Using VCO to drive an antenna

Hi, I am trying to design a simple 2.4GHz OOK transmitter, where the on-off rate is only about 1kHz.

My idea was to use the MAX2750 voltage controlled oscillator. It has a shutdown pin, so I figured I could just toggle that to provide the modulation. Interestingly, it states that it is output buffered and matched to 50Ω. Does this mean I can directly link it to the antenna?

https://datasheets.maximintegrated.com/en/ds/MAX2750-MAX2752.pdf

The second harmonic level is about -33dBm so it might not pass spurious emissions testing for FCC or european approval.
If EIRP is measured it might pass or fail depending on antenna cable loss and antenna behaviour at 4.8GHz.

If you are only making a few of them they chance of anyone complaining about the cleanliness of the output is very small.

I'm using it for a physics demonstration in an E&M class, so I'm not too worried about spurious emissions. If I WAS worried, what would be a good way to remove harmonics? Some kind of bandpass or lowpass filter? Forgive me if this is an obvious question, this is my first time working with RF.

Just looked on Digikey. Would something like this work well?

https://www.digikey.com/product-detail/en/LFB182G45SG9B740/490-4743-1-ND/1531439

VCO's are good sources to have an oscillator quickly but they have stability problems. Although you keep the tune voltage constant, the output frequency drifts unpredictably. This is a well-known problem and PLL circuits(along with VCOs) are employed to keep the output frequency stationary. Nevertheless, a VCO may work for your experimental purposes. Whenever the output frequency drifts to outside of your operation band, readjust its tune voltage and make your demonstration. The amount of the frequency uncertanity depends on the VCO properties, a narrow band VCO is expected to deviate less.
I think harmonics would not be problem because your antenna operating at 2.4GHz will not permit them to radiate, it works like a harmonic filter also.

OK, I just took a look at how PLLs work, and I'm a bit confused…

My understanding is that a PLL takes two oscillators, 1 and 2, and then compares the phase. As 1 goes faster than 2, the PLL generates a linearly increasing voltage output (after low passing). This all makes sense.

What I don't get is this: How can a VCO with a PLL control its own frequency? I'm pretty sure using itself as a reference wouldn't work. Does it need an external frequency generator? And what's to stop that external generator from drifting as well?

EDIT: I think I may have just understood this a bit better. If you take the VCO output at say 2.4GHz, and divide it by a large integer, that would bring it into the range of a crystal oscillator. You could then use the crystal as the reference, and just compare phase at every 256th cycle or something like that, right?

I guess my only concern then would be that you could have a lot of jitter and wind up a whole cycle ahead/behind, locking the wrong frequency. Or is that not an issue?

Yes, there is a reliable reference oscillator whose output compared against to divided version of the main output continuously in a control loop. As far as I know, XO, TCXO, OCXO are known reliable sources. There would be valuable tutorials about VCOs, PLLs on the web. As long as you know the divison ratio and the reference frequency, you would not see a wrong frequency at the output. Phase noise i.e. jitter in time domain is an inevitable problem, it is also one of the most important performance criteria for an PLL+VCO oscillator. The phase noise is important for phase sensitive applications.
Today, it is possible to find a single chip that VCO and PLL are integrated in it such as ADF4351. In addition to the chip, you basically need a reference oscillator and a loop filter to work your oscillator. Reading their datasheets may also help.

Yes, I've seen many of those chips at this point. However, I'm trying to build this transmitter without using a microcontroller (all the integrated PLL/VCOs seem to have digital settings). In other words, I'm trying to design it so you can just solder the parts to a pcb and turn it on. The reason is that I expect that people with no experience in electronics or code-writing may need to build this device.

With that in mind, are you aware of any VCO/PLL devices with non-code requiring settings? I.e. you set some resistor and it makes the frequency you want.

In the event that no such thing exists, I'm coming back to just using a VCO by itself. I'm wondering - how bad is the frequency drift likely to be? +/- 10MHz? 20MHz? Worse? I figure if it's less than 50-100MHz than I should be OK.

Finally, thanks for all your comments so far, you've taught me a whole lot of new things!

Amount of the frequency drift depends, I have used minicircuits' ZX95-520+ VCO, I was able to make a narrowband FM modulation demonstration which uses 10kHz BW, I was listening with my amateur radio, it was fine for 2-3 minutes. If you would like to control your circuit via a simple parallel interface, I think you need to look for older chips and possibly you may not find a complete/integrated solution. I have seen freq. synt. circuits controlled via parallel interface but their components were discrete, PLL, VCO, N Counter. You may find a discrete N counter having parallel interface.
You will eventually have a VCO. Examine it, if it does not work for you may look for counter and PLL ICs.

psemi has PLLs that you do not have to further program.
You install pull up or pull down resistors to each control bits and than it will lock to a constant frequency. you can use some DIP switches for different frequencies.

Axemaster said:

Yes, I've seen many of those chips at this point. However, I'm trying to build this transmitter without using a microcontroller (all the integrated PLL/VCOs seem to have digital settings). In other words, I'm trying to design it so you can just solder the parts to a pcb and turn it on. The reason is that I expect that people with no experience in electronics or code-writing may need to build this device.

With that in mind, are you aware of any VCO/PLL devices with non-code requiring settings? I.e. you set some resistor and it makes the frequency you want.

In the event that no such thing exists, I'm coming back to just using a VCO by itself. I'm wondering - how bad is the frequency drift likely to be? +/- 10MHz? 20MHz? Worse? I figure if it's less than 50-100MHz than I should be OK.

Finally, thanks for all your comments so far, you've taught me a whole lot of new things!

Click to expand...