[Laser] Re: Laser Digest, Vol 22, Issue 9

[wa4qal at ix.netcom.com]

> Date: Wed, 26 Apr 2006 22:48:27 EDT
> From: TWOSIG at aol.com
> Subject: Re: [Laser] Ultrasonic Modulator?
> To: laser at mailman.qth.net
> Message-ID: <2c0.905b892.31818afb at aol.com>
> Content-Type: text/plain; charset="US-ASCII"
> 
>  
> Kurt
>  
> First, let me suggest that you use a sharp cutoff low pass filter  filter on 
> the voice component.  The cut-off frequency should be  less than one half, and 
> I would recommend less than one third of the carrier frequency.  I do not 
> know what microphone you are planning to use, but I  would guess that it will 
> pick up lots of energy into the frequency spectrum that you are trying to use 
> for a carrier.  ( I am betting that it will  oscillate. )

It's generally a good idea to limit the modulating frequencies to less than
half the carrier frequency, lest you end up with a real mess at the output.
Since brick-wall filters don't really exist, it might even be better to limit
the modulating frequency to 1/3 or less the carrier frequency.  That's
probably not a big problem if you're using a 30 KHz carrier, since that 
still gives you audio up to 10 KHz, which is plenty (probably too much!)
for speech, and will even carry some music (not hi fidelity, but better than
AM BCB radio).  

> Second, I suggest that you use the highest frequency for your  carrier.  The 
> power levels that you are suggesting would be very painful  for those of us 
> that can still hear.  Even if you limit the voice component  to 3KHz with a 
> brick wall filter, the mixer products of that with a 20KHz  carrier would include 
> sounds down to 17KHz.  I still have (after more than  five decades) at least 
> 80 percent of normal hearing in that range.  If you  can push the carrier above 
> 30KHz, it would be better for your neighbors.

I'd definitely recommend at least 30 KHz if not higher, else you may discover
a new detector; "dog demodulation".  Everytime you key the thing up, the
neighbourhood dogs may start to howl.  

> Third, reconsider your modulation scheme.  A simple 555 timer can be  used to 
> frequency modulate, or even pulse width modulate in the 30 to 50 KHz  range.  
> There are also phase lock loop circuits that will work there.   If this seems 
> acceptable, then search for the circuits for an FM wireless  intercom.  The 
> idea with those was to use a 30 plus KHz carrier which was  coupled to the AC 
> house wiring.  You would just couple to the audio  amp.   ( You might even find 
> a commercial unit at a garage sale.  )

FM might work, but I'd be a little concerned about how well the transducers
will work across a very broad frequency range.  Plus, the modulation index
will be fairly large, given such a low frequency carrier and a high frequency
modulating signal.  Still, it would probably work (but, Bessel Functions make
my head hurt!).

> If you still want an AM modulator, I have a suggestion for a circuit, but  it 
> will only work for you if you do not care about generating spurious signals  
> at the odd harmonics of the carrier frequency.  The reason is that the  
> simplest way to generate the carrier is with a 555 timer circuit, and it will  
> produce square waves that are rich in harmonic content.  The modulator will  
> produce modulated waveforms in that part of the frequency spectrum as  well.  This 
> circuit is based on what I use for AM modulation of an LED for my light 
> communication demonstration.
>  
> In my circuit I use 324 op amps to produce an audio signal biased at 6  
> volts, which is fed to through two 10K resistors to ground.  The common  point of 
> the two resistors provide half the voltage of the signal to the non-inverting 
> input of one op amp.  The output of the amp is fed through a curent limiting 
> resistor of 1K to the base of a 2N2222.  The emitter of the transistor has a 
> 10K resistor to ground and a 100 ohm resistor and a diode to the output of 
> the 555.  There is also a line from the emitter to the inverting input of the 
> op amp for feedback.
>  
> I use an LED from the positive supply line to the collector of the 2N2222,  
> but you would use a transformer to couple to the audio amp.
>  
> The circuit works this way.  With no signal, the 6 volt bias goes throught 
> the divider to put 3 volts on the input of the op amp, which then drives the 
> transistor to conduct.  If the 555 output is in the high state the diode 
> blocks any current from it.  The transistor then only sees the 10K path to ground 
> and will provide only 0.3 milliamp, which is all that is needed to develope 3 
> volts across the resistor.  The 3 volts are fed back to the op amp.  If the 
> 555 output goes low, the 100 ohm resistor is now in parallel, and needs an 
> additional 30 milliamp to satisfy the op amp  feedback.  With no audio input, 
> the circuit provides a steady 300 microamp of current and 30 milliamp pulses 
> through the primary of the transformer.
>  
> Now imagine what happens if you have an audio signal comming in on the 6  
> volt bias.  The op am will try to match the input voltage (and please note that 
> the op amp never sees the pulse rate of the 555, it could be an RF signal  as 
> long as the 2N2222 can respond to the pulses).  A 4 volt peak sine wave in 
> would swing from 2 volts to 10 volts, resulting in the current through the  
> transformer going from 100 microamps to 10 milliamps on some pulses, then up to  
> 500 microamps to 50 milliamps.
>  
> You could use an opto-isolator instead of the transformer, or some other way 
> to convert the current pulses to the input you need for your audio amp.
>  
> Oh.  You will also get the base band audio signal as well as the carrier 
> frequency and all the odd harmonics.   This would be a really lousey circuit if 
> used for RF, but it works for my light communication demo.  If you can use 
> it, I am glad to share.

Another approach would be to use a balanced modulator, such as the SA602 or
SA612 (offset the input audio to prevent negative modulation).  That might 
give you a cleaner output.  Maybe.

> Best Wishes.
>  
> James 
> N5GUI

Dave