Skip to content

Instantly share code, notes, and snippets.

Last active August 20, 2023 07:43
  • Star 44 You must be signed in to star a gist
  • Fork 11 You must be signed in to fork a gist
Star You must be signed in to star a gist
What would you like to do?
Arduino Due: ADC → DMA → USB @ 1MSPS
// Arduino Due ADC->DMA->USB 1MSPS
// by stimmer
// from
// Input: Analog in A0
// Output: Raw stream of uint16_t in range 0-4095 on Native USB Serial/ACM
// on linux, to stop the OS cooking your data:
// stty -F /dev/ttyACM0 raw -iexten -echo -echoe -echok -echoctl -echoke -onlcr
volatile int bufn,obufn;
uint16_t buf[4][256]; // 4 buffers of 256 readings
void ADC_Handler(){ // move DMA pointers to next buffer
int f=ADC->ADC_ISR;
if (f&(1<<27)){
void setup(){
adc_init(ADC, SystemCoreClock, ADC_FREQ_MAX, ADC_STARTUP_FAST);
ADC->ADC_MR |=0x80; // free running
ADC->ADC_RPR=(uint32_t)buf[0]; // DMA buffer
ADC->ADC_RNPR=(uint32_t)buf[1]; // next DMA buffer
void loop(){
while(obufn==bufn); // wait for buffer to be full
SerialUSB.write((uint8_t *)buf[obufn],512); // send it - 512 bytes = 256 uint16_t
#!/usr/bin/env python
# from
import pyqtgraph as pg
import time, threading, sys
import serial
import numpy as np
class SerialReader(threading.Thread):
""" Defines a thread for reading and buffering serial data.
By default, about 5MSamples are stored in the buffer.
Data can be retrieved from the buffer by calling get(N)"""
def __init__(self, port, chunkSize=1024, chunks=5000):
# circular buffer for storing serial data until it is
# fetched by the GUI
self.buffer = np.zeros(chunks*chunkSize, dtype=np.uint16)
self.chunks = chunks # number of chunks to store in the buffer
self.chunkSize = chunkSize # size of a single chunk (items, not bytes)
self.ptr = 0 # pointer to most (recently collected buffer index) + 1
self.port = port # serial port handle
self.sps = 0.0 # holds the average sample acquisition rate
self.exitFlag = False
self.exitMutex = threading.Lock()
self.dataMutex = threading.Lock()
def run(self):
exitMutex = self.exitMutex
dataMutex = self.dataMutex
buffer = self.buffer
port = self.port
count = 0
sps = None
lastUpdate = pg.ptime.time()
while True:
# see whether an exit was requested
with exitMutex:
if self.exitFlag:
# read one full chunk from the serial port
data =*2)
# convert data to 16bit int numpy array
data = np.fromstring(data, dtype=np.uint16)
# keep track of the acquisition rate in samples-per-second
count += self.chunkSize
now = pg.ptime.time()
dt = now-lastUpdate
if dt > 1.0:
# sps is an exponential average of the running sample rate measurement
if sps is None:
sps = count / dt
sps = sps * 0.9 + (count / dt) * 0.1
count = 0
lastUpdate = now
# write the new chunk into the circular buffer
# and update the buffer pointer
with dataMutex:
buffer[self.ptr:self.ptr+self.chunkSize] = data
self.ptr = (self.ptr + self.chunkSize) % buffer.shape[0]
if sps is not None:
self.sps = sps
def get(self, num, downsample=1):
""" Return a tuple (time_values, voltage_values, rate)
- voltage_values will contain the *num* most recently-collected samples
as a 32bit float array.
- time_values assumes samples are collected at 1MS/s
- rate is the running average sample rate.
If *downsample* is > 1, then the number of values returned will be
reduced by averaging that number of consecutive samples together. In
this case, the voltage array will be returned as 32bit float.
with self.dataMutex: # lock the buffer and copy the requested data out
ptr = self.ptr
if ptr-num < 0:
data = np.empty(num, dtype=np.uint16)
data[:num-ptr] = self.buffer[ptr-num:]
data[num-ptr:] = self.buffer[:ptr]
data = self.buffer[self.ptr-num:self.ptr].copy()
rate = self.sps
# Convert array to float and rescale to voltage.
# Assume 3.3V / 12bits
# (we need calibration data to do a better job on this)
data = data.astype(np.float32) * (3.3 / 2**12)
if downsample > 1: # if downsampling is requested, average N samples together
data = data.reshape(num/downsample,downsample).mean(axis=1)
num = data.shape[0]
return np.linspace(0, (num-1)*1e-6*downsample, num), data, rate
return np.linspace(0, (num-1)*1e-6, num), data, rate
def exit(self):
""" Instruct the serial thread to exit."""
with self.exitMutex:
self.exitFlag = True
# Get handle to serial port
# (your port string may vary; windows users need 'COMn')
s = serial.Serial('/dev/ttyACM0')
# Create the GUI
app = pg.mkQApp()
plt = pg.plot()
plt.setLabels(left=('ADC Signal', 'V'), bottom=('Time', 's'))
plt.setYRange(0.0, 3.3)
# Create thread to read and buffer serial data.
thread = SerialReader(s)
# Calling update() will request a copy of the most recently-acquired
# samples and plot them.
def update():
global plt, thread
t,v,r = thread.get(1000*1024, downsample=100)
plt.plot(t, v, clear=True)
plt.setTitle('Sample Rate: %0.2f'%r)
if not plt.isVisible():
# Set up a timer with 0 interval so Qt will call update()
# as rapidly as it can handle.
timer = pg.QtCore.QTimer()
# Start Qt event loop.
if sys.flags.interactive == 0:
Copy link

hamed commented Sep 7, 2013

This sketch helped me a lot, I use it as a base to implement the recording part of my white noise spectrum analyzer.

