[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Discuss-gnuradio] USRP and 100-pin Cypress chipset

From: bb+graffiti . spam . gnuradio
Subject: [Discuss-gnuradio] USRP and 100-pin Cypress chipset
Date: Sun, 21 Dec 2003 17:45:01 -0500
User-agent: nail 9.30 2/20/02


I was looking through the USRP info, and am very excited at what I see.
A few months ago, I did a little bit of preliminary work playing around
with the Cypress chipset you're using. I ended up with two problems.
The 128 pin version was extremely irritating to solder. Also, I am
far more a software person than a hardware person. When I saw that some
experienced people were working on the USRP, I figured that I would wait
till it was done and then build a few of them for myself.

The main question I have is how important the 128k of SRAM is. On my second
version of the boards I had made, I used teh 100 pin variant, which doesn't
have the address pins brought out. Is there any hope of working with just
the 8k of internal RAM? I'm generally going to only be doing input, and
would be willing to deal with only a single channel, although I am
interested in doing direction finding, which would be made easier with
two or more inputs.

You might also be interested in a piece of hardware that NIST has designed:
Basically, it's an FPGA hooked up to 64 individual microphones, which feeds
the data out via 100Mbit ethernet. It's entirely implemented in the FPGA,
with the addition of some RAM, though they're not really using the RAM since
retransmissions are impractical at the data rates they're working at.
It's hosted on a Linux box, and has a nice toolkit for array processing.
The lab I work for has been loaned one of these for research purposes. In
small quantities, the fab cost is around $900/unit.

You guys are doing great stuff here, and I look forward to being able to play
with live signals affordably.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]