The SLIM MSA
Modularized Spectrum Analyzer, using
Standardized Lab Integration Modules (SLIMs),

with Tracking Generator and
Vector Network Analyzer extension
slim/msa_pic_front.gif slim/msa_pic_top.JPG slim/msa_pic_bot.JPG

Updated 6-03-08.  Add daisy chain warning for Wiring Diagram
Updated 6-13-08.  Updated, due to SLIM Module revisions:
SLIM-DDS-107:  SKSLIM-DDS-107 revC, PLSLIM-DDS-107 revC
SLIM-PLO-1:  SKSLIM-PLO-1 revB, PLSLIM-PLO-1 revB
SLIM-PLO-2:  SKSLIM-PLO-2 revB, PLSLIM-PLO-2 revB
SLIM-PLO-3 has been deleted from the MSA.  SLIM-PLO-1 is used in PLO 3 position.
SLIM-MO-64:  SKSLIM-MO-64 revB, PLSLIM-MO-64 revB
SLIM-MXR-1:  SKSLIM-MXR-1 revA, PLSLIM-MXR-1 revA
SLIM-MXR-2:  SKSLIM-MXR-2 revB, PLSLIM-MXR-2 revB
SLIM-MXR-3:  SKSLIM-MXR-3 revA, PLSLIM-MXR-3 revA
SLIM-MXR-4:  SKSLIM-MXR-4 revA, PLSLIM-MXR-4 revA
Updated 7-14-08.  Updated, and changes, especially: "A to D Converter, using either a 12 or 16 Bit Serial A to D Converter:"
Revised: PLSLIM-ADC-12 and PLSLIM-ADC-16. Three-way switch PN's were incorrect.
Updated 7-30-08.  Fan option.  See Addendum 1. at bottom of this page.

    The SLIM version of the Modularized Spectrum Analyzer is functionally identical to the Original MSA, but, is constructed using SLIMs.  It's operating range is 0 to 1000 MHz.  This page will describe the MSA construction, with block diagrams, a list of required modules, and a recommended integration layout.
    A Tracking Generator can be added to the Basic MSA, and this page will show the additional three SLIMs required for the SLIM MSA/TG.
    The SLIM MSA/TG can be further extended to become a Vector Network Analyzer, the SLIM MSA/TG/VNA.  This page will show those two additional SLIMs.
    This page will contain paragraphs to describe the necessary modifications to each SLIM, to integrate it into the SLIM MSA system.  Those paragraphs will contain links to the pages that describle the SLIMs in full detail, with documentation.  SLIM documentation includes everything needed to build a SLIM.  Here is the main SLIM Web Page, which describes the SLIM system and all of the SLIMs that have been designed.
    This new SLIM MSA effort does not obsolete my original MSA design.  It is just a new way to do things.  If you are interested in building only the Basic Spectrum Analyzer, simply disregard the additional requirements to add the Tracking Generator or VNA.

Jump to major items on this page: (use browser BACK button to return here)
Block Diagram for MSA/TG/VNA
List of SLIMs used in the
MSA/TG/VNA
Wiring Diagram for Basic MSA
Wiring Diagram for MSA with Tracking Generator
Wiring Diagram for Basic MSA with Tracking Generator and VNA
Layout of the MSA/TG/VNA using SLIMs
List of Coaxial Cables
MSA Specifications
Signal Flow for MSA
MSA Analysis

Here are other links supporting the SLIM MSA : not all complete and linked yet.
Printed Wiring Boards for the SLIM MSA and Tracking Generator and VNA
Main  SLIM Web Page.  Explanation and links supporting the SLIM modules
Construction Page.  Hints on constructing the SLIMs.
Much more information on the Tracking Generator addition
Much more information on the VNA extension
Operation and Calibration page for the SLIM MSA. Instructions for calibration and alignment.
Testing the SLIM Modules
Instructions for testing individual modules.
Software page for the SLIM MSA.  Description of the software code.
Link to Builder's Group for those interested in sharing ideas on the MSA.  There are several people in the process of building the MSA and can offer suggestions and comments.  This is a Yahoo Group page and you are welcome to join and contribute.
Coaxial Cavity Bandpass Filter.  A page for construction of the MSA bandpass filter.



Building the MSA, using SLIMs:
    The modules within the following Block Diagram are SLIMs, Standardized Lab Integration Modules.  Each SLIM is a functional module, with specific properties.
Block Diagram of SLIM MSA/TG/VNA System  Click for Express Drawing
slim/bdmsa_tg_vna.gif
    A power supply is not shown, and for clarity, the interconnecting wiring harness is not shown.  All of the inter-module connections shown, are coaxial connections.  SLIM part numbers are shown in the following Table.
    Notice that there is no bandwidth limiting filter preceeding the Mixer 1.  This option is left up to the builder.  Without any filtering, the 0 to 1000 MHz MSA will respond to other input frequencies, from 2000 to 3000 MHz, although rather poorly.
    Also, the Tracking Generator Output has no filtering either.  Again, this is left up to the builder.

SLIMs required for the MSA systems:
    In the following tables, SLIMs are associated with the modules within the MSA Block Diagram.  Each of the SLIMs is linked to a paragraph on this page, that gives special instructions on how to incorporate the SLIM into the SLIM MSA system.
Modules used in the Basic SLIM MSA (without TG or VNA)
Block Diagram Modules
SLIM Name Notes:
Mixer 1 SLIM-MXR-1
DDS 1
SLIM-DDS-107

PLO 1
SLIM-PLO-1 *Partial build
Mixer 2 SLIM-MXR-2
PLO 2
SLIM-PLO-2 *Partial build
I.F. Amplifier SLIM-IFA-33
Final Xtal Filter SLIM-MCF-L024 part number has changed
Log Detector SLIM-LD-8306 *Partial build
Master Oscillator SLIM-MO-64 *Partial build
Control Board
SLIM-CB-NV *Partial build
16 Bit AtoD Converter, or
SLIM-ADC-16
*Partial build
12 Bit AtoD Converter
SLIM-ADC-12 *Partial build, and modified 7-14-08
*partial build, some of the SLIMs can be constructed with minimal parts, to satisfy the minimum requirements of the Basic MSA.  This will reduce the cost a little.

