Connect with us

RF amplification

Discussion in 'Radio and Wireless' started by MBrantley, Nov 11, 2015.

  1. MBrantley

    MBrantley

    8
    0
    Jan 21, 2015
    Hello everyone,

    A DAQ system we recently built using NI components has a few RF and arb outputs that need to be amplified. In short, we use electric and magnetic fields to manipulate the movement of ions in ultra-high vacuum to study their "mass-to-charge" ratio. Our outputs are coming from a PXIe-5451 and several PXI-5404 cards but need to be amplified to reach our goal of 200-300Vp-p. I have spoken with a few companies and have been recommended products but they tend to be large rackmount amplifiers costing upwards of 10,000USD each. Our target frequencies are low KHz to ~20MHz.

    My question is: how are commercial instruments in my field able to have 4-6 of these amplifiers built into their relatively small chassis? Does it have to do with the target load, most amps I see are for 50 Ohm loads? These signals are applied on metal rods/plates with very little inherent resistance (however they do have some capacitance). I'm most interested in having the correct voltage differences between the plates and I am having trouble understanding why so much current is needed for this to happen.
     
  2. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    25,174
    2,690
    Jan 21, 2010
    I'm having difficulty figuring out why a data acquisition device would have RF outputs and why you would need to amplify them to 200-300v.

    Nevertheless, the primary issue is the output characteristics of the daq as this will determine the input characteristics of any amplifier.

    The next issue is the characteristics of the load.
     
  3. MBrantley

    MBrantley

    8
    0
    Jan 21, 2015
    Hi (*steve*), thank you for responding. In our experiments we study the way that that molecules react to various electric/magnetic fields in the gas phase - there is an amazing amount of chemistry you can study this way. As such, we first must expose them to known fields and then record how they behave accordingly (this can be destructive with items like faraday cups or electron multipliers or non-destructive by simply measuring the current induced as they move past a plate).

    The output from the 5404 are single frequency DC-100 MHz at +/- 1 V and 50 Ohms. The 5451 is an arbitrary waveform generator that will be outputting complex wave forms at +/- 2.5 V and 50 Ohms. The loads themselves are metal rods of various shapes/sizes, I expect their resistance is minimal. Will capacitance of these loads effect more than the slew times?

    Thanks,
     
  4. hevans1944

    hevans1944 Hop - AC8NS

    4,203
    1,988
    Jun 21, 2012
    The capacitance (a few pico-farads) becomes important at higher frequncies, affecting both slew rate and peak-to-peak amplitude. There are also impedance mis-match considerations at higher frequencies which can cause reflections (standing waves) on the connecting coaxial cables.

    The biggest problem I encountered, when using electrostatic deflection plates to raster-scan a high-energy ion beam across a target, was attenuation of the triangular scan voltage waveform and "rounding" of the peaks, caused by cable capacitance. This problem was eventually solved by removing the outer woven shield on the coaxial cable, leaving just a stub of shield where the coax was terminated in an RF connector. The problem only occurred on the faster horizontal scan, which was a few hundred hertz. The much slower vertical scan did not have any problem. These scan voltages were variable, up to several thousand volts peak-to-peak, derived from high-voltage transformers driven by power op-amps in a commercial product. The entire scan generator was contained in two rack-mounted cabinets, one for the oscillators and one for the power amps.

    I mounted the power amps and transformers box as close as possible to the four deflection plate terminals to minimize cable capacitance, but did not attempt to provide a "matching" impedance as a resistive load. You may need to do that, depending on distance to the ±300 V amplifiers and frequency of operation.

    Can you specify what frequency range you want to cover? If not too high a frequency, and if you have the engineering resources, you may be able to build your own high-voltage linear amplifiers. If operation at very low frequencies and DC is not required, you may be able to use your frequency sources to drive low-voltage power op-amps, which in turn can drive step-up transformers with a modest primary-to-secondary turns ratio. And, as you have discovered, there are "boat anchor" off-the-shelf solutions.
     
  5. MBrantley

    MBrantley

    8
    0
    Jan 21, 2015
    Thank you for the reply hevans1944,

    The frequencies we are working with are low KHz up to ~50 MHz. The frequencies are many times analyte dependent (i.e., their unique cyclotron frequencies) so I want to retain as much headroom as possible. Additionally, for the single frequency signals we will be using almost exclusively sin waves (although research tends to cause you to do unexpected things!). For a single frequency I can readily generate +/- 500V using a small RF amp (Henry Amplification 20B - 20 Watt) and a hand wound air-core transformer from my +/- 1 V source - however I have to use a trim pot to tune the frequency. I tried using some small broadband RF transformers from Coilcraft but was only able to get a very narrow frequency band to work. This could be due to my poor engineering skills.

    Thanks!
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Similar Threads
There are no similar threads yet.
Loading...
Electronics Point Logo
Continue to site
Quote of the day

-