The HPSDR Project
Q. I have a question that is not covered in this FAQ. How/who do I ask?
A. After searching for an answer and not finding it, usually the best way is to post your question on the HPSDR Discussion List (reflector). This allows two things to happen: (1) it permits someone other than the very busy developers to answer he question if they can, and (2) it allows everyone on the list to gain the benefit of any reply.
Q. How do I get in direct email contact with project leaders?
A. The project leaders are active on the HPSDR Discussion List and you may contact them by posting a message to the list.
Q. What is the status of the various boards or modules?
A. Here's the scoop on some as of Jaunary 1, 2009:
ATLAS - in production, order through http://tapr.org PINOCCHIO - in production, order through http://tapr.org OZY - 1st production run - currently being shipped. Order through http://tapr.org JANUS - 1st production run - currently being shipped. Order through http://tapr.org MERCURY - 1st production Order through http://tapr.org MERCURY-EU - alpha - see DJ8AY PENELOPE - Sold Out http://tapr.org ALEX - Pre-production
All others - in various stages of design/development -- see their wiki pages.
Q. What modules would I need to get a working HPSDR transceiver on the air?
A. It is important to remember the goals of HPSDR. All modules are not meant to be combined together to make a “single flavor” HPSDR transceiver. A number of different combinations will be possible (examples: Horton or Mercury for the receiver). How the modules are used and combined are in the hands of the experimenter/builder. Some users may not even wish to make an entire transceiver out of the modules (example: SDR1000 owners who only want to use Atlas, Ozy, and Janus to replace their sound cards).
In the words of Phil Covington, project leader for a number of the modules, “HPSDR was not formed to be a manufacturer of finished Ham Radio equipment. Its primary purpose is to develop a High Performance SDR in a modular fashion by experimentation with various methods.”
If your only goal is to get “on the air” with an SDR transceiver, there may be cheaper and/or easier routes to achieve this goal (Softrock or Flexradio).
If your goal is high performance software defined radio with a “roll your own” mentality, then the HPSDR modules should enable the creation of your own high performance SDR transceiver.
Q. Will the modules be offered in kit or assembled form, and what about cost?
A. Atlas and Pinocchio are offered as a bare board and kit of parts. Ozy and Janus are offered either bare board or assembled and tested. A hard to get partial parts kit is being offered or Janus. Future module costs to be determined.
Q. Will Ozy and Janus "bare boards" be available?
A. Bare boards (not kits of parts) are available through TAPR.
Q. Why doesn't TAPR offer a kit of parts or at least the hard to obtain parts for Ozy or Janus?
A. A partial kit of harder to obtain parts is being offered for Janus. Potential users may certainly get together for a group buy on other parts needed to complete the boards. There are several reasons for TAPR (or HPSDR) not offering complete kits: (1) being an all-volunteer organization, it would take tremendous manpower to break the parts down to individual kits and package them, (2) there is a very large support problem for kit builders whose boards do not work when completed, and (3) the cost of a kit of parts would be about equal or may exceed the cost of an assembled and tested board.
Q. Will the Gerber files (PCB artwork) be available for anyone's use?
A. Yes. They are released under the new TAPR open source hardware license called OHL. The board designer may restrict to non-commercial use. The OHL license was finalized and approved in May 2007. See 
Q. Why not put OZY and JANUS on a single board?
A. The overall HPSDR project design philosophy has been to partition the design into modules small enough to allow experimentation with part and design changes and to be able to put together a system meeting individual needs. Putting the ADC chip with associated circuit on the Janus board allows a future (and hopefully better) chip to be used on a similar board, but keeping Ozy for the interface and control. Flexibility is the goal.
Q. How much better will the Ozy-Janus combination be in terms of performance when used with the SDR-1000 in place of a sound card such as the Delta 44?
A. To be determined -- but of course, we expect better results. There are some preliminary results on the wiki and in the discussion list.
Q. Will a Ozy-Janus-Atlas combination work with my PowerSDR software used for my Flex Radio SDR-1000 in place of a sound card in my PC?
A. Yes, that was one of the early goals of the HPSDR group. Phil, VK6APH, did confirm with Gerald and Eric at the Flex-Radio meeting at the Dayton Hamvention 2007 that Ozy/Janus will be fully supported in future releases in their 'mainstream' releases of PowerSDR. Bill, KD5TFD, will be working with Eric from Flex to accomplish this. At this point it is not known exactly when, and what version the support will begin, but it will happen. Direct all questions regarding Janus/Ozy to the HPSDR Discussion List (and NOT the Flex-Radio list), as folks have been doing and admirably responding. The arrangement with Flex-Radio required the donation of a working Ozy/Janus to Flex-Radio and this has been accomplished after TAPR approved the expense.
