Amplitude Modulation with Function Generators

[quantumdude]

Hello everybody,

I am trying to use a pair of Tektronix FG 503 Function Generators to modulate a carrier wave and produce an AM modulated signal. I tried running the output of one generator into the VCF input of the other, but it seems to give me FM modulation. I did this because I am working from a lab handout that calls for function generators that have a modulation input feature.


What, if anything, can I do to get AM?

 

[Antiphon]

If the signal generator won't take amplitude control,
you need to get a voltage-controlled amplifier.

 

[berkeman]

Hello everybody,

I am trying to use a pair of Tektronix FG 503 Function Generators to modulate a carrier wave and produce an AM modulated signal. I tried running the output of one generator into the VCF input of the other, but it seems to give me FM modulation. I did this because I am working from a lab handout that calls for function generators that have a modulation input feature.


What, if anything, can I do to get AM?

You have a couple of options, depending on how comfortable you are with building support circuits. You can use a double-balanced mixer component:

http://tele-tech-rf.com/mixaps.htm

You can buy these from Mini-Circuits, and probably even through Digikey. Another option is to use a double-balanced mixer IC based on the Gilbert Cell:

http://members.tripod.com/michaelgellis/gilbert.html

And a 3rd way would be to brew your own AM transmitter based on opamps and transistors -- similar to the Gilbert Cell architecture, but a little more precise.

Hope that helps. -Mike-

 

[George Jones]

Hello everybody,
I am trying to use a pair of Tektronix FG 503 Function Generators to modulate a carrier wave and produce an AM modulated signal. I tried running the output of one generator into the VCF input of the other, but it seems to give me FM modulation. /QUOTE]

Sorry about the longwinded nature of this post.

A VCF (voltage contolled frequency) input produces an output frequency such that the difference between the output frequency and the reference (dial) frequency is proportional to the voltage input. Inputting a varying voltage produces a varing frequency output signal, i.e., frequency modulation, just as you say.

Amplitude modulation looks like cos(w_m t)cos(w_c t), where w_m is the frequency of the modulating signal and w_c is the frequency of the carrier, so, usually, w_m << w_c. cos(w_m) is thought of as a varying amplitude of the carrier cos(w_c t) part.

By the cos addition formulae,

2cos(w_m t)cos(w_c t) = cos[(w_c + w_m)t] + cos[(w_c - w_m)t],

or, with w_c = 1000 and w_m = 10, for example,

2cos(10t)cos(1000t) = cos(1010t) + cos(990)

Amplitude modulation is just the beat phenomenon of 2 signals with close frequencies beating. To produce AM, add the outputs of 2 generators that have almost equal frequencies and equal amplitudes.

Voltages from generators can be added just as voltages from batteries - connect them in series.

I was aware of all this "theory" last summer when you posted, but when I went into the lab to actually produce AM, I realized there was a big problem - both generators have grounded outputs. Connecting them in series would involve connecting the ungrounded output of generator A to the grounded output of generator B, effectively shorting the output of generator A.

Two days ago, while doing background reading for electronics (which I have assigned to teach this semester), I found a possible solution. Some generators have a switch on the back that allows the chassis either to be grounded or floating. If your generators have this switch, set the switch on B to the floating position.

Our tech guy, who is very good, saw me examining the generator, and asked me what I was doing. He said that occasionally he has to "float" equipment that doesn't have this switch, and he took 2 generators (no switches) down and showed me how to it.

The idea is simple - just snip off the ground pin on one of the plugs. It is best to snip the pin off an adaptor that plugs into an elctrical outlet and plug generator B into the adaptor.

Either using the switch, or by using an adapter, float B. Get 3 leads (one for each generator and one to connect to a scope so that AM can be verified) that have bnc onnectors on one end and alligator leads on the other. Connect the ungrounded lead from A to the lead from B that normally would be grounded. This connects the generators in series.

To see AM, connect the grounded lead from the scope to the grounded lead of A, and the ungrounded lead from the scope to the "ungrounded" lead of B.

After plugging A and B into the appropriate electrical outlets and turning on A and B, but before connecting any of the bnc cables, it might be wise to make a couple of checks with a multimeter.

1) Measure the voltage between the chassis of B and ground to ensure that the chassis is not "hot". This can be done by measuring the voltage between the outside of a bnc connector on A and the outside of a bnc connector on B. This voltage should be zero. If it isn't, don't proceed.

2) Measure the resistance between the same 2 points. This should be infinite.

Our tech guy and I did all this 2 days ago, and eveything worked fine.

Regards,
George

 

[berkeman]

Just a couple of caveats about cheating the ground connection on equipment that uses a 3-prong AC mains plug...

First, remember that the 3rd prong is there for a safety reason. The metal chassis of the instrument is connected to this safety ground so that if the AC mains Hot wire comes loose at the input and ends up touching the metal chassis (which is user-accessible and could shock you), the fuse will blow because of the short cicuit. When you cheat the ground on an instrument, you lose this safety feature. Granted, it's a pretty unlikely failure mode for most commercial instruments. But it's a more likely failure mode for a hand-built prototype like you might see in a lab...

Second, cheating the ground only really works for low-frequency applications. The power supply in most instruments has enough leakage capacitance from its outputs to its AC mains Hot/Neutral inputs, that for high frequencies you cannot break the connection. So cheating the ground on an instrument will likely help you at audio frequencies, but not in the 10+ MHz type range.

 

[Averagesupernova]

Amplitude modulation is just the beat phenomenon of 2 signals with close frequencies beating.

There was a discussion about this sort of thing a year or more back. There was a member (who no longer is a member) who insisted that AM is nothing more than a varying amplitude of a single frequency. I finally lost interest in arguing with him. A composite AM signal consists of 3 frequencies. They are the carrier and upper and lower sidebands. I told him he could achieve the same composite signal without actually changing the amplitude of anything and just summing 3 individual signals together. I posted quotes from textbooks and everything. He was a slippery individual who seemed to be able to make himself sound right and quite proficient at twisting around things that I and he had already posted into sounding like something else.