Bristol Braille Technology


append delete BBT

(Posted on behalf of Tom Lloyd, an independent embedded software engineer collaborating on the Quixote)

My thoughts on the XO and Pi, I am imagining a few different set ups
depending where you are in the world.

1. PC -> Quixote - (Aciting as braille display/keyboard)
2. PC -> Quixote - (Integrated Pi) (Acting as braille display/keyboard)
3. XO1 -> Quixote (Integrated in XO1) (portable device)
4. Stand Alone Quixote (portable device)
5. Quixote (Intergrated Pi) (portable device)


1. This scenarios is where someone has bought the Q as a Braille
keyboard and display only. The obvious choice would be to use USB for
this connection. However I don't know how much power the Q will take to
drive the display. If it is over the amount we can drive through USB (1,
2 or 3) ports combined then there might no longer be any advantage over
using USB over RS232. (Might be a good question for the forum / hardware


2. This scenario has potential technical problems in certain usecases, I
am considering the Q being used in a Braille Note like device where the
user maybe want to use it is a portable computer, a Braille keyboard
and also access all their files when connecting it to a pc.

a. USB Mode 1: Used as Braille Display / Keyboard
b. USB Mode 2: Used to access Pi Mass Storage

The problem arises that the The Pi only has USB Host Ports and no device
ports. Making plugging the Pi into a computer as a USB device
impossible. To solve this problem I would suggest using a
microcontroller to drive the Q hardware and to provide a usb device
interface to the outside world while providing an interface to the
internal storage either directly or indirectly.


2a.Computer -USB-> Microcontroller -> Q Hardware
                   Mass Storage Card
2b.Computer -USB-> Microcontroller -> Q Hardware(inline with Stand Alone[4])
                  Mass Storage Card


* 2b USB peripheral interfaces less common than SPI/I2C 4 Microcontroller
** USB Host hardware in more costly than SPI/I2C for Mass Storage only.

2a. if the Pi and Q are connected using a microcontoller there is the
possibility to (in hardware/software) control the Q display/keyboard
while still allowing access to the devices internal storage.

2b. The same con fig without a RPi integrated could be offered for a cheaper
version that could have a RPi added at a later date.
If a reasonable Microcontroller is chosen then basic Braille Note
functionality could still be achieved on a portable version of the device
This would be marketable at people where lower cost in paramount.


3. XO1 -> Quixote these are two options RS232 / USB depends if the Q is
integrated or not. Storage can be used from the XO1 and should not be a
problem. Powering the device might be though.


4. Stand Alone Quixote - See point 2b


5.5. Quixote (Intergrated Pi) (portable device) See point 2a


Reply RSS


append delete #1. Ed Rogers

Thanks Tom, I think that gives us a fairly solid basis to work from. I see it like this;

:: 1) Plain old Braille display

This will form the basis of our development for the immediate future. We cannot assume that the Pi will always the best computer on a chip to integrate, even if there were always to be an integrated SoC of some kind, so this gives us that flexibility. There is no reason not to present this as a plain display, but I think the added functionality of the other options make them of more general use.

:: 2) Optional stand-alone or Plain old Braille display modes

You raise an important point. Given the relative simplicity of the timing and actions it makes sense to use a separate cheap chip which allows us to swapt inputs freely. See no.5.

:: 3) Integrated into the OLPC XO1 ($100 laptop for developing nations)

I don't think it would be possible to integrate the Quixote into any XO1s. OLPC relies on very large orders (100,000+), which the Braille market probably cannot reliably supply.

The challenge will be getting the Quixote to work with their custom 'Sugar' interface. This is probably best left for a later date. We can, however, hope to get it working with those XO1s running Windows XP or Linux with the XFCE interface.

:: 4) Stand-alone with custom onboard, cheap computer

This is a very interesting suggestion. This could present a saving of approximately 10% on the final unit cost. The Natesan display has something of this kind built in for this reason. However I think it could prove to be a false economy as it is a serious reduction in features for only a small saving.

:: 5) Stand-alone with onboard Raspberry Pi computer

Coupled with the discrete micro-controller (no.2) I think this presents the best option. The Raspbery Pi would not only serve as a fully fledged PC, but also as an ebook reader and note-taker. these features would not have to be built from scratch as there are Linux distros out there that provide this. This way the Quixote could also be kept up to date with changes and improvements in technology, such as better screen readers.

append delete #2. Paul Watford

Hi I designed and built a braille display using 9 pin Braille for my Uni project years back. It used a Pic Microcontroller connected to a solenoid array which drove pins up and down in a single block. The idea was that the user just kept his/her finger in one place while the braille moved under his/her finger. Feel free to contact me on

append delete #3. Mike York

Thank you so much for these combinations. Your post has helped me a lot with my writing tasks for


(Leave this as-is, it’s a trap!)

There is no need to “register”, just enter the same name + password of your choice every time.

Pro tip: Use markup to add links, quotes and more.

Your friendly neighbourhood moderators: Ed Rogers, Steph