Re: problem testing local oscillator (definitely solved)

This contains the scottyspectrumanalyzer yahoo group backup
Post Reply
marc olanié
Posts: 59
Joined: Sat Apr 28, 2018 6:49 pm

Re: problem testing local oscillator (definitely solved)

Post by marc olanié » Sun Jan 15, 2012 6:29 pm

Hi Simon, Scotty  Problem seems to be solved.  I have rerouted a wire directly from the Convert pin (control board) to the A/N conv  and everything finally was running.  Explanation  (sort of)  When I ran the Det log “test as you build” for the first time, I had a very noisy trace and remarked that the noise was vanishing when I was putting my hand on the Data cable. So I decided to use a piece of Ethernet cable (cat 6, twisted and screened pairs) to route the 4 data wires from the command board to the A/D board. No more noise, the test was running fine.  But with a “real” signal, the twisted pair on the “Convert” data wire was messing a lot. I think the culprit should be distributed capacitance of the cable  Now the game is to find which cable must be shielded and which don’t appreciate this treatment J  What a waste of time for such a silly problem :- ))))  Thanks you both for your help. Simon, you were close to the right answer, and it helps me a lot, but the reality was a little bit more vicious.73’Marc      De : spectrumanalyzer@yahoogroups.com [mailto:spectrumanalyzer@yahoogroups.com] De la part de SFC
Envoyé : dimanche 15 janvier 2012 18:05
À : spectrumanalyzer@yahoogroups.com
Objet : Re: [spectrumanalyzer] problem testing local oscillator  


Hi, Could be a bad solder joint of pin 6 of the adc chip. Check if you have zero resistants between pin 6 (on the chip not the pad )and p3-d7. Simon  ----- Original Message ----- From: [url=mailto:Marc.olanie@... t=_blank]marc olanié[/url] To: [url=mailto:spectrumanalyzer@yahoogroups.com t=_blank]spectrumanalyzer@yahoogroups.com[/url] Sent: Sunday, January 15, 2012 4:06 PMSubject: RE: [spectrumanalyzer] problem testing local oscillator    >I think something is wrong with your a/d cnv. Magnitude voltage of 2.45V should give a value above 30,000.Thanks SimonFollowing your idea, I tried to test the A/D again… with the same results. But If I put a probe (VTVM,O’scope) on pin 6 (convert) of the AD7685, the measure seems to be right : trace on top of the screen with a 10 dBm signal on J1/Slim-AD8306, and trace à -110 dB with a 50 Ohms on  J1.If I try to put the probe directly on Control Board P3-D7 (the other tip of the Convert wire), nothing appends… a 10 dBm signal delivers a 2.34V MagV as expected, but magdata is still hooked around 15 000 and trace is located in the middle of the screen. I hate things that are “half working”.. :- /Marc



Post Reply