BATC Forum

Gday,

I build succesfully the digilite 5.7 print.
led is flashing as it should be flashing
i set the r23 and r16 at the right value with the scoop
After that i tuned the lo carrier at the lowest level with the bias pot.
i get a picture on my dreambox
but when i connect the spectrum analyser i get this picture with the signal at the right frequency but i have also a very width signal started at +- 100mhz until 1000 mhz only 20 db weaker as the digital carrier.
Even without the lo i have the same width signal coming from the digilite print.
I checked the parts and find no problems.
Does anyone give me a clue how i cam solve this unwanted signal?

thanks in advance,

arie

Can you post photos of the signal with and without the LO. That will give us a better idea.

Brian

ok, here the picture spec connected to digilite board LO connected
http://imageshack.us/photo/my-images/71 ... rum2y.jpg/

here is the digilite connected to spec. without LO connected

http://imageshack.us/photo/my-images/4/spectrum1n.jpg/

when i put my finger on c29 the interference disapeared but also the digital carrier falls off

thanks in advance for your effort

Have a look at the thread below on noise and decoupling:

http://www.batc.org.uk/forum/viewtopic.php?f=71&t=1002

Also check this, but I think it's very unlikely:

http://www.batc.org.uk/forum/viewtopic.php?f=71&t=1097

Even with lots of decoupling, I still get some noise increasing as the frequency decreases. It probably comes from the I and Q signals switching. Try selecting 'No Mod' test mode in DigiLite Config (with DL Transmit running) and see what happens on the analyser.

Brian

If you can't figure out the cause of the problem from the symptoms go back to the basics. Check all 3 power supplies for correct voltage and for noise. Inspect the board for bad connections, especially decoupling caps.

At 23 cm shielding on coax isn't 100% effective. Keep the ins and outs short and separate. See if moving cables around changes the signal.

Check your test setup, you tried it with the L.O. off, turn off the DigiLite too. More than once I've chased problems generated outside the device under test.
We had to put an attenuator before the spectrum analyzer to prevent the false signals caused by overdriving the input.

This post has pictures of our spectrum analyzer signals.
http://www.batc.org.uk/forum/viewtopic.php?f=71&t=1232

John