High speed sine wave to square wave converter

Discussion in 'The Projects Forum' started by thavamaran, Nov 1, 2010.

  1. thavamaran

    Thread Starter Active Member

    Feb 21, 2008
    67
    0
    Hi guys, I need to convert a sine wave to square wave, sine wave is running at about 1.5 GHz, so I had an idea of using comparator, but im afraid that most of the conventional op-amps cant support my speed requirement.

    Any advice and reference circuits please.
     
  2. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    To get even a halfway decent looking square wave from a 1.5GHz sinewave, you'll need a device with ~30GHz bandwidth. I wish you good luck in finding such a device, as you will need it.

    If you meant 15MHz instead of 1.5GHz, an LT1016 comparator would do nicely.
     
  3. Kermit2

    AAC Fanatic!

    Feb 5, 2010
    3,792
    948
    Have you thought of amplifying the sine wave and feeding it to a diode clipping circuit. That would just chop off the top and bottoms and leave something very very close to a squarewave.
     
  4. thavamaran

    Thread Starter Active Member

    Feb 21, 2008
    67
    0
    Hi SgtWookie, will schmitt trigger help for this application? Can I know why 30 GHz bandwidth needed?

    I found this schmitt trigger SN74LVC1G17DBVR, will this work.

    Is there any other way that I can do this?
     
  5. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    Why do you need square wave anyvay? How pecise does the sqare wave have to be, i.e. what rise and fall time do you need?
     
  6. Kermit2

    AAC Fanatic!

    Feb 5, 2010
    3,792
    948
    [​IMG]

    One unusual approach uses a series RLC circuit to step up the voltage. At the input frequency, the RLC network simply looks like a resistor since the inductor and capacitor are series resonant. The resistor value is selected to limit the current at resonance (typically 100 ohms) and the inductive reactance is selected to give the desired gate voltage. In the example below, the inductor is selected to give about 5 volts of swing for the signal input current.
     
  7. thavamaran

    Thread Starter Active Member

    Feb 21, 2008
    67
    0
    Hi Kermit2, it sounds like a good idea, i will look into it. thanks very much.

    Hi kubeek, Im going use this 1 GHz square wave as an input battery to my biasing circuit. So I have a device generating sine wave at 1 GHz, so I need to convert it to a square wave. I am not really sure about rise and fall time, cause Im not sure how it will help it to work like a battery.
     
  8. blueroomelectronics

    AAC Fanatic!

    Jul 22, 2007
    1,758
    98
    Oh please post a schematic.
     
  9. thavamaran

    Thread Starter Active Member

    Feb 21, 2008
    67
    0
    Schematic of the bias circuit? But bias circuit is just an independent circuit, it has nothing to do with converter.

    If you still need it, I can get it for you. Thank you.
     
  10. Papabravo

    Expert

    Feb 24, 2006
    10,147
    1,791
    I'm with the sarge on this one. I think you have no chance at a 1.5 GHz. square wave with ANY conventional logic or analog ICs. Using discrete RF devices or MMICs and microstripline techniques you might have a chance. The problem with overdriving and clipping is that the storage times, and the charging an discharging of the parasitics will absolutely kill 'ya.
     
  11. Papabravo

    Expert

    Feb 24, 2006
    10,147
    1,791
    Probably not possible without a discrete RF device or MMIC and microstripline techniques. The CAD tools alone for this will break your wallet, never mind the test equipment. Several hundred thousand would not be unreasonable.

    You need a 30 GHz. bandwidth because the squarewave can be thought of as a Fourier series of terms at the odd and even harmonics of 1.5GHz. The lower the bandwidth of the device the more rounded the corners of the "squarewave" will be until your output looks like a sinewave at the fundamental .
     
  12. thavamaran

    Thread Starter Active Member

    Feb 21, 2008
    67
    0
    Papabravo....I am lost with your explicit description....can you enhance a little bit....

    what do you mean by rf discrete device....

    I get your idea of rectangular and 30 GHz bandwidth because of the Sinc function in fourier transform. But in that case how about clipping? What do you mean by storage device? And what will the parasitic do?

    But thank you very much.
     
