Measuring IP3 with cheapskate equipment

Discussion in 'Homebrew and Kit Projects' started by N5HXR, Jul 19, 2021.

ad: L-HROutlet
ad: l-rl
ad: Left-2
ad: Left-3
ad: MessiPaoloni-1
ad: L-Geochron
ad: L-MFJ
ad: HRDLLC-2
ad: abrind-2
  1. SM0AOM

    SM0AOM Ham Member QRZ Page

    Really hard to tell where the problems may come from.

    My experience with FFT based spectrum analysers with not-too-large vertical resolution is that the may show very strange behaviour depending how near the quantisation noise floor the distortion products turn out to be.

    The products may very well be caused either by intermodulation distorsion in the generator, the FFT oscilloscope or both...

    A way to find out is to use a conventional receiver having some RF selectivity and feed the output signal from the generator into the receiver through a substantial attenuator, 20 dB or so. If the IM products are gone, they come from the oscilloscope, if they still are there, they are from the generator.

    N5HXR likes this.
  2. N5HXR

    N5HXR Ham Member QRZ Page

    Since the ADE-1+ doesn't have any gain, there's no P1dB, right? I don't see it in the datasheet, anyway... But it does have an IP3 specification (+15dBm).

    So if people were telling me before that putting a BGA2866 with IIP3 of -4dBm in front of the mixer clobbers its dynamic range... don't I have to measure IP3 to compare with other preamps based on that logic? Or is there some P1dB analog for the mixer so I can tell if the amplifier I put in front of it is causing more problems than it helps?

    Maybe some rule of thumb (I see the BGA2866 has IIP3 of -4dBm, and P1dB of 4dBm, so look for a P1dB that's >10dB higher than the IP3 of my mixer)?
  3. N5HXR

    N5HXR Ham Member QRZ Page

    Thanks for this info -- do you mean I could test this with, say, my FT-891 sitting nearby? Would I just tune to the frequencies where I'd expect the IMD products to be and listen?
  4. SM0AOM

    SM0AOM Ham Member QRZ Page

    Just use an attenuator to swamp out the impedance mismatch at the receiver input.
    You should of course listen both at the tone frequencies and the IM product frequencies and note the difference.
    N5HXR likes this.
  5. K7JEM

    K7JEM Ham Member QRZ Page

    The useable signal is going to be way, way below IP3. If it has that as a spec, then input levels need to be kept relatively low. That mixer is designed to be a relatively low level mixer, not suitable for high level signals. I would think the maximum signal input would be something around -30dBm, just guessing.
  6. N5HXR

    N5HXR Ham Member QRZ Page

    I'm trying to use the information I've found in my studying. E.g., EMRFD suggests putting an RF amplifier stage in front of the mixer in a DC receiver to reduce LO radiation, and a handful of other reasons. I'd like to try it, and see if I can feel the difference.

    I tried making one with one of these MMICs in front, and several people said that they're a bad choice because their IIP3 being so much lower than the mixer would make for a bad receiver on a band with any strong signals present.

    I want to build a JFET preamplifier as an alternative, as was suggested, but I also want to proceed somewhat thoughtfully, and measure the differences. Hence, the "measuring IP3" rabbit trail. I totally understand that there will be small signals going into these things, and that these are somewhat artificial figures of merit.

    Maybe this is not a useful rabbit trail, so if I'm groping in the wrong direction, let me know.
  7. SM0AOM

    SM0AOM Ham Member QRZ Page

    It is here the "intermodulation cascade" enters.
    Analysing the various contributions from the stages using AppCAD makes it possible to determine which are contributing most to both noise and IMD.

    If you put an amplifier with an input referred IP3 of -4 dBm in front of a mixer with an input referred IP3 of +15 dBm, the IM performance will still almost entirely be governed by the amplifier.

    A JFET, say, the J310 will give you an IIP3 in the region of 0 dBm, but the difference will be very difficult to spot, especially in front of a direct conversion receiver.

  8. N5HXR

    N5HXR Ham Member QRZ Page

    OK, this sounds like real science is happening... I attached my siggen to my FT891 with attenuation. I can clearly hear the 7.0 and 7.1MHz injected signals. When I tune to 6.9 or 7.2MHz, I hear nothing at all, max RF gain and max AF gain.

    So I guess that means it's not the siggen, and probably the oscilloscope?
  9. SM0AOM

    SM0AOM Ham Member QRZ Page

    You need to use sufficient signal levels so the IM products are
    above the noise floor. Say that you set the level at "S9+10" for the tones. Then the IM signals should be at S3 or so if they come from the generator.
  10. N5HXR

    N5HXR Ham Member QRZ Page

    So in your opinion, for my first complete direct conversion transceiver project, do I need to care about this or not? Wes Hayward, et al, in EMRFD use a grounded gate JFET in front of their TUF mixer in their modular DC receiver design, and seemed quite happy with the various compromises, pros, and cons. I'm trying to avoid straight-up copying an existing schematic so I know I'm really learning each of the pieces along the way... but I want to lean appropriately on the wisdom of the ages too :).

Share This Page