Modules used for the Tracking Generator Addition, omit if not needed
Block Name
SLIM Name Notes:
Mixer 3 SLIM-MXR-3
DDS 3
SLIM-DDS-107

PLO 3
SLIM-PLO-1 *Partial build

Modules used for the VNA Extension, omit if not needed
Block Name
SLIM Name Notes:
Phase Detector
SLIM-PDM

Mixer 4
SLIM-MXR-4


    To complete any form of the MSA, the following additional items are required:
1.  Coaxial Cavity Filter 


2.  external power supply
    The MSA/TG/VNA requires an input of +12 VDC to +15 VDC, with a typical current draw of 750 ma. (600 ma for Basic MSA).  I caution the use of Switching Power Supplies.  Even if the line regulation is supurb, the switching noise radiated into the air may interfere with the MSA operation.  I have always suggested using a linear power supply.  For lab use, a hefty "wall wart" will suffice.  If the builder decides to incorporate an internal, linear power supply into the MSA, I give this warning:  Use a magnetically shielded power transformer.  Keep it as far away from the SLIMs as possible.  Magnetic induction will show up as 60 Hz spurs (and harmonics).

3.  coaxial cables
, to interconnect the SLIMs with RF signals or signals that need shielding.
    I recommend RG-085 semi-rigid (hard pipe).  There are several 50 ohm varieties of this style and size, so I will not be specific with part numbers.  The interconnections are shown in the previous Block Diagram.  Descriptions of the cables are shown in the Coaxial Interconnections Table.

4.  wire harness, to interconnect the SLIMs with power and command signals.  I show three separate wiring diagrams:
    For the Basic MSA, use the Basic MSA wiring diagram
    For the MSA/TG use the MSA/TG wiring diagram
    For the MSA/TG/VNA, use the MSA/TG/VNA wiring diagram

5. 
RF panel mount connectors
    The MSA's RF panel mount connectors are not specified.  They are a builder's preference.  They only need to connect from the front panel to the bottom of the associated SLIMs.  I like the type N, because it is extremely durable. 

6.  Metal enclosure, to house the MSA system
    An ideal configuration, for a completely portable Spectrum Analyzer, would be to build the SLIM MSA into an enclosure that is large enough to house a rechargeable 13.6 volt battery, with enough capacity to supply the MSA and a Laptop Computer at the same time.


Integrating the MSA:
Layout of the MSA/TG/VNA using SLIMs
    Here is a proposed layout of how all of the SLIMs can be physically integrated to complete the MSA, Tracking Generator, and VNA.  The view is from the top, with the interconnecting cables on the bottom side of the boards.  Although this layout is probably accurate, use the Block Diagram as the master document for cable interconnections.  The coaxial cavity filter is shown, mounted vertically, in the back of the assembly.  However, the filter could be mounted horizontally, under or behind the modules, for a thinner assembly.
Layout of SLIM MSA/TG/VNA System  Click for Express Drawing
slim/laymsa_tg_vna_top.gif
  
  Here is a view of the Front, showing a phantom view of the SLIMs.
slim/laymsa_tg_vna_front.gif

    A good method, for integrating the SLIMS, is to use a 6 inch by 8 inch piece of .062 inch pwb with a top layer of copper.  A hole is cut in this board for each SLIM.  The hole size is the same size as the original SLIM pwb (ex. 1.2 x 1.2 in).  A SLIM has a fence around it's perimeters, making it a little wider than the hole.  The SLIM is then placed on the board, over the hole, and tack soldered to the board, in 4 places.  This will make it very easy to remove the SLIM in the future.  It makes each SLIM totally accessable from the top or bottom.  These are the black holes in the blue area in the MSA Layout.
    The coaxial cables can then be constructed and attached to the modules, on the bottom side.  The wiring harness can be connectorized or the wires can be soldered point to point, without connectors.          When completed, the perimeter of the carrier board will have room for drilling holes for mounting to a box or enclosure.  The LPT connector on the Control Board will protrude through the enclosure's front panel.  The power connector will be behind the enclosure's front panel, so a hole must be drilled in the enclosure to allow a power connector to mate with the Control Board's power connector.  A .25 inch hole is drilled to allow the LED to protrude through the enclosure front panel.  Another .25 inch hole is drilled in the enclosure's front panel for mounting the Video Switch.  The last hole(s) depends on the type of RF Input connector the builder prefers.
    RF inputs and outputs of each module can be either SMA (or any small connector) or directly soldered, coaxial cable.  I have been quite successful with direct connections using RG-085 and RG-141 hard pipe, and RG-188 soft coax.  Click here to view a pictoral method of Coaxial Direct Connection to a PCB.  I have gone so far as to remove the outer insulation of soft coax and "sweat" the outer braid with solder.  This makes a 100% outer jacket.  Almost as good as hard pipe!
    All of the SLIMs will accomodate the installation of RF connectors on the bottoms of their boards.  However, none of them have ground vias for mounting ground posts.  Go here to see a pictoral on how to modify an RF connector for installation on a SLIM.  These items are on the Construction Page.

