Out of band response

[PA3GSB]

Hi group,

Ed, GM3SBC found a severe problem for the radioberry; some out of band response problems.

Because the radioberry and the hermeslite share the same frontend i am curious of these problems are also valid for the hermeslite configuration.

When putting a signal generator using -70dBm @11.86MHz or @14.68MHz a response is visible at 14.000 MHz.

I did not found this issue because iam using a magnetic loop; giving nice filtering.

73 Johan PA3GSB

[Steve Haynal]

Hi Johan,

I don't see this on the Hermes-Lite 2.0. Below is a screenshot of Quisk operating FDX with two receivers. The TX signal detected by the RX is stronger than a -70 dBm signal generator yet I don't see anything at 14MHz. There is a small spur at 14.02 MHz, but this is only 15 dBm above the noise floor, not 50 like in your screenshot. Plus, this is more likely an artifact of the test using the same AD9866 for TX and RX and with a more powerful 1W TX signal into a dummy load. 

When did you fork the firmware? Maybe there is a difference in the DSP now. If I remember correctly, you are using the AD9866 to double the clock frequency. Maybe that has an effect. Also, do you have the AD9866 LPF on or off? The software is also different. Maybe some there.

73,

Steve

KF7O

[PA3GSB]

Hi Steve,

Thanks for your reaction. The problem of these out of band responses is solved.

The setting of the AD9866 was wrong. I did not double the frequency as required but 4 times!

Now it looks fine... i am seeing the same responses as you measured (https://github.com/softerhardware/Hermes-Lite/wiki/Clock-Phase-Noise).

73,

Johan

PA3GSB

Op zaterdag 7 april 2018 06:36:23 UTC+2 schreef Steve Haynal:

[Steve Haynal]

Hi Johan,

Thanks for the update. This is good to know.

73,

Steve

KF7O