Q. What will be an appropriate software for companions like Janus + Ozy + Phoenix + (Alex??) ?
A. These boards, and also with the addition of Mercury, will run using PowerSDR. --Phil VK6APH
Q. Is the HPSDR project going to use Windows or some flavor of Linux?
A. Yes! (Eventually both, that is...but, currently, the supported OS is WinXP). There is currently work being done for Linux and dttsp.
Q. What are the recommended minimum system requirements for the PC I will use for the HPSDR?
A. USB 2.0 is a requirement. Currently, the OS recommendation is WinXP. Windows 2000 is NOT recommended as the USB 2.0 stack on Windows 2000 is just too slow.
At this time, there are no solid recommendations for minimum CPU or RAM that are based on actual testing with HPSDR hardware of how low we can go.
FlexRadio does have Minimum Recommended PC Configurations for systems using the PowerSDR software. Since the HPSDR hardware may use PowerSDR, these specs are probably a good guide to what would be advisable for the HPSDR. FlexRadio's numbers from their website are as follows:
Processor: Min: 1.5GHz Recommended: 3.2GHz+ or greater
Memory: Min: 512MB Recommended: 1GB+ (use the fastest RAM available)
Q. What user name and password do I use to access the HPSDR svn repository?
A. None is required for reading the SVN, only required to place something in the repository. The IP address of the repository is shown on the resources page of the main HPSDR.org website.
Q. Will HPSDR be developed for higher frequencies like those used for satellite and space communications, e.g. VHF, UHF and Microwave?
A. There is a group doing SDR for microwave:  Current HPSDR projects could certainly be used as an IF for a transverter, but there is nothing going on with HPSDR that is specifically aimed at microwave.
The HPSDR Wiki
Q. Do I need to log in?
A. Those who contribute by editing the wiki need to have a login.
Q. How do I get a account? (a login)
A. Request it from the wiki system operator, email: Dave Larsen
Q. What if I find that a correction is needed in the wiki?
A. Reports such as this are welcomed by the wiki system operator, email: Dave Larsen
Q. What is the recommended means of powering ATLAS?
A. Demeter. Until Demeter is available...
The ATX 20 pin power connector on the Atlas board enables the use of standard PC power supplies. (Please Note: There is no reason that you cannot utilize a non-PC power supply regulated and wired to provide the proper voltages to the 20pin connector. A non-PC power supply could also enable custom current limiting of the voltages going to the 20 pin connector, an advisable setup when testing or prototyping boards plugged into ATLAS. An analog power supply may be an attractive option for users particularly concerned about spurious emissions in the HF band which some low cost PC power supplies may produce.)
If you choose to use an ATX computer power supply care should be taken that the -12V current requirement is met. (Note of warning: some versions of the attractive picoPSU do not provide proper -12V current capacity. Check before you buy.)
As a reference for current requirements (reported by Bill Tracey May, 11, 2007), Ozy/Janus used by a SDR100 had the following current usage:
Ozy/Janus +12v: 200 ma +5v: 180 ma -12v: 70 ma
Obviously additional boards connected to the Atlas board will increase these numbers.
Projections of current requirements for other boards are(as reported by Phil Harman, June 6, 2007):
Penelope +12v: 200 ma +5v: 300 ma
Mercury +12v: 200 ma +5v: 500 ma
Q. Will the Atlas be offered assembled?
A. Probably not. It is fairly easy to assemble with a very minimal amount of surface mount parts. There are quite a few solder pads due to the 96 pin connectors. If you are not able to do this work yourself, our advice is to ask on the HPSDR Discussion List (reflector) to see if you can pay someone to do the work for you.
Q. Can solder paste and a hot air heat gun (or oven) be used on Atlas for "all those connections" ?
A. It is possible, but at least one report indicates problems with the center row on the connectors. If considering doing this, we suggest you ask on the discussion list. If anyone has had success or failure, please report it to the wikisysop so we can update this reply.
