ad: AbAuRe-1

A technical point about analog Vs SDR radios

Discussion in 'General Technical Questions and Answers' started by KC6ETE, Nov 24, 2021.

ad: L-HROutlet
ad: l-rl
ad: L-MFJ
ad: L-Geochron
ad: Left-2
ad: Left-3
ad: abrind-2
  1. KC6ETE

    KC6ETE Ham Member QRZ Page

    Let me start with the good old NXP SA636 FM receiver chip.
    The RSSI output on that chip can vary by as much as 3-1 between "identical" chips and be in spec.
    Through most of it's range, the unit to unit variation is "only" 2-1, but there are points where it's about 3-1.
    It's got other issues too, but it's been an "old faithful" for FM RX designs for a lot of years.

    Looking at SDR implementations like the Tayloe detector, it seems like the amplitude of the recovered signal should be very consistent from unit to unit, dependent on the component tolerances of course, but I see far less variation there..

    Am I seeing it correctly?
  2. G8FXC

    G8FXC XML Subscriber QRZ Page

    Once you have the signal in the digital domain, one could reasonably expect that it would be constant across devices, but there is always some analogue processing in front of the digital processing block, so I don't think that you can expect identical performance across multiple examples of any real-world SDR radio... If it is a direct sampling design, then you need a decent band-pass filter in front of the A/D converter; if it is a hybrid SDR like most current Yaesus, then there is a lot of analogue circuitry to traverse before you get into the digital domain.

    Martin (G8FXC)
  3. K8XG

    K8XG XML Subscriber QRZ Page

    It is my understanding that the Yaesu SDR path is only for the Scope and it's functions. The radio you listen to is a superb SuperHet design.
  4. KC6ETE

    KC6ETE Ham Member QRZ Page

    I'm not looking for info on specific radios, this is a design question. Assume that I can keep the hardware between the antenna and ADC to a minimum, use 1% or even 0.1% resistors and 1% caps etc.
  5. G8FXC

    G8FXC XML Subscriber QRZ Page

    That is not true. The current Yaesu hybrids are analogue superhets converting input to a fixed frequency IF - about 9MHz, I believe - and then converting to digital with all the primary filtering and demodulation taking place in the digital domain. Think in terms of a fixed frequency SDR fronted up with a transverter... They also include a fairly low performance direct sampling SDR that has the bandwidth to drive the panadaptor.

    Martin (G8FXC)
    K8XG likes this.
  6. G8FXC

    G8FXC XML Subscriber QRZ Page

    I don't think you can avoid a moderate amount of electronics between the antenna and the ADC. Assuming that you want a sensitivity measured in micro-volts, you are going to need a fair amount of amplification in front of the ADC. If you are going for a direct sampling design, then you need a fairly wide band-pass filter which is going to include inductances - not easy to mass produce to very close tolerances. If you go for a hybrid design like Yaesu, then you have a large part of a regular superhet in front of the ADC...

    Martin (G8FXC)
  7. KC6ETE

    KC6ETE Ham Member QRZ Page

    You're making a ton of assumptions which don't hold true. I have an ADC that gives significantly positive SNR when tested with 0.1Vpp through 100dB of attenuation (1uV)
    I'm not looking to go DC to daylight on the front end, FAR from it. I'm looking ONLY at the 475kHz band. Between the antenna and Tayloe detector is a very low noise amplifier which I will set for the least amount of gain that I can get away with. It's very tolerant of loud out of band signals.

    But I still don't have an answer to my question.
  8. KA5IPF

    KA5IPF Ham Member QRZ Page

    Is the signal you are seeing at .1 uV real or is it leakage? Very few inline step attenuators are designed to attenuate down that far and what you are seeing is probably leakage
  9. AI5DH

    AI5DH Ham Member

    Define RSSI. The problem is you cannot define it, there is none, it is not a standard or specification. Manufacturers make it whatever they want it to be. I work on the commercial side, wireless, and P25. All the big players like Nokia, Erricson, and to a lesser degree small players like Motorola and L3Harris all use RSSI monitoring to measure RX quality issues. Not to measure signal levels, but to measure antenna and transmission quality. All those technologies use dedicated multiple RX antennas. If there is an antenna issue or PIM, RSSI levels are going to detect it by a variance in the RX Delta. Operating companies watch RSSI like a hawk and have all kinds of alarms, bells, and whistles go off when a problem is detected.

    Of all the manufacturers, only L3Harris defines their RSSI to dBm with a dynamic range of only 70 dB of -60 to -130 dBm. There lies the problem, how is it scaled and over what dynamic range. An ADC has a limited dynamic range just like analog. If you really want to know signal level will take a logarithmic amplifier like an AD-8307 with a 92 dB dynamic range. Depends on what you are after and how you use it.
    SWL37632 likes this.
  10. KC6ETE

    KC6ETE Ham Member QRZ Page

    NOT my first rodeo... Five BNC inline 20dB attenuators. Disconnect, signal gone. Change frequency, signal moves appropriately. Change levels, signal moves approptiately.
    I have a few TEK 130dB stepped units that will do this, but I trust the 20dB inline ones.

Share This Page