[eagle] Re: 70 cm Receiver IF Spurs - Problem Solved
juan-rivera at sbcglobal.net
Sun Jun 17 19:16:26 PDT 2007
At least in this particular case jumpering across the switch didn't work.
(See my log for today on Eaglepedia.)
Since the CAN-Do module is such an imbedded and pervasive component in
Eagle, I think it needs to be free of radiated and conducted EMI in both
directions. It's a great little interface but I'm afraid it has some very
serious shortcomings at the moment:
1. Conducts noise back towards the power source where it can impact other
loads on the same power bus (one ohm of resistance out there was enough to
screw up the receiver.)
2. Provides a noisy +14 to its load (the noise went right through the
receiver like fat through a goose and ended up in the IF as spurs.)
3. Radiates EMI that can adversely affect sensitive circuitry at a
distance of over two inches from the rear of the PCB
Attempting to mitigate those problems will put a huge burden on every
payload and the adverse impact might not be fully known until integration
and test, if then. Trying to go back and band-aid something like this after
all the payloads come together would be a very bad corner to get backed
into. You could do what I did and sacrifice 2 inches of space inside every
module chassis, or you could try to build a two-section chassis and filter
the lines passing through the bulkhead, adding complexity. But that doesn't
address the potential impact to another payload from the spikes that are
propagated back towards the common +14 bus.
I would expect that every receiver, block up or down converter, transmitter,
or amplifier will be impacted by this noise in a similar way. In my opinion
this is a global problem that needs a global solution. If the CAN-Do module
can be made EMI-free, or meet a yet to be determined EMI spec, then all of
the payloads teams can start their design with a known environment.
If the CAN-Do module is not modified then I think the only safe thing to do
is to build each payload so it can function using CAN-Do switched +14 volt
power (the receiver can't in the Rev-A configuration.) If your payload can
function with that noisy +14, without any adverse impact, then it is
probably a pretty well protected circuit and should be fairly immune to
whatever noise is added to the source by conducted EMI from other loads on
the power bus. It just seems simpler and cleaner to fix the problem at the
I have an idea for a noise generator that I will stick on my log... This
could be the go/no go test for each payload. I'll go sketch it up and get
it uploaded by about 2000 Pacific Time.
From: Chuck Green [mailto:greencl at mindspring.com]
Sent: Sunday, June 17, 2007 5:51 PM
To: juan-rivera at sbcglobal.net
Cc: eagle at amsat.org; Bill Ress; Dave Black (Home); Dave Black (Work); Dave
hartzell; David Smith; Don Ferguson; Juan.Rivera (Work); Samsonoff at Mac. Com
Subject: Re: [eagle] 70 cm Receiver IF Spurs - Problem Solved
One of the purposes of the CAN-Do! was to provide a standard interface
so that a module couldn't do something nasty and take out the entire
satellite. So custom mods to the CAN-Do! are not allowed. Having said
that, we have always assumed that a jumper across the CAN-Do! power
switch would be installed for those modules that needed to be powered up
Juan Rivera wrote:
> I don't want to tear into the CAN-Do module. I feel like it is a loaner
> not mine to modify. It would be interesting to disconnect the switch and
> the current monitor from the CAN-Do power input and move it to those user
> pins as you suggest. That might just work, but I also know they don't
> to have multiple versions of the module floating around and this would be
> We also can't loose sight of the radiated noise from that inductor. I can
> see signs of EMI if it gets within 2 inches of the sensitive areas of the
> -----Original Message-----
> From: Chuck Green [mailto:greencl at mindspring.com]
> Sent: Sunday, June 17, 2007 9:18 AM
> To: juan-rivera at sbcglobal.net
> Subject: Re: [eagle] 70 cm Receiver IF Spurs - Problem Solved
> Hi Juan,
> I am impressed with your testing and analysis.
> I find it curious that you were unable to filter the power from the
> CAN-Do! . But switching noise is notorious for being difficult to
> filter. So maybe that is something to just forget. But in a total
> system (satellite) it may not be possible to forget. We may need to
> find a way to filter the power for sensitive modules even if not using
> power from the CAN-Do! ; in which case, such a filter may also filter
> the noise from the CAN-Do! .
> If your analysis is correct, and it certainly seems to be, then it would
> also seem possible to remove the switch from the CAN-Do! (not just
> bypass it) and rout your clean power source through the CAN-Do! current
> measurement circuit thus saving having to do this in the module.
> Great work! And thanks.
> Juan Rivera wrote:
>> Hi Chuck,
>> Bad news I'm afraid. See my report for details... I've sent out a link
>> the group.
>> -----Original Message-----
>> From: Chuck Green [mailto:greencl at mindspring.com]
>> Sent: Saturday, June 16, 2007 8:57 AM
>> To: juan-rivera at sbcglobal.net
>> Cc: eagle at amsat.org; David Smith; Dave Black (Work); Dave Black (Home);
>> Samsonoff at Mac. Com; Juan.Rivera (Work)
>> Subject: Re: [eagle] 70 cm Receiver IF Spurs - Problem Solved
>> Hi Juan,
>> This is fantastic!
>> Please try one more experiment. Pass the module current through the
>> current measuring circuit on the CAN-Do! while bypassing the switch
>> transistor (you could just short across the switch transistor) and see
>> if the problem reappears.
>> If you end up having to use the user pins, you will need to add current
>> measuring circuitry to your module. While this is no big deal, its one
>> more thing you might not have to do.
>> Juan Rivera wrote:
>>> The 70 cm Receiver spur problem has been solved. Click here
>>> for details.
>>> Via the Eagle mailing list courtesy of AMSAT-NA
>>> Eagle at amsat.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Eagle