Coaxial Interconnections for the SLIM MSA:
    The following table shows the coaxial interconnections required for the Basic MSA, TG, and VNA.  The coax cables can use the direct connection method, or RF connectors can be used.
Coaxial Interconnection Table for the Basic SLIM MSA
Signal Name
From
To
Recommended
Cable Type
Approximate
Length, Inches
MSA Input
Front Panel
Mixer 1, J2
RG-085
6.0
IF 1 Out
Mixer 1, J3 Cavity Filter
RG-141 2.3
IF 1 In
Cavity Filter Mixer 2, J3 RG-141 2.3
IF 2 Out
Mixer 2, J2 I.F. Amp, J3
RG-085 or RG-188
2.0
IF 2 Amp 1
I.F. Amp, J4 I.F. Amp, J1 RG-085 or RG-188 2.0
IF 2 Amp 2
I.F. Amp, J2 Final Xtal Filter, J2
RG-085 or RG-188 1.6
Final IF
Final Xtal Filter, J1 Log Det, J1
RG-085 or RG-188 1.8
Mag Volts Out
Log Det, J2 AtoD Conv, J1 RG-085 or RG-188 1.8
DDS1OUT
DDS 1, J4
PLO 1, J1
RG-085 or RG-188 2.7
LO1
PLO 1, J3 Mixer 1, J1 RG-085
1.8
LO2
PLO 2, J2
Mixer 2, J1 RG-085 3.5
Master Clock 1
Master Osc, J1
DDS 1, J1 RG-085 2.5
Master Clock 2
Master Osc, J2 PLO 2, J1 RG-085 1.7

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 Master Osc, J3 DDS 3, J1
RG-085
1.8
DDS3OUT DDS 3, J4 PLO 3, J1 RG-085 or RG-188 1.6
LO3
PLO 3, J2
Mixer 3, J1 RG-085 2.5
LO2
PLO 2, J3
Mixer 3, J3 RG-085
5.0
Trk Gen Output
Mixer 3, J2
Front Panel
RG-085 or RG-141 4.0

Additional Coaxial Interconnections for the VNA Extension to the SLIM MSA/TG
Signal Name
From
To
Recommended
Cable Type
Approximate
Length, Inches
LO1 PLO 1, J2 Mixer 4, J1
RG-085
3.5
LO3 PLO 3, J3
Mixer 4, J3 RG-085 1.6
Phase Reference
Mixer 4, J2 Phase Detector, J1 RG-085 or RG-188 1.6
Limited IF
Log Det, J3 Phase Detector, J2 RG-085 or RG-188
1.8
Phase Volts
Phase Detector, J3 AtoD Conv, J2 RG-085 or RG-188 1.8


Signal and Power Interconnections for the SLIM MSA:
    First, I will show a Wiring Diagram for the complete MSA/TG/VNA.  This diagram, along with the MSA Block Diagram and Coaxial Interconnection List, is a complete interconnection scheme for the SLIM MSA/TG/VNA.  Not shown are the connections to the external power supply or LPT computer port.
WDMSA-TG-VNA, Wiring Diagram, SLIM MSA/TG/VNA.  Click for Express Drawing slim/wdmsa_tg_vna.gif
    The "option" shown between P9 and the SLIM PDM means that the SLIM ADC Module can be supplied with either 10 volts from P9, or 5 volts from SLIM PDM, P2 (modification to the ADC Module is necessary for the 5 volt option).

Next, is a Wiring Diagram for the MSA with Tracking Generator, MSA/TG.
WDMSA-TG, Wiring Diagram, SLIM MSA/TG.  Click for Express Drawing
slim/wdmsa_tg.gif

    Last, is a Wiring Diagram for the Basic MSA.
WDMSA, Wiring Diagram, Basic SLIM MSA.  Click for Express Drawing
slim/wdmsa.gif
Updated 6-03-08. 
You will notice that Connector P1-2 is supplying a common clock to (up to) five modules.  It is important that independent wires be connected to P1-2, one wire to each module.  Do not "daisy chain" this signal, even though it looks like a daisy chain in the drawing.  If daisy chained, the reflected signal may cause multiple clocking events, especially in the DDS modules.  This is a very fast rise time and fall time signal.  The data signals can be daisy chained, since they are not edge triggered.

Special Instructions and Considerations for SLIMs in the MSA:
    There are three configurations of the MSA.  The Basic MSA, the MSA with Tracking Generator, and the MSA with Tracking Generator and VNA.  The following paragraphs will discuss any special treatments a SLIM must have, to be used in one of the three MSA configurations.  For full descriptions and construction of each SLIM, click on the SLIM part number in the paragraph's header.  It will link to the Web page containing the full documentation for that SLIM, including pictures.  Use your Browser's "BACK" button to return to this page.  The SLIM's Web pages will not have these special instructions, since they can be used as modules for other construction projects that don't need special considerations that the MSA does.
    When preparing the following modules for integration into your version of an MSA, or MSA with Tracking Generator, or MSA/TG with Vector Network Analyzer, I HIGHLY SUGGEST printing out each group of documentation.  Then, use a red pencil to update or change the documentation that the following paragraphs require.  You will have a complete, "baseline" history of your system, in case there are any future design changes.  Also, if you ever have a question, as to performance, or are troubleshooting a problem, you can scan your document and send it to me for help.  A picture is worth a gazillion words.

Control Board using  SLIM-CB-NV rev B
    The SLIM-CB-NV is configured with a Noise Filter Section and a Voltage Converter Section.  The noise filter is not required for any MSA configuration.  All the components in that section can be omitted.
    In the Voltage Converter section, only the +20 volts is used in the MSA.  The -10v is not required in the MSA.  Therefore, C18 and C19 can be omitted, although leaving them will not increase power consumption or add extra noise.


Mixer 1 using  SLIM-MXR-1 rev A
    The SLIM-MXR-1 is configured with the Minicircuits, ADE-11X.  The observant may notice that my schematic conflicts with what Minicircuits calls the I port.  Pin 2, of the ADE-11X package, is internally connected to the diode bridge.  This is a "normal" I port configuration, and this is how I use it in the MSA.  This is because the diode network port has a much better low frequency response than the transformer ports.  MSA inputs down to a few KHz can be measured quite accurately.
    Special Caution: Since the MSA input is directly coupled to Mixer 1, the I port of the mixer can be destroyed by applying a signal with a DC voltage.  It can also be destroyed by a high level input signal that is AC coupled.  Good rule of thumb to prevent mixer damage: maximum input signal should not be greater than the specified LO port power (+7 dBm), and certainly, no DC voltage is allowed on the I port, J2.
   
Note here, that SLIM-MXR-1 has been revised to Rev A.  This revision adds an internal 2.5 dB attenuator inside the Mixer module.  This localized attenuation improves the conversion effeciency and port to port isolation of the ADE-11X.  The LO power level applied to Mixer 1 is +10 dBm (from PLO 1) and the attenuator drops it to 7.5 dBm, the level for the ADE-11X's L port.

