darcyrandall2004 Post subject: Latest UHF Data Radio Schematic Posted:
Mon Sep 08, 2008 8:39 am Colonel Joined: Tue Feb
27, 2007 6:16 am Posts: 46 Hello, Attached is my latest
design for a UHF Data Radio. This time I have acquired some
professional help to answer some of my questions. The main points
that came up regarding my design were: 1. There is not enough filtering,
inductors, caps on the power supplies. 2. Remove the 10K resistors
feeding the varactors and replace with reactive components. The 10k
just adds noise. 3. Instead of simply turning on and off the POWER
AMP, I should be using a DAC to adjust the gain over 30ms b4 I turn
it on or off. Otherwise switching on and off the AMP places noise on
the power supply which in turn modulates the VCO. 4. The filter
before the antennae needs to provide an impedance match between the
amp and antennae but also it needs to reduce harmonics and spurious
emissions below -30dbm. A 2nH inductor at 2.5Amp is not realizable
5. It will be difficult to get the 120MHz crystal to oscillate,
instead I should perhaps just buy one off the shelf say from some where
like Rakon. 6. My loop bandwidth is so large, it is pointless using
Two Point Modulation and in fact the second varactor will only add noise
to the VCO. Please take a look at my design and provide some
helpful suggestions. In particular can anyone reference a decent
text that discusses in depth how to go about designing kick arse filtering
on the power supplies. Thankyou I am particularly worried about
how I am going to supply a noise free modulating signal to the XTAL
reference and the VCO. Design Notes: Phase
Noise Requirements According to some documentation I read but
now can not find, most pagers provide 65dB adjacent channel rejection
for a channel spacing of 25kHz. Using a 10dB margin, at a distance of
25kHz from the carrier, the transmitter must exhibit less than -119dBc
Single Side Band Phase Noise. PLL Design A large frequency
reference oscillator was chosen so that value of "N" could be minimised
, hence phase noise could be minimised. Fref = 120MHz. Inside
the loop bandwidth of the PLL, the phase noise follows that of the low
noise reference oscillator, outside the loop bandwidth the phase noise
follows that of the noisy VCO. A Fractional Synthesizer/PLL was therefore
chosen so that the loop bandwidth could be set as large as possible.
I chose Analog Devices ADF4154 Fractional Synthesizer. With a value
of R=10, The phase comparison frequency works out to be 12MHz. To minimise
fractional spurrs, I set the loop bandwidth to be <12/10 MHz, ie
1.2MHz. The LDO regulators used on the PLL, VCO and Reference
oscillator provide about 50dB of PSRR at 25kHz. The Single SideBand
Phase noise for the VCO was obtained from ADS simulations. I was
unable to simulate the Single SideBand Phase noise for the reference
oscillator so instead I used the Single SideBand Phase noise or the
original oscillator that I later modified for my design Ktal = 5605Hz/V
KVco=15000000Hz/V Assuming: -If I was lucky enough to achieve
a constant 80dB PSSR rejection across all frequencies, -A difficult
to achieve 10mV Noise ripple, -The PLL makes no contribution to
the phase noise -My Matlab code is correct. The simulated
results of Single Side Band Phase Noise Two point modulation
will be used so that baud rate I use is not dependent on the loop bandwidth.
Two point modulation will allow signals near DC to be passed with little
distortion. Component values for the OP amps feeding the XTAl
ref and VCO were chosen such that the modulating signal would see a
flat passband response from the input to the output of the PLL.
Thankyou! _________________ Regards, Darcy Randall, Perth,
Western Australia Top IR Post subject: Posted:
Tue Sep 09, 2008 5:02 pm Site Admin Joined: Mon Jun
27, 2005 2:02 pm Posts: 373 Location: Germany Hello Darcy,
Well Done!! There is a known circuit which is used to
filter the supply voltage of VCO's. It is made of an NPN transistor
and bypass capacitors of several values (e.g. 10nF, 100nF, 1uF ...depends
on the frequency range of the VCO) that are connected at both the NPN's
base and collector. The transistor drops the voltage from the supply
rail (Vcc) to the voltage level required by the VCO. The emitter is
connected to the VCo's supply pin. The only disadvantage is that the
Vcc needs to be higher than the VCO's voltage (the drop is the Vce voltage).
The NPN's DC operating point is set with voltage divider at the base.
I don't know if it possible to be done in your application.
I will read the rest of your post and provide more feedback
if I have. Good luck! Top darcyrandall2004
Post subject: Posted: Fri Sep 12, 2008 12:36 am Colonel
Joined: Tue Feb 27, 2007 6:16 am Posts: 46 Hello,
Thanks, but unfortunately I dont think the transmitter would meet
spec in the real world. Are you referring to using an NPN transitor
to perform the biasing of the VCO? Cheers _________________
Regards, Darcy Randall, Perth, Western Australia Top
IR Post subject: Posted: Fri Sep 12, 2008 6:29 pm Site
Admin Joined: Mon Jun 27, 2005 2:02 pm Posts: 373 Location:
Germany Yes, I am referring to that. You have open questions
in the design schematic? do you need answers for them? Top
darcyrandall2004 Post subject: Posted: Fri Sep 12, 2008
7:48 pm Colonel Joined: Tue Feb 27, 2007 6:16 am
Posts: 46 Hello, Some of the questions within the schematic
already have answers within my first post. Regardless , If you have
a different opinion I would be interested to hear it. I think
the biggest problems with the design is the power supply filtering,
matching to the antennae and providing the TTL modulating signal. Any
comments or text recommendations covering these areas would be greatly
appreciated. Cheers _________________ Regards, Darcy Randall,
Perth, Western Australia Top darcyrandall2004
Post subject: Posted: Fri Sep 26, 2008 8:38 am Colonel
Joined: Tue Feb 27, 2007 6:16 am Posts: 46 Hello People,
Do you think trying to achieve -119dBc at 25kHz Single Sideband
Phase Noise is a realistic and/or necessary target? I was looking
at these specifications for a data radio: https://www.apollowireless.com/shtml/item/FullPath/2;535;1113/ipid/1113.html
Their radio: Adjacent Channel Power : < 70dBc
I am thinking this is equivalent to saying their radio is capable
of -70dBc at 25kHz Single Sideband Phase Noise. Is my thinking correct?
Thankyou _________________ Regards, Darcy Randall,
Perth, Western Australia Top IR Post subject:
Posted: Sun Sep 28, 2008 11:28 am Site Admin Joined:
Mon Jun 27, 2005 2:02 pm Posts: 373 Location: Germany This
spec. relates to the relative signal level of the adjacent channel and
not to the noise level. I think that phase noise of -119dBc
at 25kHz spacing is not feasible.
Posted 11/12/2012
|