We use cookies to provide our visitors with an optimal site experience. View our privacy notice and cookie notice to learn more about how we use cookies and how to manage your settings. By proceeding on our website you consent to the use of cookies.
I am studying as a graduate student in electrical and electronics engineering. For my thesis, with FPGA, I am going to process an input which is 240Mhz or 1.3GHz RF sine wave. After a user defined delay, I am going to generate an output digital signal which has the same period and frequency such as the input had.
I am quite new at this area, I need your advices on FPGA selection, ADC selection and the other needs for this study. I am planning to use VHDL, so any technical advise on communication protocols would be great for me as well! Cost efficient way would be prefered ofc
IMO, you are in a hard place as direct digitization of a 240 MHz is on the edge of FPGA performance when we consider Nyquist. The 1.3 GHz will require a proper RF front end to down-convert the signal.
Welcome to the DigiKey TechForum! In addition to APDahlen’s recommendation above, I encourage you to check out the “Education” section of the TechForum. It may have some insight that you’re looking for in regard to FPGA and the communication protocol used, as well as some additional considerations for your project.
Thanks for the recommendation, but it is an enstitution project that is a part of a greater project which is going on so i have to do it as the exact values or closer to these. I can of course lower the frequency but not that much like 1 MHz.
Still in need of a FPGA that contains proper ADC or that support an ADC with high speed sampling rate (MSPS at least). If you recommend anything else, it would be awesome. Plus, I am gonna apply an MSc fundation support program. Not much but one is greater than zero, right
So you know, it’s only necessary to post one thread per topic–it’s the same people reading pretty much everything on this end.
Describing the specific needs in much more detail would likely be very helpful in for identifying possible solutions. What exactly do you mean by “process” an input for example?
It’s one thing (for example) to detect whether an input has a frequency of 240Mhz or 1.3GHz for example, but a very different thing indeed to digitize a signal at those frequencies. ADCs of the sort that would let a person do the latter are not inexpensive or easy to use, and FPGAs that can deal with data at the rates that it comes out of such a conversion process aren’t cheap either. (Sampling a 1.3GHz signal at Nyquist rate with 8-bit depth makes data at 2.6GB/s)
There are techniques such as interleaving ADCs that can can be used in high-speed conversion applications, but they’re not without their own difficulties.
However, if all you’re doing is measuring the frequency (period is the same thing, expressed as an inverse…) of a sine wave, then a decent linear comparator might be a much easier/cheaper conversion device, and delay line devices do exist that might amply serve the mentioned delay function.
Defining the problem appropriately here could very easily mean the difference between $20 and $20,000 in cost of hardware to do the job…
Thanks for the answer Rick!
To clearify the topic then, first i have to detect the input if it’s the one which is needed. It is 1.3GHz or 240 MHz on this application To detect them I need to use proper interface according to the ADC and FPGA pairing. Then I have to calculate its duration, period, frequency etc. and then I need to set a delay getting by user via UART. After the delay i need to trigger an output which could be a square wave or an analog output not sure about this part which has the same frequency and period.
As an opinion, I consider that maybe i can work on lower frequencies to make a design that could be used on higher frequencies in the future, and then I can complete my thesis and provide some paper.
I thought, i need to get information about what i should buy and to do as method first. But yes, it has to be cost-efficient and applicable in the limited time like 12 months. Any help from you would be precious for me.
The best advice I can offer at this point would be to spend some time developing a very detailed description of your application’s requirements.
Start with the input signal: what’s it’s amplitude? Is it singe-ended or differential? To what is it referenced? What’s the source impedance? What kind of noise will it contain?
The measurements that are mentioned: what signal features need to be measured? To what degree of accuracy or precision? How often do the measurements need to be made?
The output signal: what it to receive it, and what requirements occur as a result? What signal features (amplitude, frequency, etc) matter? How accurate to these signal features need to be?
This delay that is mentioned: What range of delay is required? How accurate does it need to be? How rapidly does it need to be changed? What other limitations or requirements does your application have?
These are some of the questions that you will probably need answers for before you can make choices about hardware. Taking the time to carefully define the problem, so that you understand what things matter and what things do not will have a large effect on your success.