Mixer 2 using  SLIM-MXR-2 rev B
    MXR-2 uses the same pwb as MXR-1.  But, MXR-2 uses a filter in the output I port.  In the MSA, the first I.F. of 1013 MHz is mixed with 1024 MHz from LO 2.  The wanted output frequency at port I is 10.7 MHz.  However, there will be other mixing components,  R+L  = 2037 Mhz.  Some signal at the L port will feed through to the I port, and so will some signal from the R port.  The filter is a diplexer, with it's crossover at 33 MHz.  It will allow the 10.7 MHz I.F. will pass to J2, while preventing the high frequency components from leaving J2 and getting to the MSA's I.F. Amplifier.
    Note here, that SLIM-MXR-2 has been revised to Rev B.  This revision adds an internal 2.5 dB attenuator inside the Mixer module.  This localized attenuation improves the conversion effeciency
and port to port isolation of the ADE-11X.  The LO power level applied to Mixer 2 is +10 dBm (from PLO 2) and the attenuator drops it to 7.5 dBm, the level for the ADE-11X's L port.

Mixer 3 using  SLIM-MXR-3 rev A
    MXR-3 is used only when the Tracking Generator is added to the Basic MSA.  MXR-3 output is the difference frequency of the inputs, LO 3 and LO 2.  The output level is approximately -10 dBm.
   
Note here, that SLIM-MXR-3 has been revised to Rev A.  This revision adds an internal 2.5 dB attenuator in the L port path and a 14 dB attenuator in the R port path.  These localized attenuators improve the conversion effeciency and port to port isolation of the ADE-11X.  The LO power level applied to Mixer 3 is +10 dBm (from PLO 3) and the attenuator drops it to 7.5 dBm, the level for the ADE-11X's L port.  The R power level applied to Mixer 3 is +10 dBm (from PLO 2) and the attenuator drops it to -4 dBm, the level for the ADE-11X's R port.
    There is one modification (Rev A) that is on the schematic and parts list, but needs mentioning here.  The R input to Mixer 2 is operating at a fixed frequency of 1024 MHz.  The ADE-11X R port impedance is not exactly, 50 ohms.  Matching the mixer's R port, to the internal 14 dB attenuator, can be greatly improved by adding a 2.0 or 2.2 pF chip capacitor in the C9 position.  This is on the ADE-11X, pin 3 to ground.


Mixer 4 using  SLIM-MXR-4 rev A
    MXR-4 is used when the MSA/TG is extended into a VNA.  MXR-4 output is 10.7 MHz when the Tracking Generator is activated.  The output level is approximately -10 dBm.
    Note here, that SLIM-MXR-4 has been revised to Rev A.  This revision adds an internal 2.5 dB attenuator in the L port path and a 14 dB attenuator in the R port path.  These localized attenuators improve the conversion effeciency and port to port isolation of the ADE-11X.  The LO power level applied to Mixer 3 is +10 dBm (from PLO 1) and the attenuator drops it to 7.5 dBm, the level for the ADE-11X's L port.  The R power level applied to Mixer 3 is +10 dBm (from PLO 3) and the attenuator drops it to -4 dBm, the level for the ADE-11X's R port.

1013.3 MHz Coaxial Cavity Filter : Cavity Filter Construction Page
    When constructing this filter, leave enough semi-rigid cable attached to the filter to connect to Mixer 1 and Mixer 2 of the MSA.  I might even suggest that the builder construct this filter last, to insure a proper "fit" into the MSA.


DDS 1, using SLIM-DDS-107 rev C
    The MSA configurations require only one output from DDS 1, the square wave output from the Squaring Buffer, J4.  It is used as the reference clock input to J1 of PLO 1.  However, I would suggest that J3 (DDS B) be populated with a pwb RF connector, or a coaxial cable running to a connector on the front panel of the MSA assembly.  It can be used as a frequency source for other purposes.  The DDS B output will have an amplitude of approximately -8 dBm.  The signal will also contain alias frequencies and many spurious frequencies.  I will publish some experiments, using this output, at a later date.
    Omit or remove R3, a 49.9 ohm resistor.  This will allow a full 5 volt pp Master Clock, and the reflection will be retrurned to the Master Clock Module and be absorbed by the 50 ohm series driving resistor.

DDS 3, using SLIM-DDS-107 rev C
    The DDS 3 is required only when adding the Tracking Generator or VNA to the Basic MSA.  It is built and configured identically to the DDS 1.  As an option, it's J3 can also be brought out to the front panel for other experimentation.

PLO 1 using SLIM-PLO-1 rev B
    A fully configured SLIM-PLO-1 has two active outputs, J2 and J3. The Basic MSA requires only one output from the PLO 1, used as the LO drive for Mixer 1, at +10 dBm.  In this Basic MSA physical layout, the B output (J3) is preferred, but not manditory.  J3 is closer to the user, Mixer 1.  If you know you will never use the J2 output, you can perform one of the following modifications.  Either modification can be reversed, at a later date.
   Mod A.  If the SLIM-PLO-1 is fully configured with components, it can be modified to conserver power, by disconnecting the A buffer.  This decreases consumption by 38 ma.  Remove R13.  This removes power to the amplifier.  Remove C36, the input coupling capacitor to the A buffer amp.  Install a 49.9 ohm resistor, size 0803 surface mount, directly on top of R24 (piggy back).  This allows the VCO output divider circuit to see a correct 50 ohm load.  If you do this mod, I suggest taking the removed R13 and C36 components and tack solder them to a ground spot.  If this modification is reversed, at a future date, the parts are there for re-installation.
   
Mod B.  Modify the PLO to have a single output at J3.  When constructing the SLIM-PLO-1, do not populate the A side buffer.  Do not install the following components:  R13-R17, C31-C37, L1, and U7.  Install a 49.9 ohm resistor, size 0803 surface mount, directly on top of R24 (piggy back).
   If you plan to expand the MSA into the VNA configuration, either now, or at a later date, use the SLIM-PLO-1 with both buffer amplifiers populated.

