Add A SLIM Tracking Generator to the
SLIM Modularized Spectrum Analyzer to
create the SLIM MSA/TG
    You may have become familiar with the Basic SLIM MSA, which is presented on the SLIM MSA Page.  This is a guide for adding a tracking generator to the  0 to 1000 MHz, SLIM Modularized Spectrum Analyzer.

This Page was Started Oct 16, 2007.
Updated Oct 21, 2007 Added Wiring Diagram and Layout

What is a Tracking Generator?
    A tracking generator (TG) is an RF signal generator that will change frequency at the same rate as the Spectrum Analyzer (SA).  When the SA is commanded to look for a particular frequency, the TG output frequency will be that same frequency.

Block Diagram of Basic MSA, using SLIMs

    This is the same Block Diagram as the one on the Construction page for the Basic SLIM MSA.  The only frequency agile oscillator is LO 1, a hybrid synthesizer composed of PLO 1 and DDS 1.  The input signal is mixed with LO 1 to create a difference frequency of 1013.3 MHz.  In other words, LO 1 will always be set 1013.3 MHz above the input frequency.  LO 1 will tune from 1013.3 MHz (which corresponds to a 0 MHz input signal) to 2013.3 MHz (which corresponds to a 1000 MHz input signal).  This 1st IF signal is tightly filtered by the cavity filter and then presented to the second mixer.  Here, it is mixed with the LO 2 (1024 MHz) to create the final IF of 10.7 MHz.  The final IF is amplified and filtered and presented to the Log Detector for power to voltage conversion.  The Log Det output voltage is sent to the Analog to Digital converter to be read by the Home Computer (PC).
    The basic MSA will read absolute Magnitude (power) measurements vs. frequency.  It can make Relative Power measurements if an external Frequency Source is used.  If the external Frequency Source can be changed at the same rate as the frequency agile oscillator in the MSA (LO 1), then the MSA can make Relative Power measurements over a very wide frequency range.
    A Tracking Generator is the external Frequency Source.  It's frequency is changed by using the same Controller that changes the frequency of LO 1 in the MSA.  This sounds simple enough, but a Frequency Source that will tune from 0 MHz to 1000 MHz is not quite that easy.  There is no single oscillator that has that frequency range (yet).  Therefore, the Tracking Generator must use a frequency mixing scheme to create the frequency range of 0 to 1000 MHz.

The SLIM Tracking Generator
MSA with Tracking Generator Block Diagram, SLIM MSA/TG
new trk gen block diagram
    The Tracking Generator addition is composed of PLO 3, DDS 3, Mixer 3, and connections to the Control Board, Master Oscillator, and PLO 2.  It's mixing scheme utilizes the MSA's LO 2 fixed output frequency of 1024 MHz and the generated frequencies of the PLO 3 / DDS 3 hybrid synthesizer (LO 3).  The Tracking Generator's LO 3 and the basic MSA's LO 1 are identical in components and construction.  Mixer 3 is identical to Mixer 1.  Click on the above modules to take you to the construction page for each module.
    LO 3 will command from 1024 Mhz to 2024 MHz.  The +7 dBm output from LO 3 is used as the LO drive for Mixer 3.  The fixed frequency of 1024 MHz (-4 dBm) from LO 2, is input to the R port of Mixer 3.  Mixer 3 output (I port) will range from 0 MHz to 1000 MHz, thus, "tracking" the MSA at a level of approximately -10 dBm.  As a side note: if you plan to convert the MSA/TG to a MSA/VNA, build the PLO 3 with both buffers.
    Normally, the LO 1 frequency and LO 3 frequency will always have a differential frequency equal to the frequency of the Final I.F.  However, the software can be told to "offset" the Tracking Generator output.  This is particularly useful when injecting a certain frequency into a test device and a different frequency is the output of that device (such as, a transciever).
    Notice, that the Mixer 3 module has no internal low pass filtering on the TG output.  Therefore, the TG output will have two mixing products, 0 MHz to 1000 MHz, the difference product, and 2048 MHz to 3048 MHz, the sum product.  The sum product will be attenuated some, due to the poor high frequency response of the I port of the mixer.  However, if this sum product interferes with Tracking Generator usage, a low pass filter must be used on the TG output.

