[eispice] Future Direction of eispice Part 2

  • From: Charles Eidsness <charles@xxxxxxxxxxxxxxxxxxxxx>
  • To: eispice@xxxxxxxxxxxxx
  • Date: Sun, 11 Mar 2007 10:10:33 -0400

Hi All,

It's been a while since the last eispice release, and will probably be
another month or so until the next (I've been busy lately, and the next
features are proving to be toughies). Regardless I thought now would be 
a good time for a status update.

In my last "Future Direction" email I mentioned a few things I would
like to work toward, some of those features have been realized, like the 
PyB model, native Windows support, and migration to the Numpy library. 
Others are works in progress. I also added the eide tool in hopes of 
attracting more users, who may have been scared away with all of the 

Looking ahead some of the features I would like to add are:

1. Improved T-line models:
- For the next release I'm adding a W-Element-like multi-conductor, 
frequency-dependent model. It's based on the same Kuznetsov / Shutt-Aine 
paper that was used as the basis for the W-Element in HSPICE. This model 
will be the primary new feature in the next release.

2. A Netlist Import / Manipulation Tool:
- Nothing but a distant flicker on the horizon, though I think it's 
still a useful feature and it will hopefully get added someday.

3. A Digital Signalling Layer:
- I'm probably going to tackle this one next, so that IBIS models can 
drive real signals, not just a single rising or falling edge. This 
should be ready for release 0.12 (or maybe 1.1, I think I may go to 1.0 
for the next release, the simulator core is pretty stable)

4. A Graphical Schematic Entry Tool:
- This would be added to eide. I'm not a huge fan of schematic entry in 
simulators but I think it may increase the number of users, it seems 
like most commercial tools have some sort of schematic entry. I will 
always want to support the current scripting method of defining circuits 
as well.

5. An Improved Graphing Tool:
- The current graphing tool is kind of weak. I assume some folks are 
using matplotlib. I started work on an improved graphing tool under eide 
in PyQt a while ago but put it on hold, I'd like to finish it someday.

6. Integration of a 2D Field Solver:
- Once I add the W-Element it won't be of much use without a 2D field 
solver. MMTL is the best candidate, but I would like to integrate it 
into eispice with a Python API and a PyQt based GUI added onto eide 
(MMTL's TNT is TCL based).

7. Native Semi-Conductor Models:
- The main purpose of eispice is for IBIS Model based pre and post 
routing analysis so I haven't had a need for discrete semi-conductor 
models but simple diodes, BJTs, FETs, etc... may be useful for others. 
Maybe they could be added as PyB models first, followed with faster C 
based models.

8. Layout Import:
- It would be cool if there was some way to extract T-Line info from a 
layout (maybe from Gerber files) for automated post-routing analysis.

9. An Improved Scripting Interface for Large Circuits:
- Eventually I would like to build on Thomas' eispice helper tools and 
integrate them into eispice.

Is there anything else you would like eispice to do? Has anyone used 
eispice for anything beyond the trivial examples and test files that 
come with it? If so did you have any issues? Did it fail to do something 
that you were hoping for? I'd be happy to get any feedback on user 
experience and impressions to further improve the tool. Is anyone using 
eide? If so are there any comments in it?

Any input is appreciated, and if you would like to get more involved in
development just let me know what you're interested in working on and we
can coordinate development. Also, if anyone has any questions about the
architecture of eispice, i.e. "I want it to do this, what do I have to
change?", etc... or suggestions on how to do things better please don't
hesitate to ask.

If you're interested in learning about spice-like simulators, or already
know all about them and have some ideas on how to make them better now
is a great time roll up your sleeves and get involved. I'm happy to add
code, features, etc... from others, even if it's experimental stuff.

To unsubscribe from the eispice list send an email to:
eispice-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field

Other related posts:

  • » [eispice] Future Direction of eispice Part 2