fm-am transmitter/receiver

[nwar1994]

Moving onto the demodulator. I have wired it on the breadboard an am testing with sin waves between 1070 and 1270 Hz to see the binary output.

 

[RadioRon]

There are two low pass filters involved in this circuit, one being the loop filter (C5 and 3.6K ohm inside the 565 chip) and the other being a smoothing filter to eliminate the 1070 and 1270 Hz ac from the output. I'm not sure which one you are referring to, exactly. However, in both cases, the only connection between the filter and Vcc is through a few capacitors. Since capacitors block DC, it would seem that these caps can connect to +5V, to GND, or to -5V as you like. This assumes that the power supply rails are well bypassed to ground so that from an AC point of view, +5V, -5V and GND are connected together. I'm not sure why they connect to +5. I see this right on the datasheet for the NE565 as well. Sometimes, there is something inside the IC that is sensitive to transient voltages or small ground impedances, and in such cases, connecting the loop filter or low pass filter caps to + supply may be necessary to avoid some quirky problem. I can't tell if that is the case with this chip or not. On the other hand, maybe the caps are connected to +5 just to make the schematic diagram a bit tidier than taking them to ground.

Its always best to connect as recommended in the datasheet because the author of the datasheet knows more about the quirks of the IC than you or I. However, make sure you have a good sized bypass cap on your +5V rail.

To assist you with debug, I recommend that you first test the NE565 without any connections to the comparator. With power applied, You should be able to see the VCO output at pin 4 and you should be able to vary the frequency of this by varying R4. The frequency should be adjustable to about 1000Hz. Now, when you input a tone, say 1070 Hz at pin 2 through C3, you should be able to measure a DC voltage at pin 7, and a similar DC voltage at pin 6. Let me know what DC voltage you get at these pins, assuming that R10 and R7 have been removed to disconnect the comparator.

This DC voltage at pin 7 will vary up and down as you vary the input frequency.

 

[nwar1994]

Pin 4 output:
Frequency: 1.2kHz (Square wave), max: 440 mV

Pin 6 output:
Frequency: DC, max: 400 mV

Pin 7 output:
Frequency: 1.230 kHz, max: 512 mV

 

[nwar1994]

Without comparator connected, no sine signal
Pin 4 output:
Frequency: 1.174kHz (Square wave), max: 4.5V

Pin 6 output:
Frequency: DC, max: 4 V

Pin 7 output:
Frequency: DC, max: 3.92

 

[nwar1994]

Without comparator connected, with sine signal
Pin 4 output:
Frequency: 1.174kHz (Square wave), max: 4.5V

Pin 6 output:
Frequency: DC, max: 4 V

Pin 7 output:
Frequency: DC, max: 3.92

It looks like my pin 7 output will never be larger than my pin 6 output based on my frequency range 1070 - 1270. Would I use a voltage divider to tune pin 6? The base voltage I would like is 3.76V for a midpoint at 1170 Hz, since 1070 Hz is at 3.84V, and 1270 Hz is at 3.64V.

 

[RadioRon]

You must not use a voltage divider on pin 6.

Let's check that your PLL is functioning correctly. Disconnect the comparator for now. Monitor Pin 7 of the PLL with the scope. With no sine input, tune your VCO to about 1100 Hz (I think you've already done that) by varying R2. Now, put a sine input into pin 2 with a level of 300mVpp, roughly at 1000 Hz. Then, while monitoring the voltage at pin 7, tune the sine input frequency up and down over a range of 500 Hz to 5000 Hz or so. You should see the DC component of the voltage at pin 7 go up and down as you tune the frequency. Pin 7 is the loop filter voltage and when the loop is locked to the incoming sine wave, the loop filter voltage will go up and down tracking the sine input frequency and in so doing, cause the VCO frequency to track the incoming sine frequency. Make sure this works properly before doing anything with the comparator. When the loop is correctly locked, the vco frequency always matches the input frequency.

The VCO output at pin 4 should always swing between 0 and about Vcc-1 volts (in your case, about 4 volts). If it is not, then there is something wrong with the power supply voltages.

Once the VCO frequency is locked to the incoming sine frequency, you should be able to adjust the DC output voltage at pin 7 for that one frequency by adjusting R2. This changes the free-running frequency of the VCO and so also changes how much DC control voltage is needed to push the vco back to your input frequency to maintain lock. Don't go too far or you will cause the vco frequency to go out of the lock range it will not be able to lock.

