
0*a + b = -10V
3.3*a + b =
Solving for 10V , we get a=6.06 and b=-10
, which allows us to design the following circuit:
In the diagram above, the 0-3.3V signal output from the STM32H750's internal DAC is filtered by a low-pass filter and then input to the non-inverting input of the TL082, forming a non-inverting amplifier with a gain of 6. The amplified waveform is 0-19.8V. Then, utilizing the -5x amplification capability of the TL082's inverting amplifier section, the +2V obtained from the 5V voltage divider is amplified by -5x to obtain -10V. This -10V is then superimposed on the 0-19.8V signal output from the non-inverting amplifier to obtain an output of approximately ±10V. The calculation formula is: Vout = 6*Vin -10.
Similar to an ADC, to ensure the signal source output covers ±10mV to ±10V, while simultaneously balancing a large signal range and small signal accuracy, the DAC's resolution is a crucial factor. The H750's DAC is 12-bit, and its full-scale output (using all 4096 code values) is ±10V. When outputting small signals by reducing the DAC code value, to achieve a 7-bit voltage resolution (128 vertical points), the waveform must be attenuated by 128/4096 = 1/32. This translates to an output voltage range of ±10V/32 = ±0.3215V. For signals smaller than ±0.3125V, further reducing the code value results in insufficient DAC resolution and noticeable waveform steps. Therefore, we use analog voltage division. When outputting signals smaller than ±0.3125V, we use a switching resistor divider to attenuate the waveform by 1/20, ensuring sufficient voltage resolution for small signals. Simultaneously, the combination of R57 and R62 ensures a 50Ω output resistance at 1/20 attenuation, and R5 ensures a 50Ω output resistance at x1.
 
 
Comparator circuit:
To implement triggering and frequency counting functions, we designed two comparator channels on the board. These channels convert the waveforms of the two analog input channels before they enter the ADC into square wave signals for use as timer inputs in the H750. The reason for using the waveforms before they enter the ADC is that the waveforms entering the ADC, after conditioning by the front-end analog circuitry, fall within the known 0-3.3V range, making the comparator's comparison threshold easier to design.
As shown in the diagram above, the H750 uses its internal DAC2 to output a 0-3.3V DC signal to compare with the waveform of channel 2 before it enters the ADC, converting the channel 2 waveform into a square wave. This allows the H750's timer function to use the square wave signal for interrupt handling and timer capture.
 
A video demonstration is shown below.