HOME / KC901 serial VNA

KC901V spectrum analyser 30kHz RBW is non-symetrical

By VK3NH •  2017-01-06 18:25:18 •  1351 clicks
I had posted this issue I have with my KC901 before but the results were not displayed in this manner which emphasizes the difference. As I said previously, all the other resolution bandwidths 10, 3 & 1kHz are symetrical around the centre frequency as expected. For 30kHz RBW setting, the side of the response that rolls away quicker is only about 6.9kHz at its 3dB point. The measured 3dB bandwidth of the filter, ignoring that it is offset, is about 24.5kHz.

For this test I used my R&S CMT54 test set to provide a 500MHz carrier at 0dBm and then observed it with the widest 30kHz bandwidth. The KC901V span is set to 50kHz. This effectively traces the IF bandwidth response which I then stored with the hold feature. I then pressed the Lo Hi/Low button which effectively flips the IF response. The markers are set at the +/- 15kHz points so you can see that the response is way out either way you look at it. The top of the response is also about 0.8 dBm not 0dBm.

I expect the issue is a hardware one requiring a change of the pre-selection filter but there might be another cause and I would be happy to know what solutions might be available.

regards Noel

difference  previously  resolution  frequency  expected 

5 Reply | until2018-12-13 13:01Add a reply
    VK3NH  2017-03-03#1 -Reply

    It just means you are off centre from any specified channel when you go to field strength mode. Knowing that there is an error of course means you can correct the selected frequency up or down depending whether you have selected Lo Hi or Lo low.
    I also use the zero span mode for observing pulsed emissions of things like garage remotes and by varying the sweep speed you can read the code.
    The 30kHz bandwidth is a hardware limitation. All the other bandwidths being FFT or software defined are centred correctly.

    Andy  2017-03-02#2 -Reply

    I can confirm that I get exactly the same lopsided response with the 30 KHz setting on my KC901V. However, is there a reason to use a larger RBW value in the spectrum analyzer mode? In what situations would I want to use 30 KHz instead of, say, 1 KHz?

    VK3NH  2017-01-07#3 -Reply

    Your manual says the IF is 109.65MHz. From the results it looks like you are using the PDI part number 3380 at https://www.relayspec.com/specs/027351_surf_mt_crystal_filters.pdf

    A wider filter is shown at https://4donline.ihs.com/images/VipMasterCap/Capacitor/CTSC/CTSCS00155/CTSCS00155-1.pdf?hkey=EF798316E3902B6ED9A73243A3159BB0

    but I guess the case size and all the other specs make it an unlikely replacement.

    I can live with the offset if it can't be changed though wider would be better.

    regards Noel

    VK3NH  2017-01-07#4 -Reply

    Thanks for explaining that. As you say it allows good support for the smaller bandwidths established by software. Is it possible to select on test a better filter crystal shape or are they all much the same? Input and output matching could also affect the shape but I suppose you have optimised these already.

    Someone else has posted what seems to be a comparison between the 30kHz bandwidth and 5 kHz bandwidth here and it also seems lop sided like mine though maybe a little better. https://www.facebook.com/kc901s/photos/a.448457205302121.1073741828.442923015855540/603729683108205/?type=3

    BG8NPK  2017-01-07#5 -Reply


    At first our KC901V's digital IF processing just support 10K RBW at the third IF filter. When we chose the 30K RBW it will just filtered by the first IF filter. It is a Cristal filter with a unflap shape and the bandwidth about the 30KHz.

    Actually we just design the KC901V not support the 30K RBW selection but think of some situation that we need this function so we added it.


LoginPublish or Still no account?Register