The data sheet for the NE565 tells me that the output voltage at pin 7 when arranged as a demodulator (such as you have) should be close to 4 volts and that the AC swing due to your shifting frequency of 1070 to 1270 will be quite small, perhaps 200 mVpp. Measure this after the ladder filter. (You should also expect an offset voltage between pin 6 and pin 7 of 50mV, as they won't be exactly the same.) Pin 6 should only be about 50mV or perhaps a bit less different than the DC component of pin 7. Your comparator has to work with the roughly 200mVpp AC swing that will be seen on the output of the ladder filter (at the input of the comparator if it were connected).

The comparator has to function with both inputs at around 4 to 4.5 volts. If it is powered by only 5V, is it able to do that? I don't think so. The data sheet implies that the highest the input voltage can go and the comparator still function is 1 volt lower then Vcc. To test if this is the problem, is it possible for you to supply the comparator (and the comparator only) with +/- 10 or 12 volts temporarily?

 

[nwar1994]

So my pin 6 of the PLL, stays locked at a DC voltage regardless of the frequency I input

 

[RadioRon]

Does the vco frequency track the incoming frequency? If not, then the loop is not locked. You may have to adjust your vco frequency.

I'm not all that familiar with the 565 as I haven't used one in 35 years. I don't know exactly how the pin 6 voltage is derived inside the chip. If I were you I would assume that it is fixed or very slow changing, and pay attention to what is on the output of the ladder filter connected to pin 7

 

[nwar1994]

Pin 7 adjusts, but pin 6 does not.

On pin 4, it tracks the signal generator frequency

 

[RadioRon]

So it seems to be working. Now adjust your vco resistor potentiometer to set the voltage at pin 7 on one side of the voltage at pin 6 for one of your two frequencies and note that the other frequency should swing the voltage to the other side of the pin 6 value, if you get what I mean.

 

[nwar1994]

Ok pin 6 sits at 3.92V, 1070 Hz gives 4V at pin 7, and 1270 Hz gives 3.84V at pin 7. This was obtained from adjusting the VCO frequency.

 

[nwar1994]

It appears there is a discrepancy on the pin 7 output

 

[nwar1994]

I tried using the lm111J by itself

 

[Nigel Goodwin]

I tried using the lm111J by itself, but the output was locked at 1.5V. I switched to using the LM741 as a comparator and it jumped between 1.88V and 3V consistently. I must be hooking up the lm111J wrong

Sounds like you are - most comparitors have open collector outputs, so you require a pull-up resistor to make it go high - refer to the data sheet for the chip.

 

[RadioRon]

Those voltages that you measured sounded just fine. One of the ac tones gives a voltage on one side of your pin 6 reference, while the other tone is on the other side. So, if you follow this with a working comparator, it should be ok. No need to divide the voltage at pin 6. The reason I said don't do that was because I'm unsure of a couple of things about pin 6. One is that you probably should not bleed current on that pin and any resistor to ground (ie. part of a voltage divider) would do that. The other thing is that it simply should not be necessary. I frankly, don't know exactly how pin 6 voltage behaves. I mean that I don't know if it tracks an average voltage of the loop filter, or if it is completely fixed at one voltage only. Perhaps others can chime in with more experience.

 

[nwar1994]

I've tried hooking up the pull up resistor, but the varying the DC voltage to the inputs of the LM111J doesn't change anything.

 

[nwar1994]

So using the LM741 I am able to get 5.20V at 1270 Hz, and 2.24V at 1070 Hz. I grounded all the open pins on my IC

 

[RadioRon]

Op amps and comparators are not like digital logic chips. You don't automatically ground unused pins. You normally study the data sheet to figure out the function of each pin and then terminate the pin appropriately. A pin might need Vcc, or -Vcc, or gnd or a resistor load, or might have to be connected to another pin (like 5 and 6 on the LM311). My review of the LM311 datasheet implies that unused pins should be left open if you are unsure of what to do with them. That, however, does not help explain why yours isn't working.

 

[RadioRon]

When I need to figure out an IC, I like to isolate it from other circuits, and simplify my test circuit as much as possible. In the case of the 311, maybe build it on its own plug board, then put potentiometers at each input, so that you can vary the voltage at each input. Leave unusued pins open, and then see what it takes to get the thing to toggle as the input voltages are varied.

 

[unclejed613]

In addition, the only datasheets I have found were for LM311N, but I'm guessing they're all the same with respect to the LM111J

the only difference is the operating temperature range of the part. LM111 is the "military grade" part, LM311 is the "consumer grade" part.