Post Go back to editing

HMC558 Up-Converter Spurs

Hi All.

We're using a HMC558LC3B in an up-converter application and seeing, frankly, horrendous LO+2IF spurs.

IF = 950MHz, LO 9750MHz, RF 10,700MHz spur is at 11,650MHz and only -20 to -25dBc. Our spec is -50dBc, so as you might imagine, we are up the creek...

Input drive level is about -6dBm, so well backed-off.

The IF amplifier is 20dB backed off from compression at this level so the harmonic level into the IF port should be negligible. We've replaced the mixer with a fresh device (my initial thoughts were that static damage has blown one half of the diode/FET pairs and the balance was upset) but only seen an improvement of 3-5dB. (Consequently, the 2LO-RF improved 10dB - so are the mixers very sensitive to process variation? Do the MxN spurs vary this much, device to device?)

Anyway, so...

Q1. Is this performance typical of this device if used in an up-converter application?

If not then...

Q2. What are the potential causes of high LO+2IF spurs, apart from high levels of 2IF entering the mixer in the first place?

- input match?

- LO drive sin or square? We are somewhere in the middle, aiming for P1dB of the LO driver amp.

- Something else?

Thanks in advance.


  • Hi,

    The spur rejection we measured at LO+2IF is about 45dBc. The potential causes for spur may come from IF amplifier. If the output of the amplifier is not clean enough and contains high level 2IF, consequently the LO+2IF spur will be higher than normal.

    2LO-RF which equals to LO-IF, is another expected sideband main tone. It's a little confusing here that you mentioned it is improved by 10dB. Did you mean 2LO-IF?

    My suggestion is that input a clean IF signal by signal generator and test the spur level again. Double check the IF amplifier output spectrum at 2IF. If 2IF signal is high, use a filter to reject it.


  • Hello, and thank you for the reply.

    To clarify, when we changed the mixer, the LO+2IF improved by 3-5dB. Another spur (that I mis-identified as 2LO-RF) changed by 10dB - so I asked, do the mixers' MxN spur performance change that much device to device???

    So is 45dBc about the best we can expect from this mixer in an up-conversion application?

    I will check that the 2IF has been characterised properly, however we do have a LPF and the amplifier is backed off by 20dB, so there *shouldn't* be any strong 2IF component.

    I'll keep you posted.

  • Hi,

    According to your issue, we measured the mixer with two IF input tones, at 950MHz and 1.9GHz respectively. The input  signal at 1.9GHz is about 20dB lower than the 950MHz. It turns out that the spur level at LO-2IF is only around 20dBc lower than main tone.

    Then we added a bandpass filter after two inputs. 1.9GHz signal is suppressed 60dB lower than 950MHz signal. As a result, the spur level improves and about 45dBc lower than main tone.

    The mixer itself should be good enough for spur rejection. So you may need to be careful with the IF amplifier performance.

    You can also vary the LO power level to improve the spur rejection. It might help to improve it to 50dBc. But do not exceed the maximum power level datasheet indicates or the part will be damaged. After tuning the LO to the desired range, make sure the IP3 level is still good enough.

    In addition, this is the last time to buy HMC558 so this part is not recommended for new design. A new part can be recommended if need.


  • This question has been assumed as answered either offline via email or with a multi-part answer. This question has now been closed out. If you have an inquiry related to this topic please post a new question in the applicable product forum.

    Thank you,
    EZ Admin