[Top][All Lists]

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

[Discuss-gnuradio] FrontEnd4702 take 2

From: cfk
Subject: [Discuss-gnuradio] FrontEnd4702 take 2
Date: Tue, 24 Feb 2004 14:57:50 -0800

Dear Matt, Eric and others:

    I just put the schematic and parts placement I have been working on
today on my web site for review in pdf file form. The Orcad Capture ".dsn"
file and the Orcad Layout ".max" file are available to anyone that might
wish them. For some reason, I cannot get into the Wiki right now, but when I
can, I will update the links and add some description of this front end

    The changes relate to trying to make the board per the USRP mechanical
constraints firstly. I made the board 2000 mils hi X 2450 mils wide. The
growth is on the left side as viewed from the top of the USRP with the board
in the upper left slot. I made the mounting holes per the 150mil setback
(first three) and the 300mil setback (bottom right corner).

    I did add 10uf, 0.1uf and 1000pf on dvdd, avdd and +5. Additionally, I
put an opamp in place of the MAX518 and am driving it from aux_dac_a from
the USRP. Currently I am thinking MAX406, but I may change to a Linear
Technology LT1678. If anyone has a better idea, I am all ears (even have my
Mickey Mouse hat available with extra ears).

    There are now two 4702 tuners on the board. The first one has its MA0
pin grounded so it should respond to I2C address 0xC0 and the second has its
MA0 pin left open so it should respond to I2C address 0xC2. The agc voltages
are common, although if you think the opamp should be a dual and use the
aux_dac_b, that is OK with me.

    The schematic has grown to two B size pages and they are
http://home.pacbell.net/cfk/frontend4702_1.pdf and

    The layout is shown as an Xray view through the board and is at
http://home.pacbell.net/cfk/frontendcomponent.pdf . Please view it as a
parts placement idea only at this time as it isnt routed yet. I am hopeful
that I will be able to make it just a double sided board. That is, the
component side will have most of the traces. This leaves the majority of the
solder side as a ground plane with traces as required in it. I will route
the power supply lines as wide traces. Currently, signal traces are planned
at 12mils and the power supply traces at 25mils or in a couple of long
cases, 50mils.

    The two 4702 tuners just *barely* fit between the top mounting holes and
the interface connector. I just dont see any other way to fit the 3 pounds
of stuff into the 2.5 pound bag right now. Although again, I am all ears.

    On the interconnect sheet, the parts listed as X1, X2, C8, C9, C11 and
R7 are uncommitted footprints. My plan is to salt uncommitted footprints
around the board in the spare space to allow some prototyping changes for
the next design.

    I look forward to criticisms, comments, suggestions and other rowdy
behavior as appropriate.

side p.s to Matt: Is the schematic file available for study by me?


reply via email to

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