Switching Noise: My preliminary "sniff" last night didn't detect any noise near the Can-do module. The overwhelming amount of noise was picked up when the probe was near L21. I could also vary the frequency by adjusting the DC input voltage to the radio, putting both the "sniffed" and the IF output spikes on 5.0 kHz with the same input voltage. I'll look deeper later today but I believe the IF spurs are coming from the receiver switcher and not the Can-do module.
The spec sheet for the max1776 goes to lengths to warn about how critical the layout is due to high peak switching currents. It also discusses radiated noise. See the section on PC board layout and grounding. If the supply and the RF section share the same common ground to the Can-do module I can easily see how switching noise could couple into the radio.
Since we have unused space at the back of the board I'd suggest isolating the supply away from the RF section and separating the power and RF grounds. The DC from the Can-do module could be brought back to the power supply via a short section of coax or a twisted pair to keep the DC noise isolated from the receiver ground completely.
Microphonics: I spent quite a while trying to isolate the microphonic source. I softly dragged an insulated probe across every component on the board. Some areas of the board seem to produce more noise than others but no single component jumps out. John, if you're right and its several ceramic capacitors that should be film that might explain it. I'll try to zero in on this later today.
73,
Juan WA6HTP
-----Original Message----- From: Robert McGwier [mailto:rwmcgwier@gmail.com] Sent: Saturday, June 09, 2007 7:04 AM To: John B. Stephensen Cc: juan-rivera@sbcglobal.net; eagle@amsat.org; David Smith; Dave Black (Work); Dave Black (Home); Samsonoff@Mac. Com; Juan.Rivera (Work) Subject: Re: [eagle] Re: New 70 cm Receiver Test Results
John B. Stephensen wrote:
Since the noise is always symmetrical around the received signal, the switcher noise must be modulating the VCO control voltages. There are two switchers - one on the receiver PCB and one on the CAN-DO module. We must determine which is causing the problem or whether both are. Can you cut the trace on the PCB and vary the DC voltage to the receiver independently of the voltage to the CAN-Do module?
I agree with this switcher comment and did as soon as I saw the picture. I really hope this is not the CAN-DO module as FLIGHT UNITS have been DELIVERED for P3E. We need to have Stephen, et. al. check the noise output of the switcher on the Can-Do module.
I hope that you can tap each component and find the microphionic one. Good candidates for microphonic components are the capacitors in the PLL feedback loop. The capacitors in the integrator are film types so they shouldn't be microphonic but perhaps SMT parts aren't as good as leaded parts. The ceramic capactors in the feedback path might also cause problems and could be changed to film capacitors. Another possibility may be the 4.7 and 10 uF bypass capacitors on the outputs of the linear regulators. These are ceramic because of concerns about tantalums in space but they might have good piezoelectric properties and could be switched to 1206 size tantalum capacitors. There could also be a cold solder joint somewhere.
73,
John KD6OZH