This guide will show you how to use an FT232H to connect to I2C and SPI sensors and breakouts from your desktop PC running Windows, Mac OSX, or Linux. The FT232H also allows for general purpose digital input and output (GPIO) for things like buttons and LEDs.

The cool part about this is that you can then use any of the CircuitPython Libraries that have been written for the numerous sensors and breakouts.

CircuitPython and CircuitPython Libraries

As you are going through this guide, keep in mind the difference between CircuitPython and CircuitPython Libraries:

• CircuitPython - a microprocessor specific implementation of Python written in C. Here's the source code. And here's the main CircuitPython guide.
• CircuitPython Libraries - sensor and breakout specific code written in Python using the CircuitPython hardware API. There are a lot of these - check out the bundle.

There are various hardware combinations that allow for running CircuitPython and CircuitPython Libraries. In this guide we will not be using the actual CircuitPython firmware. But we will be using CircuitPython Libraries. To explain this further, let's go over the main hardware platforms and explain where the FT232H fits in to all this.

CircuitPython on Microcontrollers

This is the most straight forward setup.

If you are using a microcontroller that can run CircuitPython, then you will most likely have the low level hardware interfaces needed to access the many sensors and other breakout boards - I2C, SPI, and GPIO. Even better, there is most likely a CircuitPython library written for the sensor.

In this case, you simply install CircuitPython, add the libraries, and then follow whatever guide goes along with your sensor.

CircuitPython on Single Board Computers

This setup requires a special shim library called Blinka.

Single Board Computers (SBC), like the Raspberry Pi and Beagle, also typically have I2C, SPI, and GPIO interfaces available. These boards are also powerful enough to run complete operating systems, like Linux. They can't run CircuitPython directly, but generally don't need to. They can run the much larger Python implementations, like CPython. To allow use of SBCs running Python to use CircuitPython libraries to access sensors over I2C/SPI/GPIO, the Blinka library was created.

In this case, you pip install Blinka, pip install libraries, and then follow whatever guide goes along with your sensor.

CircuitPython on Personal Computers

This is where the FT232H comes in. Here's why.

So what about your super powerful desktop or laptop PC? They can most definitely run Python. But can they also use CircuitPython libraries and talk to I2C/SPI sensors? Can they blink LEDs? Read buttons? Generally, no. As powerful as your Windows, Mac, or Linux PC is, it most likely does not have those low level hardware interfaces. Look on the back of your computer. Do you see an I2C port? A SPI port? A cluster of pins labeled GPIO? Nope.

So what can we do? Well, look again at the back of your PC. See any USB ports? Most likely there are several. Heck, there are probably even several USB ports on the front of your PC! Can we use USB? Yes, thanks to a specialized USB bridge chip made by FTDI - the FT232H.

This will allow us to do something like this:

With FT232H support added to Blinka, we can follow a similar approach as with the SBCs. The FT232H attached to the USB port acts as our surrogate set of low level hardware interfaces.

Neat! Let's see how we can get this all setup and working.

The support for the FT232H in Blinka utilizes the pyftdi library by eblot. This in turn relies on a few other things, like libusb. So before we can actually use the FT232H, we need to get everything setup. See the OS specific sections for what we went through to get things working for each.

Additional Information

Just for reference, here's the install info from the pyftdi library itself:

But first try the install instructions on the pages that follow for your OS.

Getting this all set up on Windows is not fun - but it is possible. Follow each step below to get it working on Windows

Have Python 3 Installed

We assume you already have Python 3 installed on your computer. Note we do not support Python 2 - it's deprecated and no longer supported!

At your command line prompt of choice, check your Python version with python --version

Plug in FT232H & Fix Driver with Zadig

Unlike Mac or Linux, Windows doesn't like devices that are 'driverless' - every hardware device must have a driver attached to it. To fix the driver on Windows, we must use a tool called Zadig

Download the appropriate Zadig tool for your version of Windows at its homepage.  This tool simplifies the installation of a libusb driver for the FT232H device.

Before you run the tool unplug all FTDI devices from your computer.  This includes devices like Arduinos which include an FTDI chip as a USB to serial converter.  You want to unplug these devices to make sure you don't accidentally select one with the tool and replace its driver.

After all the FTDI devices are unplugged, plug in your FT232H breakout to the computer so it is the only FTDI device connected to the computer.

Now run the Zadig tool executable you just downloaded (there is no installation necessary, the executable is the program).  Click the Options menu and select the List All Devices item below:

The list box of devices should populate with many devices.  Select the USB Serial Converter device shown below.

