RSP SPECTRUM ANALISER VER. 0.9 Vs. VER. 1.0

[glovisol]

USING BOTH VERSIONS TO BEST ADVANTAGE

It seems to me that Ver. 1.0 does not supersede Ver. 0.9, but supplements it for the wider span ranges. The main point is that (unless I am mistaken as such early stage!) Ver. 1.0 does not see down to zero frequency, while Ver. 0.9 does quite well.

So for low frequency work, say zero to 5 MHz, Ver. 0.9 looks better to me. In the 5 to 10 MHz range the two applications overlap and above 10 MHz Ver. 1.0 is what should be used.

[SteveAndrew]

USING BOTH VERSIONS TO BEST ADVANTAGE

It seems to me that Ver. 1.0 does not supersede Ver. 0.9, but supplements it for the wider span ranges. The main point is that (unless I am mistaken as such early stage!) Ver. 1.0 does not see down to zero frequency, while Ver. 0.9 does quite well.

So for low frequency work, say zero to 5 MHz, Ver. 0.9 looks better to me. In the 5 to 10 MHz range the two applications overlap and above 10 MHz Ver. 1.0 is what should be used.

HI Glovisol
Interesting. I would have thought that having a non-zero IF, with choices of 2.048MHz or 450kHz would have given far better performance than the zero-IF mode that I used in V0.9 - Do you have some screen-shots that illustrates the differences you are looking at.

Steve

[glovisol]

Hi Steve,

I have been using VER. 1.0a extensively in these days, developing the 88-108 MHz Band-Stop filter and it has been working very well, apart from having some patience when having to re-start the sweep. So I am very happy with it at the higher frequencies. In the next few days I shall switch my attention to LF: the first tests I did there showed that it was impossible to "see" below 2 MHz, but it was quick, as I did not have the time to go in depth. Will do more testing in the next few days & advise.

Kind regards,

gianfranco

[glovisol]

Hi Steve,

My apologies. I managed to go down to 100 KHz and so far so good. My problem was I could not and cannot see the zero: the system manipulates the frequencies so that zero is forever hidden. But I am still learning...

[SteveAndrew]

Hi Steve,

My apologies. I managed to go down to 100 KHz and so far so good. My problem was I could not and cannot see the zero: the system manipulates the frequencies so that zero is forever hidden. But I am still learning...

Hi Gianfranco

I'll check the code, I think there is a minimum frequency that it's not possible to tune below.

Steve

[glovisol]

Hi Steve,

I have devoted some attention to the low frequency side, as I needed noise source calibration at low frequency and am uploading two screens, one of which down to 50 KHz. Harmonic lines of the noise generator are visible here.

Also averaging works well in low spans!

Gianfranco

[glovisol]

Hi Steve,

Here below some general findings. Please forgive me if I did something wrong, but one must try to get acquainetd with this new machine, after all...

1) In the tested range 0.5 to 15 MHz the system seems to saturate at - 44 dBm, so care should be taken to keep input signals always below this level. This should be a limitation of the RSP-1A.

2) The DISPLAY SCALING works very well from +20 to -40 dBm, but below levels shown are wrong. I tried this with a tests noise level of -64 dBm. Reading is spot on at - 64 dBm from +20 to -40 dbm of DISPLAY SCALING, but if you set DISPLAY SCALING at -50 dBm or below readings shown are wrong.

[SteveAndrew]

Hi Steve,

Here below some general findings. Please forgive me if I did something wrong, but one must try to get acquainetd with this new machine, after all...

1) In the tested range 0.5 to 15 MHz the system seems to saturate at - 44 dBm, so care should be taken to keep input signals always below this level. This should be a limitation of the RSP-1A.

2) The DISPLAY SCALING works very well from +20 to -40 dBm, but below levels shown are wrong. I tried this with a tests noise level of -64 dBm. Reading is spot on at - 64 dBm from +20 to -40 dbm of DISPLAY SCALING, but if you set DISPLAY SCALING at -50 dBm or below readings shown are wrong.

Hi Gianfranco

Thank you for the detailed information. I'll try your settings and see if I can get the same results. There is a know problem with reference levels appearing to change, or not appearing to match the level set. This has now been fixed so I'll test with both the current release and the debug version I'm currently working on.

Steve

[SteveAndrew]

Hi Gianfranco

I don't have a calibrated noise source available so I used my sig gen instead.

At 10MHz centre frequency and at various spans from 20kHz to 10MHz, with signal levels varied from 0 to -120dBm, and with the reference level set appropriately for the input signal level, I saw no sign of saturation at any level. At all times the level displayed was spot on.

I'm not sure what is causing your results. Is it possible for you to post a full screen-shot of the analyser, showing settings and the saturated signal?, also a similar screen shot showing the shift on level when setting the reference level at -50dBm or below would be good.

Thank you in advance

Steve

[glovisol]

Hi Steve,

You have given me very good news. I was hoping it was something on my side, because saturation at -40 dBm would cause very many bad measurement problems. I shall look again and provide screens unless I find what was wrong in the first place.

Perhaps there is a difference between receiving single frequency signals and noise signals. In this last case then the solution would be not to use a noise generator, but a tracking generator for critical measurements. This could happen because with noise the RSP-1A sees the input level signal over a great bandwidth AT THE SAME TIME, while with a tracking generator it sees only one signal for every time unit.

Best regards,

Gianfranco