A low-cost, open source USB 3.0 Software Defined Radio platform with many examples and tutorials to help you experiment with RF. Read more
This project was successfully funded on February 28, 2013.
About this project
If you missed the Kickstarter and are wondering about general availability, check out http://nuand.com/order.php
It's the last day for our project. Thank you everyone for the amazing support!
UPDATE: Source code is now available on Github ( https://github.com/Nuand/bladeRF )
UPDATE: Production bladeRF units will ship with removable-cap RF shields.
What is a Software Defined Radio?
As wireless technologies become ubiquitous, Software Defined Radios (SDR) are gaining popularity. Unlike most radios transceivers found in phones, WiFi devices, remote controls, etc. that can only communicate using specific wireless modulation schemes, Software Defined Radios are completely software based, which allows them to communicate with devices across the RF spectrum. Besides interacting with existing wireless devices, SDRs allow for the development of new wireless systems and protocols using intuitive software tools and APIs.
An open education
bladeRF is a platform designed to enable a community of hobbyists, and professionals to explore and experiment with the multidisciplinary facets of RF communication. By providing source code, thorough documentation, easy to grasp tutorials, and a place for open discussion, modern radio systems will be demystified by covering everything from the RF, analog, and digital hardware design to the firmware running on the ARM MCU and FPGA to Linux kernel device drivers.
A Software Defined Radio platform should not end at the hardware, which is why there is such a strong emphasis on documentation and tutorials. Starting with basic radio architecture and spanning into modulation techniques, high throughput USB Linux kernel driver design, basic telecommunication coding schemes, and MIMO, the platform aims to be the perfect tool for learning modern software radio design.
Powerful and portable
The bladeRF is a fully bus powered device that does not need to be plugged into an outlet for normal operation. For users who wish to do host processing, USB 3.0 SuperSpeed is the ideal high throughput, low latency interface that brings the PC closer to the antenna than ever before. For those looking for a standalone solution, the bladeRF accepts a 5V DC input and operates autonomously using the FPGA for signal processing.
Professional quality, amateur price
Professionally designed and verified, bladeRF prototypes were inspected through X-Ray superimposed layouts, and put through rigorous physical and electrical stress tests to ensure high quality mass production builds. Ultimately, this makes the bladeRF a high quality, low-cost Software Defined Radio capable of capturing 40MHz 12-bit full duplex quadrature samples in realtime.
A full solution in a single package
Out of the box the bladeRF can tune from 300MHz to 3.8GHz without the need for extra boards. The current open source drivers provide support for GNURadio among other things, allowing the bladeRF to be placed into immediate use. This gives the bladeRF the flexibility to act as a custom RF modem, a GSM and LTE picocell, a GPS receiver, an ATSC transmitter or a combination Bluetooth/WiFi client without the need for any expansion cards.
The bladeRF was designed from the beginning to be highly integrated and fully reprogrammable. This means more than just providing source code to modify the host software. The USB 3.0 (Cypress FX3) microcontroller firmware is available to modify, as is the Altera Cyclone IV FPGA VHDL, bringing developers as close to the RF transceiver as possible.
All the pieces were written, designed, and documented to not only teach but encourage modification at each level from the host software all the way down to the FPGA logic. The bladeRF allows for the USB 3.0 microcontroller and FPGA to be reprogrammed through JTAG or directly via USB. With freely available tools and development suites provided by the hardware vendors, the bladeRF's FPGA and USB 3.0 microcontroller firmware can be easily modified.
More than just RF
The functionality and openness of the bladeRF encourages people to use the platform as more than just an RF transceiver. The FPGA can act as an accelerator of any type from turbo decoding to video transcoding. The bladeRF can be easily adapted for use in custom embedded projects due to its low power requirements and the flexibility offered by the FPGA, FX3, and expansion port. For inquisitive developers, the platform can be used as a USB 3.0 and FPGA development kit.