    When constructing SLIM-PLO-1 for the SLIM MSA system, omit C27 and R12.  The driver to this circuit will absorb the mismatch reflection.  This may seem a like poor design but it is valid.  It improves the noise performance of the PLL.

PLO 3 using SLIM-PLO-1 rev B
    PLO-3 is used when the Tracking Generator is added to the Basic MSA.  For this purpose, only one buffer amplifier is required to drive Mixer 3.  However, I would suggest fully constructing PLO-3 with both buffer amplifiers.  The spare buffer amplifier output can be brought out to the front panel and used for future experiments, since it's output is 1000 MHz to 2000 MHz., at a high power level.  Load it with 50 ohms if not used.
    If the MSA/TG is extended into the VNA configuration, both buffer amplifiers are used.

PLO 2 using SLIM-PLO-2 rev B
    A fully configured SLIM-PLO-2 has two active outputs, J2 and J3. The Basic MSA requires only one output from the PLO 2, used as the LO drive for Mixer 2, at +7 dBm.  In this Basic MSA physical layout, the A output (J2) is used.   If you plan to add the Tracking Generator to the MSA, either now, or at a later date, use the SLIM-PLO-2, fully populated with both buffer amplifiers.  If you know you will never use the J2 output, you can perform one of the following modifications.  Either modification can be reversed, at a later date.
   Mod A.  If the SLIM-PLO-2 is fully configured with components, it can be modified to conserver power, by disconnecting the B buffer.  This decreases consumption by 38 ma.  Remove R18.  This removes power to the amplifier.  Remove C43, the input coupling capacitor to the B buffer amp.  Install a 49.9 ohm resistor, size 0803 surface mount
, directly on top of R25 (piggy back).  This allows the VCO output divider circuit to see a correct 50 ohm load.  If you do this mod, I suggest taking the removed R18 and C43 components and tack solder them to a ground spot.  If this modification is reversed at a future date, the parts are there for re-installation.
    Mod
B.  Modify the PLO to have a single output at J2.  When constructing the SLIM-PLO-2, do not populate the B side buffer.  Do not install the following components:  R18-R22, C38-C44, L21, and U8.  Install a 49.9 ohm resistor, size 0803 surface mount, directly on top of R25 (piggy back).

    When constructing SLIM-PLO-2 for the SLIM MSA system, omit C27 and R12.  The driver to this circuit will absorb the mismatch reflection.  This may seem a like poor design but it is valid.  It improves the noise performance of the PLL.

Final I.F. Xtal Filter using SLIM-MCF-L024 (place holder)
    The Final Xtal Filter determines the Resolution Bandwidth of the MSA.  Steep slopes and out of band rejection is a must for good selectivity.  The SLIM-MCF-106 is designed and placed here, giving the MSA a resolution BW of 2.2 KHz.  The center frequency, and Final I.F., is 10.695 MHz.  It is an 8 pole filter and has a loss of about 4 dB.  The monolithic crystal filter is from U.S. Electronics, part no. 10L024A.  I designed this module, "blind".  That is, I don't have this filter to test.  However, I have tested several similar filters, taken from some single sideband CB radios, with similar characteristics.  This is the only filter I could find that is on the open market, for sale.  All the filters I have experimented with, are not offered to the public.  The problem is, U.S. Electronics sells in lots of 1000.  I (we) must find a viable alternative to this filter.  That is why the term "place holder" is in the title.
    I will soon add a final filter using cascaded filters that are available from Digikey and Mouser.
Final I.F. Xtal Filter using SLIM-MCF-FL096 (place holder)
    This is the design I used when building and testing the first SLIM MSA/TG/VNA verification unit.  It gives the MSA a resolution BW of 3.8 KHz.  The center frequency, and Final I.F., is 10.695 MHz.  It is an 8 pole filter and has a loss of about 4 dB.  The filter is from an old CB radio and is probably made by Uniden.  However, I can not find it for sale anywhere.  That is why this is a place holder.

Log Detector using SLIM-LD-8306
    The Log Detector is the mechanism for converting RF power to a dc voltage.  The SLIM-LD-8306 has a broadband input transformer.  Be aware that this log detector is very responsive to wide band frequencies, and noise. The Log Det Module must be preceeded by a noise bandwidth filter.  In the MSA, this function is performed by the Final Xtal Filter.
   
The SLIM-LD-8306 has two outputs.  The Mag(nitude) Volts Output at J2 is a DC voltage, where it's level is relative to the amount of input power to the module.  The output is approximately .3 volts to 2.3 volts for a power input (Log Det input) of -90 dBm to +10 dBm.  The converion factor is 20 millivolts per dB.  This output is passed to the Analog to Digital Converter for use by the computer.
    The second output, Lim(ited) IF Out on J3, is not used in the Basic MSA.  It is a square wave of the input at J1.  It is used when the MSA is expanded to the MSA/VNA configuration.  If you plan not to use this output, do not install R4-R6, and C12.  This will disable the limiter section of the I.C.


