ad: elecraft

Design of Synthesizer with integrated vco/pll

Discussion in 'Radio Circuits, Repair & Performance' started by FOVAKIS, May 17, 2013.

Thread Status:
Not open for further replies.
ad: L-HROutlet
ad: l-assoc
ad: l-Waters
ad: l-gcopper
ad: l-innov
ad: L-rfparts
ad: l-rl

    FOVAKIS QRZ Member

    Hi experts! I am making a RF transmitter [​IMG]

    and i need to make a synthesizer as a l.o to feed my mixer 2442MHz and 5dBm need. I talked to mini-circuits and the proposed to me this:

    This a chip alone. As i understand i need more "devices" to have a complete synthesizer. Can you please help me what more i need?User interface,microcontroller etc... about the ADF4118 what is this and how the chip is connected with this.? i am confused because i am noob in this field:confused:

  2. KC9UDX

    KC9UDX Subscriber QRZ Page

    No offence, but it seems you are in way over your head. I would recommend build something much much simpler first.

    FOVAKIS QRZ Member

    Hi my friend. I do not built it in my house :) I am doing my master thesis in a Laboratory...but there is no much knowledge in this. So..can you help me about the synth?
  4. KB4QAA

    KB4QAA XML Subscriber QRZ Page

    Wow. Trying to build this as first project is like having an engine block asking 'Tell me how to build an Indy Racer".

    Not saying it can't be done, but you have bit off a lot to chew. Additional skills you are going to need are Surface Mount Devices, Printed Circuit Board design (for high UHF/low SHF). Not to mention high end Oscilloscope, RF Generator, Spectrum Analyzer.

    It appears that you are intending to transmit on Channel 1 of the WiFi band. What is the intended purpose of your device? It may be easier and more practical to modify existing devices.

    Are you a licensed Amateur? If not, are you intended to operate this device under Part 15 rules? If so what laboratory do you intend to hire to conduct the certification testing?

    The good news is that the principles of your design are straight forward and apply to transmitters at all lower frequencies. If you are new to building RF circuits, you can learn much by building circuits at MF or HF, that are simpler to get working and don't require specialized equipment.

    I recommend some of the following books:
    -Any junior college text on RF circuits

    I'm sure others will have additional practical recommendations.
  5. G0HZU

    G0HZU QRZ Member

    If I was asked to comment on your block diagram I would be slightly concerned about the cost and performance impact of the IF filters.

    eg the 30MHz filter will introduce a fair bit of group delay because of the (unrealistic?) shape factor requirements to prevent transmission of the image terms. Does your transmitter have a spec for group delay and spectral purity? The 2.4GHz BPF has a 1% BW and needs to produce good rejection of the image term of the second mix.

    Are you using a PC + soundcard for the DSP? Can you not change this to produce I and Q baseband outputs and then use a simple/cheap IQ modulator chip to directly produce a signal at 2GHz? You can then lose the expensive filters and also lose the group delay issue too.

    You would still need to design a 2GHz synth for the LO feed to the modulator chip but the part you have chosen from Mccts does make this a fairly trivial design process using modern CAD tools. It looks like you need to fit a loop filter and a signal splitter to complete the synth design. You then just need to program it via a cheap and basic PIC/AVR chip.

    But my biggest concern is the block diagram of the transmitter. It looks expensive and will introduce hard to filter image terms and also high group delay in the 30MHz BPF...

    FOVAKIS QRZ Member

    Hi and many thanks about your help ! I am doing my master thesis in a laboratory of wireless communications but the one RF Engineer we had left the lab before some months and the other stuff knows little about these. I am 24 years old not a licenced amateur. We have make a 2x2 MIMO system with DSP's that work fine with cables. We want to testify this wireless. That why i make this.

    About the l.o of 30.01MHz i have found this:

    I believe is good....

    But at 2.4 ghz i must buy a syth that cost lot of money or try to make it out on my own and this is why i am asking me to help.:D:eek:

  7. G0HZU

    G0HZU QRZ Member

    Can you afford to buy a ready built 'device evaluation' PCB from somewhere like Analog Devices?
    It's just that you could effectively buy a modern test transmitter platform in the shape of the eval board for the ADRF6703 PLL/VCO/Modulator/Tx chip. The eval PCB is listed within the datasheet below and you can order it from AD. The eval board is on page 30 of the datasheet and has the modulator, the synthesiser and the output stage.