Note: Make sure to select the device with Driver equal to FTDIBUS and USB ID equal to 0403 6014!  If you pick the wrong device you might accidentally uninstall another device's driver and make it inoperable.

Click the up/down arrows on the driver select box to the right of the green arrow and select the libusbK driver as shown above.

Now click the Replace Driver button to replace the FTDI driver with the libusbK-based driver.  After the driver replacement finishes you can close Zadig tool.

To check that the driver was successfully replaced, open Device Manager from Control Panel or searching in the Start menu.  You should see a new top level node libusbK devices and the USB Serial Converter underneath it as shown below.

If you see the libusb-win32 node and USB serial device, move on to the next step to install libftdi.

If you don't see the libusb-win32 node, try unplugging and plugging back in the FT232H breakout.  If you still don't see the node, run Zadig tool again and follow the steps above again to make sure you replace the FTDI driver for the device with the libusb-win32 driver.

Option 1 - Install pyftdi and pyusb

As a first attempt, try simply pip installing pyftdi and pyusb.

pip3 install pyusb

pip3 install pyftdi

And then move to the Test pyusb and pyftdi section below. If the install runs into issues or doesn't seem to work, you may need to try Option 2. This is what it originally took to set things up on Windows. However, it has been reported that the latest version of pyusb "just works", so you might get lucky :)

Option 2 - Install pyftdi and (fixed) pyusb

Next lets install pyusb and pyftdi. We have to do a little trickery here so before you start run

pip3 uninstall pyusb

pip3 uninstall pyftdi

(or pip instead of pip3 if that's how you have it named)

To make sure you do not have pyusb and pyftdi installed

Run them again to make absolutely sure!

We need to get a fork of pyusb to fix a bug that affects windows:

git clone https://github.com/minkustree/pyusb.git

Then

cd pyusb

python setup.py install

if you get a permission denied error, try

python setup.py install --user

Now you can run pip install pyftdi

Test pyusb and pyftdi

Now that you have pyusb and pyftdi installed correctly, run python and paste in the following (with the FT232H plugged in)

import usb
import usb.util
dev = usb.core.find(idVendor=0x0403, idProduct=0x6014)
print(dev)

You should get something like the following, not that dev is None or any other weird failure

If you get usb.core.NoBackendError: No backend available

Download libusb 1.0.23, uncompress it with WinRar or some other decompression tool that can open 7z files.

If you are on 64-bit Windows, copy the MS64\dll\libusb-1.0.dll file into C:\Windows\System32 and C:\Windows\SysWOW64I

If you are on 32-bit Windows, copy the MS32\dll\libusb-1.0.dll file into C:\Windows\System32

Install Adafruit Blinka

Run pip install adafruit-blinka

Set Environment Variable

You must do this every time before running circuitpython code, you can set it permanently in windows if you like, for now just type into the same cmd window you're using with Python

set BLINKA_FT232H=1

If you are using Windows Powershell, the syntax is a little different. In that case do:

$env:BLINKA_FT232H=1

Check Platform was detected

In the same command window you set BLINKA_FT232H=1 env var, run python and run

import board
dir(board)

at the Python REPL. If you get no errors, and you see a list of all the pins available - you're good to go!

We assume you already have Python 3 and brew available on your Mac. Thankfully, setup on MacOS X is not so bad!

Note: If you are running VMWare Fusion on MacOS, then you can also try the Windows install process.

Install libusb

Start by installing libusb with

brew install libusb

Install pyftdi and Blinka

Then, pip3 install pyftdi which will also install some other libraries

Then pip3 install adafruit-blinka

Test!

Finally, set the environment variable with

export BLINKA_FT232H=1

and run python3. In the REPL, type

import board
dir(board)

You should get no errors, and a list of the pins available

The following shows a typical run through installing and setting things up on Linux.

Install libusb

Run the following:

and answer Y to the prompt. This should install libusb.

Setup udev rules

Use a text editor to create and edit the file /etc/udev/rules.d/11-ftdi.rules and add the following contents.

Here we use nano, so run:

like this:

add the contents from above:

and then press CTRL-X and Y to save and exit.

The settings will take effect the next time you plug in the FT232H.

Install pyftdi

To install pyftdi and its dependencies, run:

Install Blinka

To install Blinka and its dependencies, run:

Set environment variable

We need to manually signal to Blinka that we have a FT232H attached. To do this, we set the environment variable BLINKA_FT232H. The value doesn't matter, just use 1:

Run the sanity check.

