Noise figure, preamps, and another random question

Discussion in 'General Technical Questions and Answers' started by KK4YWN, Aug 12, 2017.

ad: L-HROutlet
ad: l-rl
ad: l-gcopper
ad: Subscribe
ad: L-rfparts
  1. KK4YWN

    KK4YWN Ham Member QRZ Page

    The transverter I use has a noise figure of .8db. It has its own preamp but includes instructions for bypassing it in the event that a mast-mount unit is used.

    The built-in preamp is an ERA6 mmic with an advertised gain of 11db and 4db NF.

    So how are they able to say .8db noise figure? I don't understand.

    The lna I have is capable of 23.5 db gain and .75 db noise figure. I don't need that much gain but I am curious to know what happens to overall system noise figure if I added it and bypassed the built-in unit.

    Lastly, I'm looking at a 500w pallet amp. Its class B push pull. Is that linear enough? I have vague memories of studying this stuff years ago. I dont want to spend a lot of time scrounging an amplifier only to find out its dirty.

    AF6LJ likes this.
  2. AF6LJ

    AF6LJ Premium Subscriber QRZ Page

    Your Class B pallet amp should be good enough, as for the noise figure business, I think I know but I would rather defer to someone who truly does know.
  3. WB2WIK

    WB2WIK Premium Subscriber QRZ Page

    The MMIC probably doesn't have a 0.8 dB NF.

    The preamp likely does, but 23.5 dB gain (not uncommon) is way more than you need and will often tend to saturate following stages, degrading performance.

    I remember speaking with Russ, W1VD, who runs Advanced Receiver Research (ARR), a big provider of low-noise VHF-UHF preamps, about this a long time ago (maybe like 30 years ago, but not all that much has changed). He tunes preamp for best NF (50 Ohms) and unfortunately that's also the tuning point that produces a lot of gain. You don't need 23.5 dB gain with a 0.5 dB NF preamp to override a 4 dB NF following stage (only need about 7 dB), so it can be best to throw some of that gain away with an attenuator after the preamp and before the next stage.

    I do that, and receiver performance is much better than using all the gain the low noise preamp provides.

    Having used ANFIs before (automatic noise figure measuring instruments), I've noted you can tune a preamp for best NF, most gain, or best match to 50 Ohms (noise match, gain match or impedance match). Those are three different tuning points.:p Most commercial units are tuned for lowest NF and end up with too much gain.
    AF6LJ likes this.
  4. KM1H

    KM1H Subscriber QRZ Page

    The last time I talked to W1VD he still called himself Jay
    AF6LJ likes this.
  5. WB2WIK

    WB2WIK Premium Subscriber QRZ Page

    You're right, my mistake!

    I haven't talked with him in a long time.

    At Dayton in the mid-70s he was dating Rosalie, and they were a couple.

    "Time flies."
    AF6LJ likes this.
  6. N1OOQ

    N1OOQ Ham Member QRZ Page

    What bands? You usually only need sub-1 dB NF for vhf/uhf.
    AF6LJ likes this.
  7. W6RZ

    W6RZ Ham Member QRZ Page

    Google "cascaded noise figure calculator".

    Also, there's an ATF 34143 GaAsFET in front of the ERA6.
    K8AI, AF6LJ and AI3V like this.
  8. AI3V

    AI3V Ham Member QRZ Page

    A couple points:

    To tune a amplifier for best nf you don't need any equipment other than your ears, and a steady, weak signal. Weak enough so there is more static than signal.

    Then you simply tune the preamplifier (which should be in circuit and connected as you are going to use it on air) for the clearest signal. You will be as good as ANY other method.

    NEVER tune with a strong signal for a simple peak on a meter. (Unless of course you want maximum gain).


    23db is a LOT of preamplifier gain, unless you truly live out in the sticks, I predict you will be learning a lot about bandpass filters and preselectors. Expect intermodulation issues with your combo.

    AF6LJ likes this.
  9. AI3V

    AI3V Ham Member QRZ Page

    Another issue universally overlooked in ham radio preamps is wideband noise.

    I once built a series of preamps for 432mhz using Mitsubishi mgf1302 gasfets.

    I happened to have access to some hp wideband power meters.

    Integrated over the frequency range of 10 MHz to 12 GHz, the preamps , with their inputs terminated with a 50 ohm load, outputted -30dbm of noise.

    It took a day of head scratching, and another quarter million dollars of test gear (thanks intelsat!)
    Before I convinced myself it wasn't some parasitic oscillation, just plain old KTB noise, amplified by 20 or so DB.

    Think about hitting the front end of your radio with that kind of RX signal.

    Bandpass filters are your friend.

    K6CLS and AF6LJ like this.
  10. KM1H

    KM1H Subscriber QRZ Page

    Sam Harris, W1FZJ, taught me the benefit of bandpass filters in the 1960's. I currently have all homebrew lumped LC filter on 6 and cavity filters above that for receive which takes care of all the crud from the UHF TV plus several commercial FM stations, and a host of other 2 way FM crud on the hilltop about 2 miles away. Plus the 150' tower across the street with the towns police, fire, highway dept, and several other small business users.

    Just watching the combined garbage on a spectrum analyzer was an eye opener. I never considered hooking up my HP-437B with a 50 gHz head to see what crud levels it measured, gotta do that soon.

    Except for one of Jay's GaAs 432 preamps on an ancient IC-451A I use for band opening monitoring 432.1, all my other preamps are home brew up to 10 gHz built when I was in R&D at a microwave company doing daily development from 12-75 gHz.

    AI3V and AF6LJ like this.

Share This Page