A to D Converter, using either a 12 or 16 Bit Serial A to D Converter:
16 Bit Serial A to D Converter using SLIM-ADC-16
Changed, 7-14-08
    The SLIM-ADC-16 design is a similar to the 16 Bit design in the original MSA.  This module has no adjustment potentiometer for calibration.  With 16 bit resolution, adjustment is not necessary for the MSA.  This module has two A/D circuits, but only one is used in the Basic MSA.  The other is used when the MSA is expanded to VNA operation.  Therefore, if the builder would like, he can delete the "PHA VOLTS" section.  That would be U3 and all of it's supporting components.
    In the MSA, the "MAG VOLTS" input is connected back to the Log Detector's Magnitude output.
  J1 will accept an input range of 0 volts to +5 volts, but the Log Det. output voltage is expected to range only from +0.4 volts to +2.4 volts (it's maximum 100 dB range).  This 16 Bit AtoD will convert +0.4 volts to a bit value of "5243".  The bit value of +2.4 volts is "31457".  The dynamic bit range is 31457 - 5243 = 26214 bits.  Therefore, the conversion factor for the MSA's combination of Log Det and 16 Bit AtoD Converter is: 100 dB/26214 bits = .0038 dB per bit.  However, the 2 least significant bits are specified to be indeterminite, ie, 14 bit resolution.  With 14 Bits, the conversion factor is .015 dB per bit.  My testing only showed a single bit error, ie, 15 bit resolution.
    The "PHA VOLTS" (J2)
is configured for an input dynamic range of 0 volts to +5 volts, and that is the range expected from the Phase Detector Module.  A delta 5 volt range equates to 360 degrees.  Therefore, the resolution of the SLIM-ADC-16 is 360/65536 = .0055 degrees per bit.  However, 14 bits is possible, which still gives a resolution of 360/16384 = .022 degrees per bit.  Here again, my testing showed only a single bit of undetermined resolution, ie, 15 bits of resolution = .011 degrees.  In reality, the MSA system noise is much higher than this resolution.
    If this module is part of the MSA/VNA expansion, there is one (optional) modification that can increase Phase Measurement accuracy.  It is a simple mod:  Remove (or do not install) U1, the 5 volt regulator.  Add a jumper wire between the remaining pads associated with U1 pin 3 and U1 pin 1.
    When interconnecting the modules in the system, connect SLIM-ADC-16, P1, pins 1 and 2 to SLIM-PDM (SLIM Phase Detector), P2, pins 1 and 2.  This allows both modules to use the same 5 volt source as a common reference.  If this modification is performed after an MSA Magnitude calibration, the calibration must be retaken, due to a probable change in reference voltage.

12 Bit Serial A to D Converter using SLIM-ADC-12
Changed, 7-14-08.  For better accuracy, at the expense of resolution.
    The SLIM-ADC-12 design is an option to the 16 Bit SLIM.  It is less expensive to build, yet still has good resolution.  This module has no adjustment potentiometer for calibration.  With 12 bit resolution, adjustment is not necessary for the MSA or VNA.  This module has two A/D circuits, but only one is used in the Basic MSA.  The other is used when the MSA is expanded to VNA operation.  For a Basic MSA, the "PHA VOLTS" section can be unpopulated.  That would be U3 and all of it's supporting components.
    The J1 (MAG VOLTS) section is configured for an input dynamic range of 0 volts to +2.8 volts.
  In the MSA, the "MAG VOLTS" input is connected back to the Log Detector's Magnitude output, which is expected to range from +0.4 volts to +2.4 volts, (the maximum 100 dB range of the Log Det).  This 12 Bit AtoD will convert +0.4 volts to a bit value of "585".  The bit value of +2.4 volts is "3511".  The dynamic bit range is 3511 - 585 = 2926 bits.  Therefore, the conversion factor for the MSA's combination of Log Det and 12 Bit AtoD Converter is: 100 dB/2926 bits = .034 dB per bit.  However, the A/D chip is specified to have a 1/2 bit resolution.  With 11.5 Bits, the conversion factor is closer to  .0483 dB per bit.  I have not tested this resolution.

Note: the following paragraph has a major change, 7-14-08
    The "PHA VOLTS" (J2) was originally configured for an input dynamic range of  +1 volt to +4 volts, but I am changing the range to 0 volts to +5 volts, the expected input from the Phase Detector Module.  The reason for this change is that the voltage divider in the SLIM-ADC-12 can create as much as a 2% error in phase accuracy.  This equates to an error of 7.2 degrees.
    Therefore, a modification to the
SLIM-ADC-12 is required.  If the SLIM-ADC-12 has been fully assembled, just remove R6, a 665 ohm resistor.  If you have not assembled the SLIM-ADC-12, change R5 and R7 to 0 (zero) ohm resistors.  R6,C15, C16, C17, and C18 are not installed. 
    Now, the resolution of the SLIM-ADC-12 is 360/4095 = .0879 degrees per bit.  However, 11.5 bits is possible, which still gives a resolution of 360/2896.3 = .124 degrees.  I have not tested this resolution.
    If this module is part of the MSA/VNA expansion, there is one (optional) modification that can increase Phase Measurement accuracy.  It is a simple mod:  Remove (or do not install) U1, the 5 volt regulator.  Add a jumper wire between the remaining pads associated with U1 pin 3 and U1 pin 1.
    When interconnecting the modules in the system, connect SLIM-ADC-12, P1, pins 1 and 2 to SLIM-PDM (SLIM Phase Detector), P2, pins 1 and 2.  This allows both modules to use the same 5 volt source as a common reference.  If this modification is performed after an MSA Magnitude calibration, the calibration must be retaken, due to a probable change in reference voltage.

Master Oscillator using SLIM-MO-64 rev B
    The SLIM-MO-64 has three outputs.  Only two are used in the Basic MSA.  If an output is not used, leave it unterminated.  Or, to conserve power in the Basic MSA, U6, R3, R6, and C9 could be deleted.

I.F. Amplifier using SLIM-IFA-33 rev A
    The SLIM-IFA-33 has two independent amplifiers with an operating bandwidth of 3 to30 MHz.  The output of one (J4) is connected to the input of the other (J1) with a short piece of coaxial cable.  This gives a total gain of 40 dB.  Saturated output can be as much as +14 dBm.

PDM, Phase Detector Module, using SLIM-PDM rev A
    The SLIM-PDM is used only when expanding the MSA/TG into the VNA.  This module operates at 10.7 MHz and has squaring circuits within it.  Consequently, it is a potential radiator of harmonic noise.  It is extremely important that this module be well shielded.  When the VNA is not running, this module is still actively amplifying noise.  "Funnies" in the MSA spectrum could be attributed to an "unclean" PDM.
    The internal +5 volt reference is tied to connector P2.  To increase phase measurement accuracy in the MSA/VNA system, this voltage source can be optionally used to supply the Analog to Digital Converter Module.  Either the SLIM-ADC-12 or SLIM-ADC-16.  Do not use it for any other purpose.