Now move on to the Sanity Check section and run the commands there to make sure everything is installed correctly.

After going through all the install steps for your OS, run this as a simple test to make sure everything is installed correctly. See the rest of the page for some potential hiccups you may run into.

Go ahead and plug in your FT232H to a USB port on your PC.

Check pyftdi is installed correctly

Launch Python:

and enter these commands:

You should get a list of connected FTDI devices. Most likely, there will be only one.

If an FTDI device can not be found for some reason, you'll see something like this:

If you get

NotImplementedError: Operation not supported or unimplemented on this platform

Go back to the setup steps and Zadig your board if on Windows

Check environment variable within python

At the Python REPL, type

import os
os.environ["BLINKA_FT232H"]

if you get a KeyError it means you did not set the environment variable right

If you see some other errors, go back through the install process for your OS and make sure everything ran correctly.

If you have set it correctly, you'll get a value back

Now type in import board

You should get no errors at all, in which case you can continue onto the next steps!

Board not supported None

If you get

NotImplementedError: Board not supported None

that could mean you did not set the FT232H environmental variable or you don't have the latest python libraries or the FT232H is not plugged in.

BLINKA_FT232H environment variable set, but no FT232H device found

If you get this error, check your USB cable - it could be that you have a charge-only not charge+sync cable.

If you're running Windows, verify you ran Zadig to install the libusbK driver

The device has no langid

On Windows, if you get ValueError: The device has no langid on an open_from_url() call that may mean you dont have libusbk.dll placed in C:\Windows\System32

NotImplementedError: Operation not supported or unimplemented on this platform

If you're running Windows, run Zadig to select/install the libusbK driver on this device

pyftdi.ftdi.FtdiError: UsbError: [Errno 13] Access denied (insufficient permissions)

If you get this on a Mac, it could be you have conflicting ftdi drivers installed, you'll need to remove them!

Versions of OS X up to and including 10.8

The easiest way we found to do that is run

brew remove libftdi

This may have been installed by avrdude, open-ocd or other tools! You may have to --force uninstall it depending on what tools you are using.

For OS X versions 10.3 to 10.8, there is a procedure outlined in section 4.1 of Application Note AN 134 from FTDI:

In summary, start a terminal session:

Go -> Applications -> Utilities -> Terminal

and run the following commands:

and then reboot the system for the changes to take effect.

Versions of OS X 10.9 through 10.13

For OS X versions 10.9 though 10.13, you can unload the kext driver by running the following command:

Please note, the system will automatically reload the kext driver upon restart, so this would need to be run each time before using the board.

You may want to use an alias or script to easily run this command.

Note that there are two versions of this board in the wild. It was updated on Feb. 12, 2020 to have a USB C connector, additional pinouts, and other features (see product page for more info).

Pinouts for both the new version (USB C) and the original version are below.

Power Pins

• 5V - this is the 5V power from the USB input.
• GND - this is the common ground for all power and logic.
• 3V power output - The new version has a 3.3V power output pin for up to 500mA

GPIO Pins

• D4 to D7 - can be used as either digital inputs or outputs.
• C0 to C7 - can be used as either digital inputs or outputs.
  

I2C Pins

• SCL - the I2C clock signal is on D0.
• SDA - the I2C data is on D1+D2.
• I2C switch - The new version has a switch that connects D1 and D2 for easy I2C interfacing. Move the switch to ON to use I2C and/or the STEMMA QT connector. You can then use either D1 or D2 for SDA.

On the original version only: Note that there are two pins (D1 and D2) which must be tied together and treated as one to use I2C.

SPI Pins

• SCLK - the SPI clock signal is on D0.
• MOSI - Microcontroller Out, Serial In is on D1.
• MISO - Microcontroller In, Serial Out is on D2.
• CS0 - Chip Select is on D3. This is not used by Blinka, instead use one of the GPIO pins from above (see example section).

STEMMA QT

• STEMMA QT connector - The new version has a Stemma QT connector that lets you plug & play any Stemma QT or Qwiic devices, sensors and displays. To use: make sure the I2C switch is set to ON to use the STEMMA QT connector.

All right, now that all that annoying install stuff is done, let's have some fun.

The following sections will provide some basic examples for the main use cases - GPIO, I2C, and SPI.

Installing Libraries for Breakouts

The general process for installing the CircuitPython library you are interested in will be the same as shown in the Python section of the Learn guide for your sensor. Just use pip3.

