Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Amplifier Circuit Advice -- AF/RF

  1. Apr 9, 2015 #1
    Not sure if this is the right place to post, but here we go, my first question:
    I've been searching around for a simple circuit for amplifying rf at around 7MHz, and went ahead and followed this guide, 2nd picture down, and built the amplifier. It works, even with some small resistor value shifts (82k->79k and 39k->38.7k), but only in audio frequencies. At 600Hz, the amplitude of the resulting waveform is roughly 3 times the input. However at higher frequencies (7MHz), the resulting amplitude is only about 1/8th of the input. I also tried this design (from another site) using 10k ohm and 4k7 ohm for the voltage divider (which gets roughly the same voltage at the transistor base, but current??)

    So my question: is it possible to make this design work at rf? If so, what component values should I use?

    Here are two of my attempts,another two I did on my breadboard(class C & AB), which got taken apart:
    upload_2015-4-9_14-14-9.png
    Unsurprisingly, neither the class AB or C worked :smile:.
    Thanks!
     
    Last edited: Apr 9, 2015
  2. jcsd
  3. Apr 9, 2015 #2

    analogdesign

    User Avatar
    Science Advisor

    Obviously your amplifier is not fast enough to amplify at 7 MHz (which isn't generally considered RF anymore btw). The basic design is standard and should be fine well beyond 7 MHz.

    Do you know how to calculate amplifier bandwidth? There are a million webpages to help. Basically the product of the gain and the bandwidth (the point at which the small-signal gain is 1) is gm/C where gm is the transconductance of your transistor (also called hfe and equal to IC/vt or Ic/26mV) and C is your load (including self-loading effects). You can use the load resistor to trade off gain for bandwidth once its working.

    Now you can see what's going on. What is your load capacitance? If it is small you might be loading it with a probe. Also, what is your collector current? You can increase your bandwidth by running the transistor hotter but make sure you don't put too much current through it (read the data sheet).

    You might also need a faster device but I'm not convinced about that. You probably just need more current.
     
    Last edited: Apr 9, 2015
  4. Apr 9, 2015 #3

    berkeman

    User Avatar

    Staff: Mentor

    Welcome to the PF.

    Yeah, the Miller capacitance is going to limit that design to lower frequencies. Have a look at "cascode" transistor amplifier circuits to get rid of the Miller capacitance issue... :smile:
     
  5. Apr 9, 2015 #4
    I think the designer has made the amplifier for 2 MHz, so the output will be very reduced at 7 MHz.
     
  6. Apr 9, 2015 #5
    Thanks for your responses, I now have much to think about.
    I'm really not good with the math at all and only know a bit of the theory behind these amplifiers (my goal to fix by building one of these), but I do have some tools to make a few measurements:
    Collector current: 127uA with 9V input, 751uA with 12V. I just de-soldered and measured it with multimeter
    Load capacitance: 608pF (stuck a cheap capacitance meter at the load, an antenna), esr=17 ohm (haven't figured out what "equivalent series resistance" is either)
    hfe of transistor: 310 (also measured)

    I really don't need much bandwidth, and only have a basic understanding of how it relates to gain. It might take me a while to figure this topic out through Google.

    I skimmed through the wikipedia cascode page quickly, and I do have to say, it looks promising. I also happen to have a few 2N5951's "N-Channel RF Amplifier"s around (datasheet). Maybe that would be a better path, but for now, should I play with resistor values at the collector?

    But other than that, I still don't really understand how to set the maximum gain of the amplifier to a certain frequency, and reduce bandwidth.
    Thanks for bearing with me!
     
  7. Apr 9, 2015 #6

    Averagesupernova

    User Avatar
    Gold Member

    Most amplifier designs that get above a couple of Mhz are going to use choke in place of the collector resistor.
     
  8. Apr 9, 2015 #7

    analogdesign

    User Avatar
    Science Advisor

    If you don't need a lot of bandwidth, but you want gain at a specific frequency (for example, say you want to amplify a 100 kHz band around 7 MHz) you should look into using a tuned amplifier. This goes along with the choke suggestion from Averagesupernova.

    http://en.wikipedia.org/wiki/Tuned_amplifier


    Basically you put an LC take in your collector circuit and set the resonance frequency of the tank to the center frequency. In your case you would chose sq-rt(1/LC) = 2*pi*7MHz
     
  9. Apr 9, 2015 #8
    Tuned amplifier... interesting.
    I just went and replaced the collector resistor with a choke, and it's now at 1/2 the original input (rather than 1/8).
    Perhaps I should just scrap this amplifier and try the tuned one.
     
  10. Apr 9, 2015 #9

    analogdesign

    User Avatar
    Science Advisor

    I don't think you're going to have to scrape it. Just replace the choke with an LC tank. You may need to AC couple the output if it isn't configured that way already.
     
  11. Apr 9, 2015 #10
    Output is AC coupled with a 0.1uF cap.
    I replaced the choke with a tank (470pF+47pF, 1uH --> 6.99963MHz, pretty close)
    Results are just about the same as the choke. On my scope, the wave is a bit less fuzzy (why?) , but the amplitude is just about the same.
     
  12. Apr 9, 2015 #11

    Averagesupernova

    User Avatar
    Gold Member

    Watch out so your tuned amplifier doesn't become a class C amplifier which is not a linear amplifier. This is acceptable for amplifying an CW, FM or FSK signal for instance. However it would be unsuitable for amplifying an AM signal. Amplifiers in the IF strip in receivers would be tuned amplifiers however they are generally not class C.
     
  13. Apr 9, 2015 #12
    If class C's aren't good for AM, would that mean that AM signals would be distorted enough to almost be "filtered" out, or would it just turn into tons of noise on the receiving end? That would be fun to test, but I just don't have the time this week, or next week.

    I'll take some time to consider a few of the suggestions above, and come back in a week or two to (hopefully) report on any successes.
    Once again, thanks to all who responded!
     
  14. Apr 9, 2015 #13

    Averagesupernova

    User Avatar
    Gold Member

    A class C amplifier is a sort of limiter. It relies on the tank circuits flywheel effect. The transistor is in conduction less than 50% of the time.
     
  15. Apr 10, 2015 #14

    f95toli

    User Avatar
    Science Advisor
    Gold Member

    Are you building the amplifier because you actually need an amplifier? Or because you want to learn more about electronics?
    If it is the former you could just use a fast op-amp which is quite a bit easier to use. There are many op-amps out there with a gain BW product of over 1 GHz, meaning you can get a lot of gain at 7 MHz. All you need is the op-amp, a few resistors and some decoupling caps.

    .
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Amplifier Circuit Advice -- AF/RF
  1. RF circuit parameters (Replies: 7)

  2. RF Mixer circuit help (Replies: 25)

Loading...