Specifications for the Basic MSA using SLIMs
    Dual Conversion                      1013.3 MHz first I.F., 10.7 MHz,  final I.F.
    Frequency Response               2 KHz to 1038 MHz (with 2 KHz resolution bandwidth filter)
                                                      
15 KHz to 1038 MHz (with 15 KHz resolution bandwidth filter), etc.
    Frequency Step Size               1.4 Hz at 0 Mhz, 2.8 Hz at 1000 MHz
    Input Power Sensitivity            greater than -20 dBm, to -105 dBm
    Dynamic Range                        > 85 dB
    Amplitude Resolution              0.1 dB
    Selectivity (BW)                     Depends on builder's choice of Resolution Filter (Final Filter),
                                                     I use 4 different Final I.F. BW filters: 9.5 MHz/200 Hz, 11.15 Mz/2 KHz,

                                                     10.695 MHz/4 KHz, 10.7 MHz/15 KHz
    Noise Figure                           < 23 dB
    Phase Noise                            better than -91 dBc/Hz  (1 Hz BW), 3 KHz away from carrier
                                                    -90 dBc/Hz  10 KHz away from carrier
                                                    -87 dBc/Hz  40 KHz away from carrier (noise peaking)
                                                    -101 dBc/Hz  100 KHz away from carrier
                                                    -114 dBc/Hz  300 KHz away from carrier
                                                    better than -116 dBc/Hz  500 KHz away from carrier or above
    Image Rejection                     In-Band Image Rejection is better than -112 dBc,
   IM Distortion                   Two tone, better than -60 dBc, worse case.  Avg better than -70 dBc

Cost of the Basic MSA using SLIMs
Cost of SLIMs ($USA), fully built with connectors, without connectors, and with minimum parts required for the Basic MSA:
SLIMs Full Boards wo Conns w Rqd Pts
SLIM-ADC-12 33.44 23.96 14.23
SLIM-CB-NV 39.3 29.82 25.86
SLIM-DDS-107 52.33 33.37 33.37
SLIM-IFA-33 30.52 11.56 11.56
SLIM-LD-8306 45.16 30.94 30.69
SLIM-MCF-L024 25.84 16.36 16.36
SLIM-MO-64 25.15 10.93 10.47
SLIM-MXR-1 20.63 6.41 6.41
SLIM-MXR-2 21.75 7.53 7.53
SLIM-PLO-1 68.54 54.32 51.89
SLIM-PLO-2 49.52 35.3 32.87
Total for SLIMs Only 412.18 260.5 241.24

Total cost of Basic MSA ($USA), with other materials:
Other Materials
Conns
wo Conns
w Rqd Pts
28 in. RG-085@$5 ft.
11.67 11.67 11.67
23 Conns@$4.74
109.02 0 0
Cavity Filter
9.48 4 4
Wall Wart Pwr Supply
15 15 15
Front Panel Conn.
6 6 6
Enclosure Box
20 20 20
Total for Basic MSA
583.35 317.17 297.91
    These costs can vary quite a bit.  I used the minimum order requirements of Digikey.  Minimum order requirements from ExpressPCB and Minicircuits would require sharing among builders.  The cost can be lowered by using "junk box" parts.  As you can see, the cost of the Basic MSA using the Direct Coax Method (wo Conns) is substancially lower.  I have not included the cost of solder, fencing material, interconnecting wires, and certainly not the value of labor.  I can be more accurate in the future, but I calculate a total of 104 hours of construction time.

Added cost for Tracking Generator Addition:
SLIMs Full Boards wo Conns
SLIM-MXR-3 21.75 7.53
SLIM-PLO-3 68.54 54.32
SLIM-DDS-107 52.33 33.37

Added cost for VNA Extension:
SLIMs Full Boards wo Conns
SLIM-MXR-4 21.75 7.53
SLIM-PDM 23.66
9.44


Signal Flow in the Basic Spectrum Analyzer
    The signal (to be measured) is input to the I port of the first mixer (Mxr 1).  It is high side mixed with the first Local Oscillator (LO 1) to create the first intermediate frequency of 1013.3 MHz (1st IF=LO-RFin).  This output, from the first mixer R port, is passed to the coaxial cavity filter.  The main purpose of this filter is to attenuate the I.F. Image Frequency that will occur at 1034.7 MHz.  The 1st IF is passed to Mixer 2, where is is mixed with a fixed frequency of 1024 MHz., LO 2.  The mixing result is the final I.F. frequency of 10.7 MHz, and other higher frequency components.  The low pass filter passes only the final I.F. for amplification, and then filtering by the Final Xtal Filter.  The Final Xtal Filter determines the Resolution Bandwidth of the SA.
    The final IF is sent to the amplitude measuring device, the Logrithmic Detector (Log Det).  The output of the Log Detector (video) is a DC voltage corresponding to the logrithmic power of the incoming signal.  This video voltage is converted to a Data word by the action of the Analog to Digital Converter (A/D).  The computer program converts the Data word to a power level and is graphed on the computer's monitor in a window called Graph Window.
    The Local Oscillator (LO1) for Mixer 1 will tune from 1013.3 MHz to 2013.3 MHz in as little as 1.4 Hz steps.  This is accomplished by using a Hybrid Synthesizer, consisting of PLL 1, steered by DDS 1.
    The Local Oscillator (LO2) for Mixer 2 is a fixed frequency of 1024 MHz.