Digital Output

Let's blink a LED!

Here's the bread board layout. The resistor can be something around 1kOhm. We don't need to make the LED super bright.

First, let's do things interactively so you can see how it all works one line at a time. Start by launching Python:

Then, at the Python >>> prompt, enter the following to import the needed modules:

Next we'll create our LED digital pin and set the mode to output:

And that should be it. You should be able to turn ON the LED with:

And turn it OFF with:

And here's a complete blink program you can run to make the LED blink forever.

Save it as something like blink.py and then you can run it with:

Digital Input

Let's read a button!

Here's the bread board layout. Use something like a 10kOhm resistor.

We'll do this interactively also. So launch python:

Then, at the Python >>> prompt, enter the following to import the needed modules:

And now we create our button digital pin and set it to input.

And that's it. To read the current state of the button use:

This will return False when the button is not pressed and True when it is pressed.

Digtal Input and Output

Ok, let's put those two together and make the button turn on the LED. So we'll use two digital pins - one will be an input (button) and one will be an output (LED).

Here's the bread board layout.

And here's the code.

Save that to a file with a name like button_and_led.py and then you can run it with:

and the button should turn on the LED when pressed.

Let's talk to an I2C sensor.

We'll use the BME280 sensor which can read temperature, humidity, and pressure. Here's the wiring diagram:

For the older version of the board, you'll need to jumper D1 and D2 together:

Install BME280 Library

To install the BME280 library, run the following:

Note that this step is the same as shown in the main BME280 guide. You would do the same thing for any other sensor.

Run Example

Now we can run the simple test example from the library. Here's the code:

Copy and save the above code and then run it with:

and you should see it print out sensor readings over and over:

Using STEMMA QT

The new version of the FT232H features a STEMMA QT connector. This makes connecting to newer breakouts that also have a STEMMA QT connector super easy. Just use a STEMMA QT cable:

And then you can just wire the FT232H and the STEMMA QT breakout directly together. You don't even need a breadboard or soldering.

For breakout boards without a STEMMA QT connector, you can either use the header pins on both boards, that is, don't even use the STEMMA QT connector:

Or use the STEMMA QT connector on the FT232H breakout along with a STEMMA QT cable with one end of the cable cut off. You can then terminate the cut off end to the breakout in whatever way works best for your application.

There are also some STEMMA QT cables with header connectors which might be useful:

Let's talk to a SPI sensor.

We can use the BME280 again, since it supports both I2C and SPI. Here's the SPI wiring:

Install BME280 Library

To install the BME280 library, run the following:

Note that this step is the same as shown in the main BME280 guide. You would do the same thing for any other sensor.

Run Example

We can run the simple test example from the library. It is set up to use I2C by default, so we need to comment out the I2C stuff and uncomment the SPI stuff.

We also add an import for digitalio (since it wasn't in there) and change the CS pin to the one we are using with the FT232H - C0.

Here's the complete code:

Copy and save that code and then run it with:

and you should see it print out sensor readings over and over:

Let's draw to a display!

Here we'll use SPI to talk to a TFT display. And since we are on a powerful desktop PC, we can even use Pillow (PIL) to do the graphical heavy lifting.

Connect the Display

Here is the wiring diagram for connecting a 2.4" TFT LCD display via SPI. We aren't using touch, so don't need to worry about those pins.

Install Pillow/PIL

We'll just point you to the official Pillow/PIL page so you can follow the install instruction there.

You should be able to just use pip though.

Install CircuitPython RGB Display Library

This library provides support for the TFT display. To install, run:

Example Image

In this example, we'll load and display an image. Download the image below and save it to the same directory where the Python code will be stored. Make sure it is named blinka.bmp.

The Code

And here is the complete code listing to load and display the BMP file.

Save the above with a filename like tft_image.py and then run it with:

and boom! you should get a happy little Blinka showing up on your display.

Fancier Demo for Linux

There's a lot of power in Pillow/PIL. It can do more than just load and display images. Here's another example that creates a Draw object, uses a TTF font, and then grabs some stats from the PC and displays them.

The example has import and code initialization for multiple displays. Uncomment the one for the display you are using, and comment out all others. If you are using the same 2.4" TFT (ILI9341) from above, then you don't need to change anything. The example uses that as default.

You will need to change the pins being used to match those on the FT232H. Find these lines of code and make things changes:

With those changes in place, you should be able to run it and get some system stats showing up.

And it will update in near real-time.