[SOLVED] Using CW as frequency synthesizer

[bolton00]

Hi,

I have a transmitter sending a continuous wave at some channel in the ISM band, 902-928MHz. Can this signal be used as a input to the mixers on the receiver, instead of an LO?

 

[jiripolivka]

For mixer LO you can use any RF signal, check if your mixer can operate with the RF power you get from the source (typically +7...+13 dBm).

The other mixer output is "RF" but you do not indicate what you use for that.

 

[bolton00]

I attached a diagram to make the problem a bit more clear.

 

[vfone]

Because is radiated, and perhaps lose a lot of power, might need to put an amplifier before LO input of the mixer.

 

[jiripolivka]

From your schematic I do not see what you expect to get at mixer IF output? There would be only a DC component due to the product of equal frequency signals generated by the mixer. If you expect some response to an obstacle in one of the two paths connected to the mixer, a better result may be obtained by using only a detector.

For very short ranges, I think Toshiba "invented" a similar concept to introduce the LO signal in the receiver along with a modulated RF carrier. The RF and LO frequencies, however, should differ to generate an IF in the receiver.

What is the purpose of your system?

 

[bolton00]

Attached is a diagram of the complete system. The continuous wave is used to power up a tag and the tag modulates this wave for a response. The receiver uses the high-power continuous wave as mixer LO input, to convert the tag response to baseband. Why not just use an internal VCO on the receiver side you might ask? Because of frequency hopping and frequency offset issues.


jiripolivka: Can you give some pointers/papers to the Toshiba invention?

 

[jiripolivka]

Thanks for the details. Then I wonder why the receiver is separated from the transmitter? If you have both (as usual) next to each other, the RF signal generated in the TX can directly be used as the LO in the RX, and even the antennas may be not separated ,(but they can if it helps). Then the mixer output will contain a component generated by a responding tag.

I am sorry, I only noticed the Toshiba "invention" several years ago in the news. It reminded me about one patent application I made in ~1970s; the referee decried my using both RF and LO to be transmitted to a remote receiver as "nonsense". My invention was not accepted but Toshiba reinvented it ~40 years later ) I have used it back then with a success in my project on mm-wave propagation.

 

[BigBoss]

First LO power that is transmitted by transmitter will have very low power level that is unsufficient to drive mixer at receiver side
Second, you should amplifiy 915MHz signal much to able to drive LO port but you LNA will amplify all band with spurious,jammers,other signals etc...

 

[biff44]

You could try to do as you drew it. But you would find that the receiver input power required for it to work would be very high--meaning that the operating distance would be very small (maybe a few meters). If you want a long range ISM communications system that might work at 100 meters, then the standard superheterodyne receiver with an on-board local oscillator is the way to go.

 

[torbai]

If your purpose is demodulation,please check the phrase "Carrier Recovery".

 

[biff44]

Not knowing what you are tyring to do exactly, I would be remiss to not point out that you can make a receiver that does not have a mixer at the front of it! Then you would not need an LO.

If you pulse amplitude modulate the transmit signal (OOK modulation), then the receiver only has to figure out the on/off transmit energy. You can use a bandpass filter at the receiver input, and some sort of analog RF detector. IF there is not too much interference at the same frequency, an Bandpass filter, LNA, and a log detector would work fairly well out to maybe 15 meters.

 

[bolton00]

jiripolivka: Using the TX signal directly as the LO input would be ideal, but in my system I have distributed receivers across a room, so it wouldn't work.

BigBoss: The transmitter is transmitting at 1 Watt, so LO power shouldn't be that much of an issue.

biff44: The problem with a simple superheterodyne structure in my case is frequency hopping. I'm in no control over the transmitter hopping, so configuring the local LO would pose a problem.
Using something like an envelope detector or log detector would simplify things immensely, but the signal I'm trying to capture is FSK modulation. Also, I believe mixer-based receivers have higher sensitivity.

 

[torbai]

The advantage of using external LO is that the carrier can match the signal well.For example, the LO of demodulation can have the same phase as that of modulation.But the disadvantage is obvious that it wastes another frequency.
There is some way to synthesize an internal LO which could match the signal as well.That is called "Carrier Recovery".So we don't need to have an external LO.

 

[bolton00]

torbai: Thank you for the pointer. Do you have an example of an RFIC that has carrier recovery capability?

 

[torbai]

HSP50210 is a Digital COSTAS Loop.You can use it to demodulate BPSK, QPSK, 8-PSK, OQPSK, FSK, AM and FM.
You should add a down-converter and an ADC before it.The LO for down-converter is not the one we need to recover.Only its phase noise should we need to care about,I think.
HSP50210 will recover the carrier which is used for demodulation.

 

[torbai]

There is an another document using a TDA8040 to form a COSTAS Loop,but it is written in Chinese.