    I'm not sure if this is what you want but this would support modern (high) data rates. You feed your baseband I and Q outputs into the PCB and program up the transmit frequency via USB from a PC (using the free bundled AD software) and you then have a decent and very versatile transmitter at 2.4GHz.

    All you need to provide is 5V power, an external reference (XTAL OSC) and you would have to ensure your DSP correctly outputs I and Q info at baseband.

    This way you could test your MIMO system up at very high data rates rather than the low data rate your current system is limited to above. Also you could do the whole thing without needing a soldering iron. So you could throw away all the parts in your block diagram and just use the ADRF6703 eval PCB.

    If you say the right things to the Analog Devices sales rep you might get the eval PCB for free. At my place of work we get stuff like this free from AD but if you tell them you are doing a study for a large company sponsor etc you might get the eval board free :)
  8. G0HZU

    G0HZU QRZ Member

    Alternatively, if you really do want to stick with your current block diagram then maybe you can try and get an eval PCB for the Analog Devices ADF4350 or AD4351 synthesiser chip.

    I have eval boards for both these devices from AD and you just connect up the prebuilt eval PCB to the USB port of a PC and use the free software GUI to control the synthesiser chip on the eval PCB that tunes from 137MHz up to 4400MHz in one tiny chip with an SMA output connector for the RF output. You just connect a USB cable and run the software to get your 2.4GHz LO from the PCB!

    Otherwise you are going to have to make a PCB with a little AVR/PIC MCU to control the Mccts synth chip you plan to use. This is fairly easy to do if you have experience of building PCBs and programming up a small MCU chip. But you will need to understand how to program all the internal registers for the synth chip to get it to run correctly.

    FOVAKIS QRZ Member

    Hi GOHZU and sorry for the delay. You have already writen very interesting topics..

    <a href="" target="_blank"><img src="" border="0" alt=" photo P5181092_zps5ee33fa8.jpg"/></a>

    So i want to continue with the buliding block job in this task. So maybe a board like ADRF6703 would be a challenge for me later when i have complete this task.

    I am not interest to program any chip synthesizer because i have small experience with microcontroller so i want something almost ready to use... please can you propose me something? Frequency at ~2.4 Ghz Zone..

    The problem is that my DSP has one output port. The output signal is a 16QAM and it upconverted to 30MHz through a mixer...

    If i could have I and Q port separately from the DSP i would use maybe an I/Q Modulator to upconvert with SSB-SC.

    So, now i don't know what to do..the only good thing is that the signal's power is low -20dBm so the interferences from the mixer's ouput are low.. tell me more about the group delay of the filter and what to do to fix my problem please...

    Thanks for all George

    FOVAKIS QRZ Member

  11. G0HZU

    G0HZU QRZ Member

    If you don't want to do any microcontroller programming then one solution for the 2.432GHz LO signal is to buy the eval board for the 'all in one' ADF4351 synthesiser/LO chip from AD.

    I have one here I use as a crude signal source and you install the free control SW from the CDROM then connect with USB and you then configure the user interface to set it to a LO frequency of your own choice anywhere from about 40MHz to 4400MHz. No programming required. Just point and click to set it up on the computer screen.

    The output power is adjustable up to a few milliwatts using the software interface so you can set it to about +4dBm. You don't even need a power supply for it because you can flick the switches to make it self powered from USB. See the image below of the board I have here.


    See below for a link to the AD technical literature about the eval PCB. Note that you do not have to build this PCB yourself as it arrives from Analog Devices fully built and tested by them when you buy it from them. I don't know how much it costs but you might be able to get it for free if you are a student doing a research project on MIMO. AD might be happy to give it to you for free if you tell them it is used for MIMO research by a student at a research laboratory :)

    Your 30MHz BPF is only about 17kHz wide and needs to be a high order filter so it could be that it will have about 50 microseconds of differential group delay across its passband. This means that some parts of your signal will take 50 microseconds longer than other parts to pass throught the filter. i.e. the signals at the edge of the filter passband will take 50 microseconds longer than the signals at the centre of the filter. Will this form of time delay distortion be a problem for your system?

    FOVAKIS QRZ Member

    WOW!!! That is really cool ! I will read about this evaluation board and i will ask you in some days more things. many thanks
Thread Status:
Not open for further replies.

Share This Page