Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

need help on uwb synthesizer

Status
Not open for further replies.
W

wuweizi

Newbie level 4
Joined
Feb 28, 2007
Messages
7
Helped
1
Reputation
2
Reaction score
0
Trophy points
1,281
Activity points
1,338
Hi everyone,
I'm going to design a synthesizer of uwb, you know, the bandwith(7.5GHz) is too large. so it's very difficult to minize the settling time. I searched papers about uwb synthesizer, but the results are dispointed, some of them only focused on parts of bandwith, or used many plls...

Please help me if you know, or tell me something how to minize the settling time, thank you very much!

Thanks,
Alex
 

what is ur requriment
u need to design PLL with output frequency from 3 to 10 GHz , or what u need ?

khouly
 

Yes! the freqency is from 3.1G to 10.6G, and the most difficult is the settling time is less than 9.5ns!
Thank you very much!

Alex
 

but in the UWB , there is ni RF carrier , u have a pulse generation circuit , and the digital pulses just shaped , ad transmitted directly through the antenna

i think u donot need the PLL with this sever requirements

khouly
 

    W

    wuweizi

    Points: 2
    Helpful Answer Positive Rating
I too am confused. Why would you need to tune a synthesizer quickly over that bandwidth? Exactly what type of modulation are you trying to impart to the signal?
 

the 9.5ns is due to frequency hopping. from what i have seen, ppl use multiple PLLs or SSB mixer to generate the frequency quickly.
 

    W

    wuweizi

    Points: 2
    Helpful Answer Positive Rating
Thank you!
I'm not familiar with uwb, after little study I found the synthesizer is very hard to design.

kuouly, thank you! But I only know little about uwb, could you please give me the details why the pll is not needed, and what's the pulse circuit? I'm very interested in that! thank you!

biff44, I'm confused too. We just assumed the architecture is direct convertion, in fact, it's popular architecture, and the synthesizer gives I/Q LO signals, that's all, the base frequency is about several hunded MHz. Maybe I made a mistake, could you give me more suggestion? Thank you!

szekit, thank you, I just want to know, is there any chance to design a simple pll to realize it? If so, I'll try it, if not, I don't want to waste time on it, I'll just design it use many plls and ssb!
Thank you!

Best,
Alex
 

There are two kinds of standards about UWB. One is Wimedia/MBOA, which needs fast setting PLL. The other one is Pulse based UWB, which needn't PLL.You can read some papers from ISSCC2004-2006 to study that.
 
  • Like
Reactions: tony_lth

    W

    wuweizi

    Points: 2
    Helpful Answer Positive Rating
    tony_lth

    tony_lth

    Points: 2
    Helpful Answer Positive Rating
Well, forget about a PLL settling in 7.5 nS! You would need a control loop bandwidth of around 500 MHz, which is probably impossible with standard PLL technology.

So, it sounds like you need some sort of "direct synthesis" (NOT to be confused with DDS synthesizer). A number of CW sources, a matrix of switches, and some mixers to mix them all together. If you make the switches ultrafast (say 1 nS type) and refrain from using any filters, you could get 7.5 nS settling time. You might be able to do much of it monoliithically, using some sort of cmos crosspoint switch chip, and then multiplying the frequency up. I suspect that there will be a tradeoff between settling time and spurious output (since you will not be able to use any narrowband filters to clean up spurs).
 

    W

    wuweizi

    Points: 2
    Helpful Answer Positive Rating
Thank kolarbear! Of course, it's MB UWB :) Actually, I'm reading these papers!

Added after 9 minutes:

Thank you, biff44! Actually, I'm trying in this way! because the very short settling time, there is no way to tune the vco frequency, so the only way is to use mixers and selectors,etc to generate the 14 bands frequencies!
Yes, I'm worring the spurious too, you know, the ssb mixers are too big,so if possible, I'll use two or three plls to generate it, then the spurious output will be better! right?
Thank you again, you are a great synthesizer expert! May I ask you more detail questions when I design the circuit? Thank you!

Best Regards,
Alex
 

Does the phase noise have to be good, or can it be poor? There are millimeter wave or optical methods that might work, but will be fairly noisy.
 

it's a wide band channel ~500Mhz. so noise is not as strigent.
 

    W

    wuweizi

    Points: 2
    Helpful Answer Positive Rating
Thank biff44 and szekit, I hope the noise is relax, if too tight, it's more difficut!
 

any body here

i need help plz i wana david m pozar manual for 3rd edition plx some body help
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top