Q. Will a larger (or smaller?) number of slots version be offered?
A. Possibly, if the need and demand warrant. Nothing is in the plans right now (as of May 2007).
Q. I don't see assignment of all the bus pins. Is there a list somewhere?
A. Some are not assigned a function yet, due to the developing nature of the HPSDR project and the use of the FPGA.
A. The bare board and connectors are now available from TAPR http://tapr.org
Q. Will the USB connection from Ozy to my PC require anything special in terms of USB port specification or drivers?
A. A USB 2 connection will be required on the PC. Most modern PCs have this as standard. With MS Windows, for the USB driver we are using the LibUsb-Win32 library which is a free download from http://libusb-win32.sourceforge.net/ A Linux version is also available, see http://www.linux-usb.org/ and http://libusb.sourceforge.net/ . Experience will tell us if there are any problems with certain types of USB2 ports.
Q. Why do we need a "configuration device" when the software can just load the FPGA via USB and the Cypress CY7C68013 (FX2) chip? The schematic shows the programming pins connected from FX2 GPIO pins to FPGA.
A. It does load via USB and this is how OZY is normally used. BUT, there will come a time when someone wants to use the OZY without PC attached and the configuraton device allows this possibility.
Q. Is the design of Ozy such that it can be used for other purposes than SDR?
A. We certainly hope so and expect that some will use it as a learning tool or development platform for other projects not even remotely related to SDR. It provides an inexpensive piece of hardware for many purposes.
Q. Is Janus a "sound card" ?
A. NO! The usual meaning of a sound card is one which plugs into a personal computer (ISA, PCI, or other bus). The Janus module plugs into our Atlas bus and contains some of the components of the usual sound card. It also requires the Ozy or similar interface to use it in applications which call for a PC sound card.
Q. Will I be able to use Janus for other non-SDR sound applications with my PC?
A. In theory, Yes! This will require a Windows or Linux driver; there is no reason one can't be written, we just need a volunteer!
Q. Why are there no input RF filters on the Mercury PCB?
A. This is due to the multi-function nature of the Mercury board. Whilst primarily a high performance HF (and VHF/UHF in alias mode) receiver, Mercury can also be used for other functions. For example, it can be used as a high performance 0-55MHz spectrum analyzer, sampling oscilloscope and VNA (by using Penelope as the signal source for the VNA and also as a tracking generator). In these cases it is desirable not to have any RF filtering before the ADC on Mercury.
When used as a receiver the amount of filtering will vary depending on the antenna system the user has. For most situations with say an HF beam or dipoles the natural frequency response of the antenna may provide adequate filtering. For the user with a high performance broadband antenna e.g. a rhombic, or very strong local signals, then additional filtering may be required.
An external set of filters will be provided as part of the Alex project.
Additionally, HPDR is a journey and not a destination! We fully expect higher performance ADCs to be come available in the future. These newer devices will still require some form of input filtering. By using external filters the cost of replacing the ADC board is reduced.
Q. Why are there no output RF filters on the Penelope PCB?
A. This is due to a number of reasons. Firstly, whilst Penelope is primarily an HF ( and VHF/UHF on alias) exciter it can be used for other functions. For example, when used with Mercury it can form a low level signal source as a tracking generator or VNA. For these functions the lack of output filters is an advantage.
Secondly, Penelope generates RF directly at the desired output frequency by synthesizing the required RF waveform using a DAC. The lack of mixers, DDS, frequency synthesizer etc means the output spectrum of Penelope is particularly clean. In fact the spurious output at 0.5w meets the FCC requirements without additional filtering.
Thirdly, Penelope is an exciter. Whilst we expect it will be used by QRP operators as is we also expect it to be used to drive a higher power amplifier. In the latter case the user will most likely provided external filtering as part of this power amplification.
Fourthly, Penelope does provide a 55MHz LPF that can be placed in circuit after the DAC and prior to the 0.5W PA. If desired the user can add external bandpass filters here. Alternatively, the filter can be bypassed and/or an external VHF/UHF filter fitted such that the alias output of the DAC can be used on the higher bands.
Fifthly, if is desirable to use LPFs that may be also be used before Mercury. The IP3 performance of Mercury is very good and using small inductors, that are quite acceptable for removing the harmonics from Penelope, results in a significant degradation in IP3 performance.
An external set of filters will be provided as part of the Alex project.
Additionally, HPDR is a journey and not a destination! We fully expect higher performance DACs to be come available in the future. These newer devices will still require some form of output filtering. By using external filters the cost of replacing the exciter board is reduced.
Project leaders, developers, documenters: feel free to contribute answers -- especially where it says "TBD" or "Answer pending."
General Readership: have a suggested question that should be here? Email: Dave, KV0S