Modifications to the Basic SLIM MSA when adding the TG:
    The Tracking Generator requires power and signal commands from the SLIM Control Board.  The SLIM-CB-NV, Control Board, has been built with these connections in place; therefore, no mods are necessary to the Control Board.
    The Tracking Generator requires a Clock from the Master Oscillator Module.  It is possible that the SLIM-MO-64, Master Clock Module, was built with with only two buffers.  If so, a third buffer must be added.
    The Tracking Generator requires a 1024 MHz signal at -4 dBm from LO 2.  It is possible that the SLIM-PLO-2, PLO 2 Module, was built with only one buffer section.  If so, the second buffer must be added.

Additional Coaxial Interconnections for the Tracking Generator addition to the Basic SLIM MSA
Signal Name
 From
 To
 Recommended
Cable Type
 Approximate
Length, Inches
Master Clock 3 SLIM-MO-64, J3 SLIM-DDS-3, J1
 RG-085
 1.8
DDS3OUT SLIM-DDS-3, J4 SLIM-PLO-3, J1 RG-085 or RG-188 1.6
LO3
 SLIM-PLO-3, J3 SLIM-MXR 3, J1 RG-085 2.5
LO2
 SLIM-PLO-2, J2 SLIM-MXR 3, J3 RG-085
 5.0
Trk Gen Output
 SLIM-MXR 3, J2 Front Panel
 RG-085 or RG-141 4.0


Wiring Diagram for the SLIM MSA/TG
    The following diagram consists of all the Power and Control signals for the SLIM MSA/TG.  For RF and coaxial connections, refer to the MSA with Tracking Generator Block Diagram, shown previously.
wiring diagram for slim msa/trk gen

Layout for the SLIM MSA/TG
    The following layout is proposed for integrating the Tracking Generator into the Basic MSA.
layout of slim msa and tg

Some Notes on the SLIM MSA/TG Combination
1.    I would like to point out to the reader, that, the absolute frequency of the MSA and the Tracking Generator Output will not be precisely, the same frequency as the software command, even if the Master Oscillator is precise.  This is because a DDS is used as the steering frequency for the MSA's LO1, and the TG's LO 3.  The DDS is quite accurate, but it is not "phase locked" to the Master Oscillator.  It is treated as a frequency divider, and the software will command the DDS to the closest significant bit.  Therefore, the DDS's output frequency will be accurate within one bit, or 0.014 Hz.  This seems pretty good, but remember, the DDS output is "multiplied" in the PLO by anywhere from 93 to 190 times.  This equates to a frequency resolution of 1.3 Hz to 2.6 Hz when the PLO is tuned from 1000 MHz to 2000 MHz.  For use in a Spectrum Analyzer or properly designed into a Vector Network Analyzer, this is not of concern.  But, it is just a general warning that the Tracking Generator is not a precision frequency source.  With certain software conditions, it can be made a precision step frequency source.  However, the MSA does not do that.  I just wanted to make this statement before someone emails me and says, "Hey Scotty, you jerk, I commanded the MSA to zero MHz and the Tracking Generator has an output of one fifth Hertz, what gives?"
2.    This Tracking Generator scheme does have the probility of producing interference within the MSA.  There are two interfering frequency sources.
    One is the Tracking Generator Output frequency created by Mixer 3.  I will call this TGO.  We would like all of the TGO to exit Mixer 3 at the I port.  But, some of the TGO (0 MHz to 1000 MHz) will reflect back out of the R port of Mixer 3 and enter PLO 2 at J2.  The TGO will pass through the PLO 2 and exit J3.  Although highly attenuated, the TGO will enter Mixer 2, via the L port, causing some interference with the mixing action of Mixer 2.  This interference should only be noticeable when the Tracking Generator is commanded to the same frequency as the Final I.F (10.7 MHz).  However, since the buffer amplifiers have excellent isolation at low frequencies, the interference should be minor.
    The second interfering source is a bit more complex.  And, I didn't notice it in my original test environment.  This source is a combination of LO 1, LO 3, and Mixer 2.  Here is how it is created:
    All mixers have a certain amount of port to port isolation.  Some of LO 1 will pass through Mixer 1 from the L port to the R port.  The cavity filter will allow frequencies from 1013.3 MHz to about 1015 MHz to freely pass through it and get to Mixer 2.  The roll-off of the cavity filter will attenuate frequencies above 1015 MHz.  But, we have some of the LO 1 present at the R port of Mixer 2.
    The same port to port isolation of Mixer 3 will allow the LO 3 energy (1024 to 2024 MHz) to pass through from the L port to the R port.  This signal will travel like the first interfering souce, described above.  Although there is isolation in the PLO 2 module, some of the LO 3 energy will pass through the PLO 2 module and be present on the L port of Mixer 2.  Now, we have some of LO 3 on the L port of Mixer 2.
    Mixer 2 will mix the two interfering signals, LO 1 and LO 3, and output a differential frequency at it's I port.  As the MSA/TG is swept, the two frequencies are always 10.7 MHz in difference.  Of course, this 10.7 MHz difference is exactly what the MSA's Final I.F. is tuned to.  This interference is prominent only when the MSA/TG is at low frequencies.  The bandwidth of the cavity filter will determine the cut-off frequency of interference.  In my MSA/TG the interference becomes unnoticeable when the MSA/TG is tuned above 3 MHz.
    The level of both of these interferences are dependent on the port to port isolation of the mixers when their L port frequencies are close to 1000 MHz.  The ADE-11X mixer isolation is rather poor here.  Minicircuits specs it much better than it actually is.  I found this out after much testing.  If the builder would like better interference performance, Mixer 1 and Mixer 3 could be replaced with mixers of higher quality (and expense).  If so, try to find a mixer with L to R port isolation of greater than 35 dB at 1000 MHz.  My SSAProto used a WJ-M1J for both Mixer 1 and Mixer 3 and had isolation greater than 42 dB.  I never saw any interference.  (M1J's are not cheap).
    I am showing a Tracking Generator Output of -10 dBm, but that is dependent on the Mixer 3 loss, and the originating power level of the 1024 MHz sample from LO 2.  This is much more power than is usually required to make reference magnitude measurements.  Therefore, the attenuation of LO 3 to the Mixer 3 could be increased to reduce the Tracking Generator output.  This has the benefit of reducing the amount of both types of Tracking Generator interference.

3.    This Tracking Generator addition can be used as a frequency source for other experiments when the Spectrum Analyzer portion is not being used.  Simply command the MSA to a stationary frequency with the Sweep at 0.  Command the "Track Gen On", and the TG will output the same frequency that the MSA is tuned to.  The range is from 0 Hz to over 1000 Mhz, possibly as high as 1200 MHz.

4.    This Tracking Generator addition can be used as a stand alone UHF synthesizer to create frequencies from about 950 MHz to 2050 MHz.  If the PLO 3 is totally populated with components, an unused output is available at PLO 3-J2.  Of course, the MSA's PLO 1 has a spare port, also.  It is J2 and can be used as a stand alone UHF synthesizer.

5.    The Tracking Generator is never actually turned off.  When the software commands the "Track Gen Off" it simply commands the Tracking Generator's LO 3 to idle at a frequency that will not interfere with normal MSA operation.  At the present time, that frequency is 970 MHz.  The Tracking Generator will output a frequency of  970 +/- 1024 =  54 MHz and 1994 MHz.  And, it can be changed in the Global Variables portion of the software code.  Look in the software line of code called:
PLL3off = 970       'Output freq of PLL3/VCO3 when Trk Gen "off"
If  970 MHz is a bug for the user, he can change the (970) to anywhere between 950 and 2030.  This is where the LO 3 will idle, in MHz.  When the operator commands the "Track Gen On", the TG will command to the same frequency that the MSA is tuned to.

6.    The Mixer 1 input and the Tracking Generator Mixer 3 outputs are not 50 ohms over frequency (close, but no brass ring).  They can be brought closer to 50 ohms by adding 10 dB pads to the ports.  With the high level of -10 dBm output of the TG, there is still plenty of power available (-20 dBm), after the addition of the 10 dB pads.

Scotty Sprowls, email is wsprowls (at) yahoo.com
End of page (so far).