Copy link

This code is excellent! I am experiencing one issue though.
About 7 min 40 sec into the code running, it just stops. I ran it 3 times and came up with the exact same time. I have not touched the computer, no errors are displayed, there is not a time limit in the code that I can see. In order to get the program to run again, I need to press the reset button on the Due and rebuild the python script.

What would cause the code to all of a sudden stop working? Is there a limit to the number of loop iterations in the Due? Is there a limit to the number of iterations in python? Is this an OS issue? (I am using Windows 10, 64-bit)

I have tried removing all code pertaining to exiting in the Python code, yet the issue persists.

Any advice is appreciated.

Copy link

It seems there is a bug in arduino part. See here for the explanation and a fix:

Copy link

ghost commented Dec 8, 2016

Nice job! I runt it, work very well. But I have a big problem because I am new with arduino DUE and the registers. I want to make 4 channels acquisition and It was very difficult for me. Do you have an exemple, or give me a direction to go. Thank's . Patrick
And I want to keep the plot of the 4 channels in Python, of course.

Copy link

Hi, very nice! I'm trying to make 2 channels acquisition but I don't fully understand your code. Could you explain it a little bit so I can modify to use 2 channels instead of one?

Copy link

I think there is an error in ADC_Handler where the buffer pointers are updated. By the time the handler is called, the DMA is already filling the "next" buffer = bufn+1. So not only must bufn be incremented, but the "next" buffer should then be bufn+2. I think the updates work better like this:

void ADC_Handler(){ // move DMA pointers to next buffer
int f=ADC->ADC_ISR;
if (f&(1<<27)){

Copy link

hoomanj commented Sep 14, 2017

Hi, Awesome code.

can you help me write a python code that will just simply write the data in a text file. so a text file containing time values,voltage,rate

no need for graphing etc. just receive the data from the Due and store it in a text file.


Copy link

Im trying to run the python code on my pc using python3.5.4. and I get an error:
Exception: PyQtGraph requires one of PyQt4, PyQt5 or PySide; none of these packages could be imported.
I have already installed PyQtGraph so which one of these others mentioned above should I install with my version of python?

Copy link

trycage commented May 12, 2018

Hi all

In order to get the program run (I did amend the buffer increment in the ADC_Handler ), I need to open a serial when I use the native port.

More specifically, every time I close the serial from the PC side the Arduino DUE stops, I use a led blinking as a check. My led blinking is related to the actual transfer, hence I guess the process just hangs at the SerialUSB.write.
There are other posts in which users noted issue when a serial transfer is initiated on the native port without an actual connection. The weird thing is that is I re-open the port on the PC side the Arduino actually continues from the point it was interrupted.



Copy link

I see that you use SeriaUSB together with Python. Did you do anything particular? In my case, when using the Native port, nothing is getting to the PC... (I use Linux, in case it matters)

Copy link

Hi, I want use the Arduino due or teensy 4.0 like SPI-USB convwersion and I need 1Msps, I want use TCL or c++ in linux but I cant get more the 7ksps, with Arduino Serial console I get 1Msps but I want a TCL script,
what I doing wrong?
Can be something like Baudrate? Cant use more than 250000 in linux
What i can do?

Copy link

pklaus commented Dec 3, 2019

Hi @emanavas,

Hi, I want use the Arduino due or teensy 4.0 like SPI-USB convwersion and I need 1Msps, I want use TCL or c++ in linux but I cant get more the 7ksps, with Arduino Serial console I get 1Msps but I want a TCL script,
what I doing wrong?

First of all, could you confirm you're connecting to the Native USB Port, not the Programming USB Port? Are you running the sketch as in the original Gist on top of this page or a modified version?

Can be something like Baudrate? Cant use more than 250000 in linux

Baudrate only matters if there's a physical UART connection in your transmission line. (Such as in the case of the "Programming USB Port" where the transmission takes this route: PC <- USB -> ATmega16U2 <- UART (LV-TTL) -> Arduino Due)

What i can do?

Show us your minimal code example of your TCL / C++ code so that we can help further.


Copy link

Hi pklaus,
Finally I Decided use a C++ and I get around 3MB/s Using printng a FIX value from TEENSY 4.0, My goal its attach a external ADC 1-3Msps to teensy and send to PC.
ABout a DUE, I used a Native USB and I now I know this Baudrate its a virtual and depends of the driver og the host (Linux/PC).
Basically always work by the USB 2.0 (480Mbps) because we dont have a FTDI.

Copy link

I use this code, it works fine
But the result shows there is only 666kHzSPS, not 1MSPS.
Do you give me a direction to go?

Copy link

aleksas commented Sep 22, 2020

Minor correction: it should be (3.3/(2**12 - 1)) instead of (3.3/2**12) on this line.

Copy link

I use this code, it works fine
But the result shows there is only 666kHzSPS, not 1MSPS.
Do you give me a direction to go?

ADC_FREQ_MAX is defined as 20000000. The datasheet of the SAM3X specifies a max ADC frequency of 22MHz.
Use the following call to adc_init to get full 1MSPS:

adc_init(ADC, SystemCoreClock, 22000000, ADC_STARTUP_FAST);

Copy link

This is great code! Is there any possibility to add more than one channel to the ADC? I am interested in using up to 5 analog inputs

Copy link

Hello, may I ask why after executing app.exec_()
will keep running interactively
Traceback (most recent call last):
t,v,r = thread.get(1000*1024, downsample=100)
data = data.reshape(num/downsample,downsample).mean(axis=1)
TypeError: 'float' object cannot be interpreted as an integer

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment