A flexible signal generator based on the AD9914 DDS chip. Striving for lab-quality noise figures and supreme flexibility.
To make the experience fit your profile, pick a username and tell us what interests you.
We found and based on your interests.
Work has resumed on this project after a long hiatus due to time constraints, etc. etc. etc. - check out the smaller project I have being fabbed at OSH Park right now: https://hackaday.io/project/9084-mini-dvm
I was able to get most of the Easy-PHI template added to the schematic - an Altium version of the template module will be an unexpected but rewarding side-effect of this project. I am doing a fairly faithful reproduction of the KiCad schematics - part because no importer but also to gain familiarity with the template circuits. It's based largely on an Arduino Due but there are lots of features for use in a high-end physics test platform (central 10MHz clock, RF connectors by default, Eurocard sizing and connectors, etc.) which made it very attractive as a starting point.
Check those out on github and thanks for all the follows and skulls!
As I near the capture/layout phase of the DDS section of this project, I'd appreciate any feedback and/or usecases anyone can provide. For example, is it more useful (for you) to have an ultra low-noise signal < say 100MHz, or would you prefer to maximize the modulation capabilities of the DDS? I'm toying with some ideas on switchable output filters, as well as trying to get an idea of down the road, what kind of processing power the daughtercard should exhibit...
Thanks and hack on!
Hi Folks! I uploaded a quick intro video detailing the project here: http://youtu.be/D6CwkmWaq7M
I've added a very quick design document/block diagram to google drive here:
https://docs.google.com/document/d/1JdDacofMxN9zo3MFpjHVc2o74f6vh41sJwJNfwpOXBM/edit?usp=sharing
The FPGA daughter card is FMC compatible although it uses the "exception" style in that the processing is happening on the daughter card, not on the motherboard as is the norm. The FPGA will be programmed by the ATSAM3x8E and should enable the full 140MSPS modulation the AD9914 is capable of.
Currently a Spartan 6 LX16 is planned, although this may change to a Zynq depending on cost and flexibility.
One of the hardest things about any electronics design project is capturing requirements. That's why after some careful consideration I have taken Mathieu Stephan's recommendation to take a look at the Easy-Phi project to heart.
The Easy-Phi project is a very well thought out modular system that is compatible with the size and interface constraints I had already set for this project. Easy-Phi defines a backplane connector that will supply USB 2.0 connectivity as well as ample power and a reference timing signal. This, in addition to the eurocard form-factor I had already chosen makes this a really good fit and eliminates a lot of decisions I still needed to make about how I was to connect to the board.
One issue I'm facing is that the template is in KiCAD and I plan to use Altium Designer at least for the initial layout, although I am very interested in re-creating my Designer work in KiCAD and learning what looks like a really strong truly Open EDA tool.
Mathieu's very nice introduction to the Easy-Phi template finally convinced me. You can read it on his blog here: http://www.limpkin.fr/index.php?post/2013/10/09/Easy-phi-project%3A-the-template-module
Arduino IDE compatibility FTW!
Create an account to leave a comment. Already have an account? Log In.
Hello, as i understand you have 100 Mhz ocxo, but to achieve 3.5 GHz reference you have to multiply it by 35. It is quite chalenging task. What scehematic do you use for multiplier ? I ask because i have AD9910 and i need good reference signal with very low phase noise and jitter. And you probably have similar task.
Thanks for this project, I am a univercity student now and I want to make this signal generator project. I looked your files project. I learned some think about signal generator. But that is not unough for make that project myself. You add this site or other site the photo of this project step to step and how did you design this project by raspry pi ang altium designer,, Thanks,,
[this comment has been deleted]
Hi Stuart,
The AD9914 is only the latest of a very similar DDS family from Analog Devices - it can be clocked at 3.5GSPS - making 1.4GHz output doable.
You may be thinking of the AD9954, which was clocked at 400MSPS - giving a reasonable output of ~150MHz given nyquist and aliasing.
The tricky frequency multiplication voodoo in this project is trying to keep low phase noise for the REFCLK - currently taking 100MHz from a crystal oscillator and multiplying it to 3.5GHz.
More info coming shortly - I'm making time to work on this again.
Thanks!
-Logan
Become a member to follow this project and never miss any updates
By using our website and services, you expressly agree to the placement of our performance, functionality, and advertising cookies. Learn More
Do you have any details on the NLTL comb generator. Diodes used, design literature, schematic etc. Many Thanks.