Signal Flow for the Tracking Generator (to be added)
Signal Flow VNA, Vector Network Analyzer (to be added)

MSA Analysis
    This is a specific analysis of the MSA using the recommended SLIMs.
    First, approximate gains and losses in the SLIM MSA. 
Each mixer has -6.5 dB of loss.  The coaxial cavity filter has -7 dB of loss.  The two I.F. Amplifiers have a total of +40 dB of gain. The Final Xtal Filter is optional, but I will show a filter that has -4 dB loss and a bandwidth of 2.2 KHz.  The total MSA gain = -6.5dB -7dB -6.5dB +40dB -4dB = +16 dB gain.
    The Log Detector SLIM has a signal input range of -90 dBm to +10 dBm.  -90 dBm is the amount of signal at the input to the Log Det Module that equals the self generated noise of the Log Detector I.C.   +10 dBm is the amount signal level at the input to the Log Detector Module to saturate the Log Detector I.CTherefore, the instantaneous dynamic range of the Log Detector Module is 100 dB.  Consequently, the instantaneous dynamic range of the MSA can not be greater than that of the Log Detector Module.
    The maximum signal input level to the MSA is the level of input power that causes the Log Detector to saturate.  The minimum signal input to the MSA (sensitivity) is determined by the Noise Floor of the MSA.  The difference between these two power levels is the Instantaneous Dynamic Range of the MSA.
    The maximum signal at the input to the Log Detector module is +10dBm.  The signal level at the input to the MSA, to be equal to the maximum power at the input of the Log Detector Module, is calculated by:  PWRin = +10dBm -total gain(+16 dB) = -6 dBm.  This is the maximum signal input level of the MSA.
    The Noise Floor of the MSA is determined by the self generated noise of the circuits within the MSA.  The noise created in the MSA is the combination of the two I.F. Amplifiers and the Log Detector.  The first I.F. Amplifier has a noise figure of 3 dB and a gain of 20 dB.  The noise generated by the first amplifier is = -174dBm +3dB(amp noise figure) +20dB(gain) =  -151 dBm /sqrtHz.  The second amplifier increases the noise to -131 dBm /sqrtHz.  The noise is limited by the 2.2 KHz bandwidth of the Final Xtal Filter, and its loss.  Total noise at the input to the Log Det SLIM = -131 dBm /sqrtHz + 10logBW(2.2KHz) - 4dB(filter loss) =  -101.6 dBm.  This is the total noise at the input to the Log Det Module, which is internally generated by the MSA.
    This noise level is 11.6 dB less than the -90 dBm self generated noise of the Log Detector.  With this filter, the Noise Floor of the MSA is the actual noise of the Log Detector, and not the circuitry preceding the Log Detector. 
Since the equivalent noise level at the input to the Log Detector is -90 dBm, the level of input power to the MSA that would create a signal level of -90 dBm at the input to the Log Detector Module is:  PWRin = -90 dBm - total gain(+16 dB) = -106 dBm.  This is the noise floor of the MSA (with this filter resolution).  It is also the minimum detectable signal input of the MSA (sensitivity).  Therefore, the Instantaneous Dynamic Range of the MSA (with this filter) is = -6 dBm - (-106 dBm) = 100 dB.
     If the Final Xtal Filter is replaced with a wider bandwidth filter, the internally generated noise level will increase.  Replacing the Final Xtal Filter with one having a bandwidth of 15 KHz, the total noise at the input to the Log Det will be:  Total noise = -174dBm +3dB(amp noise figure) +20dB +20dB +10logBW(15KHz) -4dB(filt loss) =  -89.2 dBm.  This noise level is .8 dB greater than the -90 dBm self generated noise of the Log Detector.  Therefore, the circuitry in front of the Log Detector Module determines the noise floor.  The signal level at the input to the MSA to equal the noise power at the input to the Log Detector of -89.2 dBm is now: PWRin = -89.2dBm - total gain(+16 dB) = -105.2 dBm.  This is the minimum input level to the MSA, with this wider filter.  Therefore, the Instantaneous Dynamic Range of the MSA (with the 15 KHz filter) is = -6 dBm - (-105.2 dBm) = 99.2 dB.
    A dynamic range of 99 dB is extremely good.  However, there is a potential problem that I have not previously addressed.  And that is, the Final Xtal Filter may not be able to accomodate a large signal without internal destruction or significant distortion.  With a -4 dB loss, that means a signal of +14 dBm is on the filter's input when the maximum input of -6 dBm is applied to the MSA.  This level of +14 dBm is also the 1 dB compression point of the amplifier used in the RF Amplifier module.  So, not only can we destroy the crystal filter, the signal will be distorted. Therefore, it is a good idea to operate the Final Xtal Filter with no more than 0 dBm on it's input.  This is 14 dB below our calculations for maximum input, so we derate the maximum power input by 14 dB.   Therefore, the maximum input to the MSA, instead of -6 dBm, is -20 dBm.  This lowers the MSA dynamic range by 14 dB, from 99 dB to 85 dB.  Still, a very good dynamic range.  Personally, I think that an instantaneous dynamic range of 70 dB or greater is a good system.  The total dynamic range of the MSA can be increased by using a selectable attenuator at the input of the MSA.  This is a user's preference, and I have restricted all calculations to the internal aspects of the MSA.  One further note; -6dBm input to the first mixer is acceptable, but does have the potential to create higher IMD products.  Lowering the maximum signal input to the mixer, to -20 dBm, has the advantage of lowering IMD products.

Addendum:
    1. Update 7-30-08.    I had previously stated, that, it was not a good idea to use a fan inside the MSA/VNA for cooling.  The reason is vibration.  There are several components in the MSA that are sensitive to vibration.  Probably the most critical, is the Final Crystal Filter, due to it's piezo characteristics.
    I am going to retract my concerns a bit.  I have tested the Verification MSA, having installed a muffin fan on the bottom cover.  It is pointed directly up, at the bottom of the SLIM Log Detector module.  The Master Oscillator is adjacent to it and runs quite warm.  The muffin fan is a 2.5 inch, 24 volt, rated at 90 ma.  I am running it off the +13.6 volt input line, and it is drawing 40 ma.  I cut a 2.5 inch hole in the top cover for the air exit.  With it "half blowing" it is keeping the MSA, cool as a cucumber.
    I have tested for vibration effects in MSA Mode and VNA Mode and see absolutely no ill effects.  Therefore, a really "quiet" fan, is acceptable.

    A little about me:  My name is Scotty Sprowls.  I am a retired RF Design Engineer from E-Systems / Raytheon.  Although I am not an Amateur Radio Operator, I do repair radios as a hobby.  My frustration in tuning cavity filters in diplexers for a couple of Hams caused me create this Spectrum Analyzer to aid me.  You can get in touch with me, via email at,   wsprowls(at)yahoo.com

  I will try to answer your questions or comments as soon as possible.
(end of page)