  13. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    As far as the 30GHz bandwidth requirement, have a look at the Wikipedia entry for "Square Wave":
    http://en.wikipedia.org/wiki/Square_wave
    Scroll halfway down, and click on the image on the right to animate it, or click here:
    http://upload.wikimedia.org/wikipedia/commons/f/f8/SquareWave.gif

    A theoretically perfect square wave consists of the fundamental frequency, plus ALL of the odd harmonics of the fundamental frequency. This implies that unlimited bandwidth is required to faithfully reproduce a true square wave, which is of course impossible. Actually, you could probably get a fairly decent-looking square wave if you just went up to the 9th harmonic, but you'd have a good bit of ringing on the leading/trailing edges.

    Going with Kermit2's idea, you might check MiniCircuits amplifier selection, which is on this page:
    http://www.minicircuits.com/products/amplifiers_main.html

    Their components/devices are remarkably inexpensive (usually in the $20-$50 range) for the frequencies they're capable of handling - and they are ready to install.

    You'd either have to use coaxial cable (I'd suggest SMA type with semi-rigid coax) or stripline if really low-power.

    You'd also need to find a pair of extremely high speed diodes to clamp the output at the rails.

    Since I don't know what your input signal level is, you'll need to determine the required gain.
     
    Last edited: Nov 1, 2010
  14. blueroomelectronics

    AAC Fanatic!

    Jul 22, 2007
    1,758
    98
    Why would anyone want to pump 1.5GHz into a battery?
     
  15. Kermit2

    AAC Fanatic!

    Feb 5, 2010
    3,792
    948
  16. Papabravo

    Expert

    Feb 24, 2006
    10,147
    1,791
    An RF discrete device is an RF transistor that is specially fabricated and characterized to be used for a particular application at a particular range of frequencies. The problem is that actual devices represent a distribution of values with a mean and a variance. In practice we get a handful of devices and characterize them on a VNA (Vector Network Analyzer) or similar device so we have some idea of how the parameters may vary. Then we try to design circuits that don't depend so much on actual values but will work over some range of values. This is where the expensive test equipment comes in. I don't know many folks who can design this kind of stuff without some sophisticated CAD tools and equally sophisticated test equipment. You could be the exception.

    An MMIC is a Monolithic Microwave Integrated Circuit. These can be useful in building up a system from pieces parts, but you still have to know how to match the guzintas and gazatas (aka input and output impedances) and how to layout either a PCB or a dead bug breadboard. This stuff can be done but you'll need some luck, experience or a dutch uncle to get it right.

    Storage time refers to the buildup of charge in the base region. When the bias voltage drops below the cutoff value there is some amount of time before the transistor stops conducting. The time delay is referred to as storage time. You can easily see that with several nano-seconds of storage time that beating up and down at 1.5 GHz would be difficult.

    Parasitic capacitance exists between each of the electrodes on a transistor. Whenever you change the transistor state these parasitic capacitances have to be charged and discharged. Even a few picofarads with any appreciable source resistance represents a time constant in the nanosecond range; this makes sub nanosecond square wave oscillation oscillations difficult to establish and maintain.

    Can I ask if this is for some kind of overunity project?
     
    Last edited: Nov 1, 2010
  17. thavamaran

    Thread Starter Active Member

    Feb 21, 2008
    67
    0
    Guys, I need this bias current device to bias my semiconductor optical amplifier for switching purposes.

    And this bias current device can only work according to my requirement by having a 1ns on and 1ns off time, which has 50 % duty cycle and gives 500 MHz signal. I wanted 1.5 GHz because I might increase the rate in future.

    I have a signal generator that generates sine wave at this rate, so if I could convert it, then it will make my life easier.

    So can I just use fast diodes to clip of the peak of positive and negative of the signal which will give some sort rectangular wave. I saw one of this circuits call diode clipping circuit.

    Papabravo, thanks a lot!
     
  18. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    Have you tried feeding the sine wave straight into the "mystical device"? It might work as well, because at that frequency nothing does really look like a square wave and everything starts looking as a sinewave anyway.
     
  19. blueroomelectronics

    AAC Fanatic!

    Jul 22, 2007
    1,758
    98
    I second the question is this an overunity project? Just so we're not all wasting our time.
     
  20. thavamaran

    Thread Starter Active Member

    Feb 21, 2008
    67
    0
    Mystical device....lol! I have designed diode clipping circuit today...dont worry, you are not wasting your time, its a huge project.

    The diode is quite slow, so I am looking for fast diodes now, any suggestion?

    Anyway, sorry for cross posting, I have another thread on OFDM yesterday, do anyone of you have worked on OFDM before? Thanks guys.
     
Loading...