- Fully bus-powered USB 3.0 SuperSpeed Software Defined Radio
- Portable, handheld form factor: 5" by 3.5"
- Extensible gold plated RF SMA connectors
- 300MHz - 3.8GHz RF frequency range
- Independent RX/TX 12-bit 40MSPS quadrature sampling
- Capable of achieving full-duplex 28MHz channels
- 16-bit DAC factory calibrated 38.4MHz +/-1ppm VCTCXO
- On-board 200MHz ARM9 with 512KB embedded SRAM (JTAG port available)
- On-board 15KLE or 115KLE Altera Cyclone 4 E FPGA (JTAG port available)
- 2x2 MIMO configurable with SMB cable, expandable up to 4x4
- Modular expansion board design for adding GPIO, Ethernet, and 1PPS sync signal and expanding frequency range, and power limits
- DC power jack for running headless
- Highly efficient, low noise power architecture
- Stable Linux and GNURadio software support
- Hardware capable of operating as a spectrum analyzer, vector signal analyzer, and vector signal generator
We have a set of fully functional prototypes that we have been using to verify the design and finish the driver development.
The bladeRF was very verified early on to work well with embedded devices such as a Raspberry Pi. The highly efficient power architecture makes the bladeRF the perfect embedded SDR.
The prototypes have been verified electrically and physically to ensure that the design is indeed ready for mass production. We have gone as far as X-Raying the finest pitch components to inspect the quality of the PCBs and of the assembly.
The USB 3.0 transceiver's 0.8mm package could have caused manufacturability issues, however by inspecting it with the X-Ray machine the design was proved to be correct. It is this attention to detail that will ensure high quality bladeRFs are shipped to backers.
Why do we need your help?
We've invested a great deal of our time and money into bringing the bladeRF to this point. With your help we can purchase all of components in volume at price points that will make the bladeRF's cost much more accessible to hobbyists and enthusiasts. However, we have a lot to do before we can start shipping the bladeRF. By funding this project you are also helping us:
- Get a USB Vendor ID from USB-IF
- Purchase test equipment necessary for accurately calibrating units before they are shipped
- Purchase radio test equipment and allot resources to developing the more complicated examples and tutorials
- Fund the development of expansion boards
Who should purchase a bladeRF?
The bladeRF is meant for new as well as experienced wireless developers. With drivers for many platforms and GNURadio support, the bladeRF is an ideal drop-in replacement for experienced developers looking for a highly capable and cost effective SDR.
The bladeRF is also meant for hobbyists and enthusiasts looking to learn about modern radios. By providing examples and tutorials along with open source support and many of the design files, the bladeRF provides an unparalleled guide to wireless development, especially for newcomers.
Who we are...
We are a group of software, hardware, and RF engineers dedicated to making understanding and experimenting with RF more accessible to hobbyists and enthusiasts by creating low-cost professional grade equipment. Drawing from years of industry experience, and many months of work we've put together the bladeRF platform. Most importantly, we are active developers with a strong desire to see our platform bring RF experimentation to a whole new level.
Make sure to check out our blog at http://nuand.com/blog/ or visit us on IRC in #bladeRF on Freenode.
Risks and challenges
Due to the nature of test equipment, we have sought legal advice to ensure that we are in compliance with American and international regulatory bodies.
As with any other hardware project, our delivery dates can be adversely affected by component inventory. To avoid such surprises during a mass production run, only parts with healthy stock estimates were selected during the design phase of the bladeRF. Additionally, we have approached and remained in contact with many component distributors about their inventory.
A successful mass production run has been carefully planned since the very beginning of the project. We selected a PCB manufacturer that has experience with prototype and mass production runs for boards that are similar to ours. By using the same manufacturer all the way through, we were able to correct and avoid issues that would affect mass production runs early on.
As of now our delivery dates are based on worst case scenario lead times with additional headroom for error for all our parts. Although we won't go as far as promising earlier delivery dates, there exists a good chance of it happening.
To further ensure the quality of shipped bladeRFs, we will run an extensive physical and electrical tests to verify each device. Additionally, each bladeRF's VCTCXO will be individually calibrated to a near perfect 38.4MHz frequency using a very high fidelity frequency counter.Learn about accountability on Kickstarter
The reference we are using is at 38.4MHz and feeds the FPGA. The FPGA feed the ADC/DAC clock with a maximum rate of 40MHz (as per the LMS6002D RF transceiver specifications). There are PLL's inside the FPGA which we envisioned being used for arbitrary frequency synthesis. The PLLs can be dynamically reconfigured after initial programming, but we fully intend to open source all the VHDL associated with the FPGA for full control over the entire chain. Moreover, the ADC and DAC clocks are independent in case you wanted to have asymmetric sampling.
