I am trying to build simple receiver and transmitter that does as follows: at the transmitter end detects not enough daylight and motion to send signal to the receiver to close the switch. Keep the switch close at receiver until transmitter end detects no motion or enough daylight to send signal to open the switch. Keeps it open until transmitter sends signal to close based on daylight and motion.

It could be as simple as AM or FM transmitter/receiver. It is short disrance signal travel through brick and thru some conrete. The transmitter is outside the home while receiver is inside the home next to light switch. I do not want any rf pre built module or any ics etc. want to built it from transistors, resistors, indictors, capacitors etc.

With all the wireless signal how does receiver know to pickup correct signal recognize yes it is the signal intended for the receiver and decode it? What is the main principle and logic behind it? Also how can I build such a circuit as described above?

 

Mechanisms such as garage door openers use a 8 section DIP switch to set a digital recognition code string .

 

The transmitters sends a code in some digital format, and the receiver then decodes the signal and looks for the coded format.
As Max, noted, old garage door openers often used a 8-bit digital code, giving 256 possible codes (modern openers usually use a rolling code with thousands of codes, to avoid thieves using scammers which would automatically sequentially try all codes to open the door).
Building such a circuit is not trivial, so is typically purchased as a dedicated IC designed for that purpose.
But if you want to reinvent the wheel, have fun.

 

There are two conditions the receiver will accept the signal:

1) Matched frequency between receiver and transmitter. Even 2MHz difference in frequency at 100Mhz can/may cause the receiver will not accept the transmitter signal.

2) Coding. There are sooo many codes that can be implemented: analog levels, timed sequence, digital coding, even combinations of all mentioned.

 

Do you know how an AM radio can tune one of many stations? That would be the first (but not last) level you would need to do this.

Think of what it sounds like when you tune between stations. That is what is called noise and it is the next level you would have to deal with.

If you have a transmitter transmitting at a frequency carrying no information, it will make an AM radio tuned to it go quiet. That could be enough for your simple on / off remote.

But, for a a reliable system, one needs to go beyond that.

Sending a specific audio frequncy coukd be the next iteratuon.

Beyond that, you would want to send digital data. Without ICs that would be way too difficult.

 

Yes, there are also simple analog codes, such as sending a particular sinewave or DTMF frequency.

 

There are two conditions the receiver will accept the signal:

1) Matched frequency between receiver and transmitter. Even 2MHz difference in frequency at 100Mhz can/may cause the receiver will not accept the transmitter signal.

2) Coding. There are sooo many codes that can be implemented: analog levels, timed sequence, digital coding, even combinations of all mentioned.

How do you do all that? Any circuits you can give??

 

I am just looking for circuit that detects daylight and motion and sends signal wirelessly to receiver to turn 120vac light on if their is no daylight and motion otherwise light off.

Know any circuit that can do that?

 

One solution is to buy a wireless doorbell. You will have to figure out the light and motion sensing part.
At the receiving end, you can activate a relay to turn on the 120VAC light.

 

How do you do all that? Any circuits you can give??

It sounds like you want to do this all without digital IC's in which case I'd say your in your some trouble. This kind of puts you back to the spark gap transmitter days. I think Bob's idea in post #5 is a good place to start because you can use Frequency Shift Keying to transmit binary (digital) data over an analog carrier. The trouble is generating the carrier and transmitting it. Unless you tune the transmitter and receiver, it will require a lot of power to penetrate the walls you mentioned. It's also possible you may violate local laws and cause interference to your own devices. Still interested?

 

It sounds like you want to do this all without digital IC's in which case I'd say your in your some trouble. This kind of puts you back to the spark gap transmitter days. I think Bob's idea in post #5 is a good place to start because you can use Frequency Shift Keying to transmit binary (digital) data over an analog carrier. The trouble is generating the carrier and transmitting it. Unless you tune the transmitter and receiver, it will require a lot of power to penetrate the walls you mentioned. It's also possible you may violate local laws and cause interference to your own devices. Still interested?

I dont follow. Its just sending on and off signal. The transmitter would need batteries and receiver can get power from 120vac. How would circuit look like?

 

Its just sending on and off signal.

For a simple design I implemented some time ago, I learnt how easy was having interference (environment) or noise interpreted as the above. Not what you will appreciate for sure.

 

To expand upon @K1ng 1337’s expansion on my post, here us a proposed solution:

AM modulated transmitter that sends a short burst of 1000Hz for ON and 4000Hz for off.

Receiver and two filters, one for each modulation frequency. Peak detector at output of each filter.

A flip flop or latching relay controlled by the two peak detectors.

This would be a nice project for someone with several years experience designing their own circuits.

 

I dont follow. Its just sending on and off signal. The transmitter would need batteries and receiver can get power from 120vac. How would circuit look like?

If you were to build the circuit from scratch, I think it would need at least 4 major parts: 1) Modulator 2) Transmitter 3) Receiver 4) Demodulator. To demonstrate Frequency Shift Keying, I built the 1 and 4 in LTspice. Interestingly, almost any kind of transmitter / receiver can be used because the same principles apply. I designed a similar circuit to use laser light and a photodiode for #2 and #3 but you could adapt this circuit for radio antennas. I don't know much antenna theory so hopefully the lads can do better. Anyway, here's how my circuit works:

1) DATA is simulating some kind of 5V digital data. This could from a micro-controller or from a daylight / motion sensor. I have it pulsing ON and OFF every 2 seconds.

2) R1, R2, R4, R3, C1, C2, Q2 and Q1 form a typical astable multi-vibrator circuit. When DATA is LOW, the frequency at A is ~483Hz. When DATA is HIGH, the frequency changes to ~350Hz. The shift happens because DATA and R7 affect the parameters of the oscillator altering the frequency (and duty cycle). Essentially we get a circuit with two distinct states. Next, R6, Q3 and Q4 amplify the output of the oscillator. The circuit up to this point forms the Modulator and the MOD output could be connected to just about anything. But for the sake of my simulation, I have it coupled directly to the Demodulator.

3) The MOD signal comes in (could be a photodiode or antenna) and is filtered by a parallel LC tank circuit C3, L1, and R8. These three components form a band-pass filter around 350Hz. Because 483Hz is much greater than 350Hz, the 483Hz signal is greatly attenuated compared to the 350Hz signal. The signal at this point is AC, so I used a peak detector (D1 and C4) to convert the signal to DC. Finally, R9, Q5, R5 and Q6 reshape the signal to a nice square wave and the original data is restored!






This third image is pretty cool. It shows the transition of DATA from LOW to HIGH. If you look closely, the 483Hz signal has a voltage of +-0.4V and the 350Hz signal +-0.9V. This is part of my design because it takes at least 0.6V to turn on Q5. The utility of this method comes from the final output only going HIGH when there is a 350Hz signal detected. All other signals are shunted by the inductor. You only need two states but this method can be applied to any number of inputs.