Choosing a mixer for HF transceiver

Hello guys,
I'm trying to design an analog portable single side band receiver (ham radio transceiver in future).

I am looking for a good matching components.
Initially I started with SA612 mixers but I was disappointed by their noise level.
So I tried AD831 which had almost no gain while dissipating tonnes of heat. I understood I was foolish underestimating their power consumption and made many mistakes.

I have drawn this block diagram:


It uses Si5351 as VFO. First mixer VFO is set above receive frequency. RF BPF is up to 2MHz wide. IF 8MHz BPF is 3 KHz wide. LO can move up & down to be able to receive USB or LSB.

Now I am trying to build something which will be power efficient, compact.
So I picked LT5526 as double balanced mixers.
But I am looking at LT5560 and LTC5562.
LTC5562 seems having better characteristics and claims to have ESD protecting which sounds important for HF applications having an external antenna. The only caveat it has some note about external matching for lower frequencies which is my case.

For IF amplifier I picked AD8338. It has attracted me by presence of AGC and 80dB of dynamic range. But I struggle a little bit with proper configuration for this chip. I had to manually set AGC setpoint. Also I think I made poor decisions regarding my PCB design, I guess it starts self excitation. Many people recommend 4 layers board but it is an overkill for ham radio. I think it is doable with 2 layers. Also I was thinking to tap to AGC loop and use it for S-meter.

Next, I am thinking about some LNA after RF bandpass filters but unsure about it. In different designs I see LNAs and attenuators there. If signal will be stong it will overload LNA and small signal might be lost. From the other side, ham radio signals are usually very weak unlike broadcast radio stations so it worth to have it.

For audio amplifier I used SSM2211 and LM358 as preamp/filter. It is the only one which has option to have single ended output which is one my requirements.

TBH, I am not RF engineer and radio is my hobby so I am very welcome to any advice.



typo fix
[edited by: zoonman at 4:21 AM (GMT 0) on 21 Jan 2020]
Parents
  • Hello,

    I think nice design, well described. Did you check your frequency plan? I am not sure, but if I understand the solution correctly, 1st IF (8MHz) is in the input RF band. Or is there a gap in the RF received band? If no, you should revise the frequency plan and use e.g. 1st IF above the RF band. You have to decide about the noise figure. Usually the noise figure is requested ~3dB (or less) for the VHF, UHF bands but I don't have experience with the HF band, I assume the noise figure can be higher because of higher noise background. To reach better NF, LNA shall be placed in front of mixer of course at the expense of linearity. Define what you want to reach (NF, IIP3) and calculated level diagram gives you answers whether use or not to use gain blocks somewhere in the RF chain. 

  • 0
    •  Analog Employees 
    on Jan 21, 2020 7:36 PM over 1 year ago in reply to viktor_vojta

    I read SA612 datasheet and noticed that it has just 5dB noise figure which is smaller than all ADI mixers. Now I am very confused.

  • 0
    •  Analog Employees 
    on Feb 1, 2020 7:10 AM 11 months ago in reply to viktor_vojta

    I can grab something like FT-891 Specifications:
    Sensitivity for SSB (2.4 kHz, 10 dB S+N/N): 0.16uV which translates to -123 dBm.

    I know it is a good and popular transceiver.
    My goal is to be capable to recieve S1 signal which is -111dBm. If we have 10dB for signal to noise ratio that gives me -121dBm sensitivity baseline.

    Let me workout signal chain backwards.

    If I want to have S1 signal to be 1W audio it means 30dBm.
    SSM2211 allows up to 40 dBm gain. Let's stick to 30 dB gain here to ensure stability of audio amplifier. Please, forgive me my newbie assumptions.
    This gives me 0 dBm at the LF amp input. I was planning to add tiny pre-amp + filter on LM358 which can give 10 db of gain or more.
    So at this point for S1 signal we will have -10 dBm at the mixer output.
    LTC5562 has 1dB conversion gain which I totally disregard due to losses in filters and bad impedance matching.
    AD8338 has 80dB of gain which lowers our input to -90dB.
    Basically this threshold dictates our sensitivity and gain requirements for pre-amp. So far it has to be around +30dB. If I will play around with audio amplifier I can lower it to 20 dB. Again, it is ballpark estimations.
    TBH, I don't know which part will introduce more noise and where is better to get maximum of the gain.

    Here I drew some schematics so you can get idea of tranceiver design.

    First mixer:

    And the second

    I have concerns about getting audio out and proper audio injection but this idea seems ok to me. I got some inspiration from phase detector schematic inside LTC5562 manual.

  • Philipp,

    I have looked on FT-891 spec, it quite "brief". Thus I found something more detailed e.g. here: https://static.dxengineering.com/global/images/technicalarticles/ysu-ft-891_sn.pdf

    NF is mentioned there NF=7dB (LNA is ON) and I calculated from measurement IIP3=-1dBm (LNA is ON).  Gain you know, I think, from your prev. post. This 3 numbers are important for you. Modify NF, IIP3. Gain is fixed. And calculate level diagram. IIP3 you calculate of course without audio amp. Use e.g. ADIsimRF in order to build RX chain, it is quite useful.

    Thank you for pictures, but I can not do a review of whole schematic. And I mean this EZone should be oriented on the AD devices, we are now quite far away with this topic, we are discussion general theory of receivers. 

  • 0
    •  Analog Employees 
    on Feb 8, 2020 4:52 PM 11 months ago in reply to viktor_vojta

    I had to jump over several humps to get ADISumRF running. It doesn't work on MacOS. When I finally got it working I was unable to find mixers. It looks slightly abandoned to my taste.

    I think I need to learn more theory of receivers and all these parameters.

  • Aha, OK, I did not know you use MacOS, it gets complicated a bit. And I did know, that ADISimRF does not support LTC5562. In the past I tried use ADISimRF, but I did not match my expectations. Thus I went back to calculatin (evaluation) of level diagram in the MS excel sheet. In principle it is very simple to write down formulas for: total gain, NF, IP3. For each component you need to know also these 3 mentioned parameters: gain, input IP3 and NF, so the calculation can be made in any spreadsheet.

    Or stay with ADISimRF, it offers more calculations, but in the first step, you will not need them. If you don't find the component in the AD list supported components, nevermind, write corresponding values from the datasheet. 

    Very simple calculation find also here: www.qorvo.com/.../cascade-calculator

  • 0
    •  Analog Employees 
    on Mar 3, 2020 2:08 AM 10 months ago in reply to viktor_vojta

    I tried to calculate some metrics:

    I tried guesstimate certain values from this thread https://ez.analog.com/amplifiers/f/q-a/14752/ad8338-compression

    I am not sure if it gives me anything. Total gain is just sum of gain on every stage.

    I don't understand how the noise figure works. I really need to learn this.

    Also noise floor is given in dBm/Hz for mixer, and in nV/√Hz for amplifiers. I don't know how to relate these values and I have to factor them in.

    P1dB for AD8338 is not specified in datasheet.

    If I had the math behind I can easily build online calculator for those things.

Reply Children