Are the two SMA jacks one each dedicated to input/output or can you do half-duplex on one (or both) of them?
There is no T/R switch on the board, so one port is for TX and the other is for RX. An external duplexer is required if you want to operate with a single antenna, but given the frequency selective nature of those devices we figured we'd keep those off of the actual board itself.
The LMS6002D is capable of doing around +6dBm maximum output power. This is at complete saturation and will not produce nice linear constellations. Obviously the PAPR of your waveform will dictate the average output power, but expect anywhere from -3dBm to 0dBm to be a good and linear range for transmissions with a non-constant envelope.
The schematics are currently available from the support page on the nuand website.
The firmware for the FX3 USB controller as well as the FPGA and host software are all going to be open source with a mixture of GPL and BSD licenses.
We are currently keeping the layout of the board closed, but intend to have an easy to use net finder to easily identify and cross probe between the schematic and your bladeRF sitting in front of you without having to resort to reading silkscreen.
The total amount of baseband bandwidth that is able to be captured is 40MHz via the quadrature ADC/DACs. The analog inputs/outputs are brought to a header for debug purposes or to sample a different RF chain altogether.
Internal to the LMS6002D RF transceiver chip are a set of low pass filters - one for transmit reconstruction filtering and one for receive channel filtering. They have a programmable range from 0.75MHz to 14MHz yielding a total bandwidth of 1.5MHz to 28MHz. These low pass filters are capable of being bypassed if desired, but it is highly recommend to use them at least for transmission reconstruction filtering.
Software Defined Radios, due to their wide frequency coverage and programmable modulation type, make it a bit difficult to ask the FCC to certify. Exactly what they're certifying it against (modulation type, emissions mask, etc) can not be completely known since it's so flexible.
As such, bladeRF, much like other SDR's, is considered a piece of test equipment. If you plan on connecting to an antenna, it is your responsibility to ensure you are in compliance with any of your local laws regarding spectral usage in both the TX and RX cases!
Pledge $5 or more
SUPPORTER - Receive a thank you on our site for supporting the bladeRF project.Estimated delivery:Ships anywhere in the world
Pledge $25 or more
VOTER - Obtain voting rights on the chapters we will write and make direct requests for examples and tutorials.Estimated delivery:Ships anywhere in the world
Pledge $400 or more
BACKER - A bladeRF production unit, USB3 SS cable, and two SMA cables delivered by May. This is for backers who want to see the project get funded, and want a finished product with support under Linux, Windows and Mac OS X.Estimated delivery:Ships anywhere in the world
Pledge $600 or more
BACKER - Same as the bladeRF production unit with a larger 115KLE FPGA. Although not necessary for most applications, the larger FPGA gives the bladeRF more signal processing power.Estimated delivery:Ships anywhere in the world
Pledge $650 or more
14 backers Limited (1 left of 15)
ENTHUSIAST - A bladeRF pre-production prototype, USB 3.0 SS cable, and two SMA cables delivered by April. This is for backers who want to wait out the initial hardware cleanup spin. Full support for a C interface as well as GNU Radio integration. If issues exist with this spin backers can exchange their boards for a later spin.Estimated delivery:Ships anywhere in the world
Pledge $800 or more
6 backers All gone!
DEVELOPER - A bladeRF beta prototype, USB 3.0 SS cable, and two SMA cables delivered by March with early access to source code. This is for backers who want to get in on the ground level and want to help shape the direction of the device and software. Drivers will be limited to a C interface and, rudimentary GNU Radio support - all under Linux. If issues exist with this spin backers can exchange their boards for a later spin.Estimated delivery:Ships anywhere in the world
Pledge $800 or more
Two 15KLE FPGA bladeRFs. This is identical to other May deliveries, except it ships with an SMB cable, and two units ready to run in synchronized clock mode.Estimated delivery:Ships anywhere in the world
Pledge $1,200 or more
Two 115KLE FPGA bladeRFs. This is identical to other May deliveries, except it ships with an SMB cable, and two units ready to run in synchronized clock mode.Estimated delivery:Ships anywhere in the world
- (30 days)