[eagle] Re: CAN-Do noise problem is proving to be a difficult one to fix

Juan Rivera juan-rivera at sbcglobal.net
Fri Jun 15 15:30:22 PDT 2007



I'll need the CAN-Do folks to come up with a shielded inductor that will
have the same form, fit, and function.


I know John has been working on a two-chassis implementation for the
receiver to try to get around the thermal issues that have been raised, but
I am personally apprehensive due to the added complexity of any
interconnects that will be required.  I don't want to beat this milled
chassis idea to death but it might be worth considering from a thermal
standpoint.  If you had enough thermal mass in the chassis perhaps it would
never have time to reach those extreme low temperatures during an eclipse,
rendering the need for a two-chassis approach unnecessary.  It would
eliminate my concern about flexing and with a milled enclosure you could
also move the noisy CAN-Do module into a separate compartment and filter the
power through the common wall.


I've been thinking about DC power distribution in general.  Can someone
explain the plan for Eagle to me?  In aircraft, at least in the one I am
familiar with, they do not use the chassis as a return.  All DC power is
treated the same way as AC power and distributed from a common point as a
twisted pair to minimize common mode noise.  Great care is taken to
eliminate ground loops by insuring that the ground returns are back through
the twisted pair to the source and not through a sneak path to the frame.





From: Jim Sanford [mailto:wb4gcs at amsat.org] 
Sent: Friday, June 15, 2007 2:16 PM
To: juan-rivera at sbcglobal.net
Cc: eagle at amsat.org; eagle at projectoscar.net
Subject: Re: [eagle] CAN-Do noise problem is proving to be a difficult one
to fix


My thoughts:
1.  We must stop the radiated noise.  Let's try obtaining and placing a
shielded inductor on the CAN-Do! module.
2.  I just read a great suggestion from Lou, amplified by a phone call I
just had with Rick, regarding three terminal filter devices.  Seems like a
good start.
3.  I agree that we need to review the bidding on packaging.  That will
probably be the LAST of these issues we take on.

Thanks & 73,
wb4gcs at amsat.org

Juan Rivera wrote: 



Please see my write-up here
ting.htm> .


The CAN-Do noise is not lending itself to an easy cure.  I think this whole
subject of RFI/EMI needs to be looked at closely.  Every payload will be
affected by noise on its inputs and can affect other payloads by creating
noise that is propagated back out from within its module chassis.


With the current scheme, of having a CAN-Do module located in the same case
as the main payload, there are several issues that need immediate attention:


The CAN-Do module provides dirty power - both the +13 VDC output and the
ground are noisy

The CAN-Do's switching power supply inductor, L3, radiates noise that can
affect sensitive circuits several inches away.


Both of these factors are adversely impacting the 70 cm breadboard receiver
in the form of 5 kHz spurs on the IF output.  As I mentions earlier, Rather
than work to clean up the CAN-Do module immediately, we should focus on
making the receiver function In the presence of this noise.  This approach
will result in a more robust design.  The CAN-Do module needs to be cleaned
up, but for now it acts as a great real-world noise generator.


I'm currently out of ideas.  I'd like to hear what you all have to say.








Via the Eagle mailing list courtesy of AMSAT-NA
Eagle at amsat.org
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://amsat.org/mailman/private/eagle/attachments/20070615/df2f007f/attachment.html

More information about the Eagle mailing list