Project blog


Primcalc Upgrade

posted Apr 8, 2012, 9:42 PM by Gregory Kiesel

Primcalc has gotten a major upgrade- you can now display model information in PrimCalc, to explain what the model is.  It's really just peeking at a JPEG, if present, next to the XML files for the models.  Currently working for the microstrip, stripline, and lumped element models.  I'll get the coplanar libraries added in the next couple of days. 

The next major upgrade I'm looking to add is a GUI to allow for manual tuning in Mason.  You'd get various optimization parameters as a tuning bar to tweak values and update the plots in real time.

GUI Updates

posted Sep 22, 2011, 7:06 PM by Gregory Kiesel

It has been a while since I posted comments... major update for GUI.

In TinyCAD, in the optimizer or statistical analysis block, if you set the "optimizer" or "argument" to not be shown ("Show?"), then that will disable that optimizer or goal.

If you set the Reference for ports or devices (Like _Port1, or TL1) to be not shown, that will remove the device from the circuit.  This will not affect Flags, Variable blocks (including microstrip and stripline properties), calculation blocks, or Frequency blocks.

If you set the argument for a variable block (including microstrip or stripline properties) or a calculation block to not be shown, that will remove that variable or calculation.

If you set the argument in the Flag block to be not shown, that flag setting will not be used.

The big exception is that devices (like TL1) can have arguments that are not shown, but are still used.

Also, I recently made the 30% - 50% faster. 

Circuit Mason works with Wine

posted Apr 19, 2011, 9:53 PM by Gregory Kiesel   [ updated May 10, 2011, 8:13 AM ]

Battled through naming conventions, incomplete thread APIs, and missing DLLs, and have gotten the GUI tools to work with Wine.  Now on a stable release, so far nothing seems broken.


Improved Schematic Simulation and Layout

posted Oct 8, 2010, 6:43 AM by Gregory Kiesel

I've improved the way Mason handles the simulation and layout from the schematic with version 3.4.1.  I think I've fixed the layout problems I was initially seeing.  I've also added two new libraries, Murata with some caps and Hittite for some amplifiers.  I don't have a lot in there yet, but I'll try to add more parts soon.  

I will also add some examples of how to use these features, probably under the "Power Users" section (use the search site feature to find them).

I've added a SOT26 "smart footprint" component to support a new Hittite library, you can find it in the "lumped elements" folder.


New Schematic Feature

posted Oct 5, 2010, 5:22 PM by Gregory Kiesel   [ updated Oct 7, 2010, 5:22 AM ]

I've upgraded the way Mason simulates and performs layouts; check out the description and example on the Power Users page.  You can now map the ports of schematic item to the ports of the S-Parameters... so if you have an LNA in a SOT23-5 package, you can use that 5-pin part (with the non-RF power and ground pins) for the layout and the simulation.  The LNA just needs to have the Nodes defined as to which pins of the device map to the S-Parameter terminals.  This works well for the schematic and the simulation... but the layout is actually not quite right yet.  I have to work on mapping the power planes better, and mapping the footprints better.  Layout works okay for just microstrip and simple parts, just seem to have problems relating to mapping parts where the simulation ports is less than the number of pins on the schematic.

I've also clarified on that page how grounds can be used in TinyCad.  That's not technically a new feature, just newly documented.  I need to work on how I do RF grounds for power nets.

Also, should have some more parts libraries for download before the week is up... working on some Murata and Hittite parts.

Coupled lines, documentation

posted Sep 24, 2010, 3:07 AM by Gregory Kiesel

The model for the edge-coupled microstrip line is now complete, and in related news the 2D Finite Difference code for generic edge coupled lines is also completed.  I'd treat the 2D Finite Difference code as beta for now, it seems to check out but has not been rigorously tested.  Both of these are described in the microstrip section of the documentation.

I've working on my first draft of my Mason workbook.  The workbook is written as a series of labs geared for supplementing a high school or physics level physics or engineering class.  It would also be appropriate for those who have a basic understanding of RF and want the step-by-step tutorials for using Mason. 

Coupled Lines

posted Aug 28, 2010, 11:20 AM by Gregory Kiesel

I'm currently attacking the coupled lines problem on two fronts.  Firstly, we're working on a microstrip coupled model, using the design equations (like every other microstrip model).  Second, I'm working on a 2D Method of Relaxation code to add to Mason.  This will be slower than the design equations, but initial tests show it isn't too bad. 

The 2D MOR code will be accessible through the math parser, specifically through several new functions.  Devices taking advantage of the new feature will be added using the same file format as the existing dynamic (primitive) components.  The changes should be available in the next couple weeks, should be considered "beta".


Variable Validation

posted Aug 12, 2010, 11:25 PM by Gregory Kiesel

Added a warning if a variable is used before being initialized.  Useful for catching mispelled (sic) variables- remember that Capitalization Matters.

Improvements to Layout Export

posted Aug 12, 2010, 12:57 AM by Gregory Kiesel

Fixed some annoyances with the output to Layout.  Curves and bends come in more cleanly now, and fixed a bug for the combined (single part) output.


GUI updates

posted Aug 2, 2010, 7:54 AM by Gregory Kiesel   [ updated Aug 12, 2010, 12:57 AM ]

This week I focused on updating the GUI parts of Circuit Mason.  The biggest update is with the schematic capture tool.  I've made some usability improvements on how things are displayed on the schematic.  I've also tweaked the plotting program and compiling it statically so hopefully people don't need to download the VC++ 10 run-time library anymore.

Now, if you run Mason and Update  the schematic (the "U" button), components based on a collection of Touchstone files will display not just the "value" (which file is being used), they will also display whatever is in the corresponding "File_desc" parameter... which should be the actual value.  So, rather than getting "value = 1", you will see "value = 1 (1: 0.1pF)"... a bit more human readable.

I no longer display optimization and statistical details for each part on the schematic.  I found that made the schematic too cluttered.  I may make this a flag that you can set in the GUI eventually.  


1-10 of 19