3rd south Italy homebuilders meeting (MAdS) and 22nd south Italy V-UHF meeting
Acicastello (Catania province), Sicily, Italy, JM77NN, E15.1268/N37.56036, April 17th and 18th, 2010.
Phase noise measurements using amateur equipment
by Eraldo Sbarbati I4SBX (Eraldo Sbarbati I4SBX is SK since August, 23rd 2014)
I already had: the Intermodulation, the cross-modulation, the thermal noise, the QRM, the QRN
and so on. Now you bring me also the phase noise: if you know it, you can avoid it.
What is the "phase noise"
The signal generated by a real oscillator doesn't consist of a single spectrum line, but
is a set of rows, with decreasing amplitudes moving away from the central frequency f0.
If we could measure the frequency of each sinusoidal component, we would see that this value is not
constant and equal to f0, but the frequency values of each period are arranged in a stochastic way
on a bell with the summit centered on f0 and with a slope similar to the quality factor Q.
At the base of the spectrum we find the usual thermal noise
Pn= KTB [k= 1.38*10E-23 j/°K]
Pn= 290°K*1.38*10E-23 J/°K*1Hz= 4*10E-21W= -174dBm/Hz@17°C for a BW= 1Hz
[see I4SBX Radio Rivista 11/2006 page 101; ARRL Handbook 2005 Fig. 10.7]
Some effects of phase noise
Disturbs nearby communications. Increases the uncertainty in the RADAR results.
Produces reciprocal mixing in receivers.
Reciprocal Mixing: (see figure on page 4 of the original paper)
Red line, Icom IC-706 IF response in SSB.
Blue line, effective response due to the VFO noise.
Green line, improved response with an improved VFO.
How to represent the phase noise
(see the figure on page 5 of the original paper)
on the X-axis, the offset in Hz.
on the Y-axis the dBc/Hz.
How to represents the phase noise
(see the figures on page 6 of the original paper)
first figure is © ARRL QST March 2007
second figure is © Funkamateur 1/2010
Measurement of phase noise with amateur instrumentation
(see the picture on page 7 of the original paper)
This is an amateur piece of equipment, belongs to Ulrich Rohde N1UL.
refurbished R&S FSUP are on offer from 39000 to 59000 EUR.
On the screen you can see the phase noise of the ADAT-200A -134dBc/Hz@1kHz
How to measure the phase noise
-With bench equipment and appropriate set-up. [notch, PLL, discriminators, delay-lines, ...]
-With dedicated equipment such as the example in the previous page (the R&S FSUP)
-By measuring the effects, such as the reciprocal mixing
-With a spectrum analyzer
-With FFT systems
-managing it out, someway
An example: (see the figure on page 8 of the original paper)
phase noise measure by means of a spectrum analyzer with automatic filter compensation
red Line: IF filter and analyzer noise
the measure is limited to -95 dBc/Hz
Images taken with pic-plot © Hotwater
Phase noise measurement by a spectrum analyzer
(see the figures on page 9 of the original paper)
Same plot of the previous page with the reference frequency axis to the left.
first figure:
100Hz/div
BW filter 10Hz so subtract 10dB/Hz@1Hz
measurement limit is -100dBc/Hz
red Line: analyzer noise
second figure:
10kHz/div
BW filter 1KHz so subtract 30dB/Hz@1Hz
measurement limit is -120dBc/Hz
Simplified diagram of a spectrum analyzer
(see the figure on page 10 of the original paper)
1) A spectrum analyzer is similar to a receiver, with a variable LO, without AGC,
and generally with the output represented in logarithmic form (dB) on a monitor.
2) The noise of the analyzer's LO affects the quality of the measure, therefore
the LO must be better than the signal under measurement.
FFT (Fast Fourier Transform) spectrum analyzer
(see the figure on page 11 of the original paper)
An FFT spectrum analyzer is usually realized by a numeric computation, DSP, on the input
signal's samples. We can think of it as a multitude of frequency filters (bin) one near the other.
Spectrum Lab - DL4YHF (HP-651B)
(see the figures on page 12 and 13 of the original paper)
PC audio cards features
card input type max input mV noise dBc/Hz
Sound Max [MME 16 bit] Mic. 100 -118
Sound Max [MME 16 bit] Aux (line) 400 -125
M-Audio Delta 44 [ASIO 24bit] -10dBV 1000 -141
M-Audio Delta 44 [ASIO 24bit] +4dBu 4000 -147
E-MU 1212m [ASIO 24bit] -10dBV 1200 -158
E-MU 1212m [ASIO 24bit] +4dBu 5000 -159
E-MU 0404 [ASIO 24bit] Vol_Max. 3,4 -120
E-MU 0404 [ASIO 24bit] Vol_Min. 2300 -151
E-MU 0202 [ASIO 24bit] Vol_Max. 3 -120
E-MU 0202 [ASIO 24bit] Vol_Min. 2240 -151
* The cards Delta 44 and E-MU have differential inputs and sample rate up to 96-192kB/s
* E-MU-0404 & E-MU 0202 are USB soundcards
Mixer + Spectrum Lab
(see the figure on page 15 of the original paper)
Both oscillators, at the mixer's inputs, are contributing to the noise so you have to use:
1) a reference oscillator better than the DUT
2) a reference oscillator equal to the DUT by subtracting 3dB to the Phase-Noise measure*
* For simplicity, the oscillators' contribution to the noise is assumed to be the same.
Double Balanced Mixer (SRA-1 & equivalent)
(see the schematic on page 16 of the original paper)
Low noise amplifier (LT6231, OPA2228, NE5532, [AD797] and others)
(see the schematic on page 17 of the original paper)
if the differential output is not necessary then U1B can be used to generate Vcc/2
Dual balanced low noise amplifier OPA1632
(see the schematic on page 18 of the original paper)
OPAmp gain for the professional soundcards with -10dBV inputs
mixer type LO dBm 1dB compression dB loss dB output level mV gain min for 1,2V dB
SRA-1 7 1 6 126 20
SRA-1H 17 10 6 354 11
RAY-1 23 15 6 630 6
SAY-1 23 20 6 1112 1
NE602 or SA602 or SA612 mixers
(see the schematic on page 20 of the original paper)
Soundcards noise Sound-Max / E-MU 1212m
(see the figure on page 21 of the original paper)
Marconi 2019A (HP-8640) Sound-Max (C-Media)
(see the figure on page 22 of the original paper)
PLL IC-706 Marconi 2019A
(see the figure on page 23 of the original paper)
2xDDS (12bit DAC)
(see the figure on page 24 of the original paper)
4046-9046 Osc. UHF / PLL ~ 6.4MHz
(see the figure on page 25 of the original paper)
N1UL's R&S FSPU (Phase-Noise ADAT-200A)
(see the figure on page 26 of the original paper)
2x Xtal Osc. & 2x 2019A (f0 27MHz)
(see the figure on page 27 of the original paper)
2x Xtal Osc. & R&S SMFS (f0 27 MHz)
(see the figure on page 28 of the original paper)
3rd south Italy homebuilders meeting
This is what I was able to get with a few euro and a lot of patience.
Thank you for your attention
73 de
I4SBX Eraldo