Favorite free engineering CAD tools

LTSpice, no limit whatsoever.

Xilinx WebPack... pretty much covers any of the devices someone doing development on a shoestring budget is going to be able to afford. :-) Altera's freebies are decent too.

Good choice.

I left out the items where I don't have a favorite or have little experience.

Hello Mook,

Like Joel I prefer LTSpice. The wiewing window for sim results is IMHO a bit hokey (x-axis keeps popping back to unusable ranges) but other than that a fine program.

Only FIR filters. I like some old DOS programs better but my favorite one from Mildenberger talks German so I guess it wouldn't help you. I wish the more fancy ones would go into wave digital filters and stuff. But they don't.

If you design a lot of analog filters of the LC kind try Aade's Filter Design. Not very comfy but fast.

Very good. Licenses for "real" versions are much more reasonable priced than other tools. This is one case where the willingness to offer a useful free version has paid off. I bought it shortly after kicking the tires on the free version.

Never tried it but my experience is that this stuff really needs to be tested in the lab, and very thoroughly.

Try the EZ430 kit from TI. Not free but only $20 for USB programmer, target board and C- plus assembler tool. Can't beat that, I believe. The IAR basic Kickstart tool is free BTW in case you decide to set up your own hardware. But you couldn't possibly do that for under $20.

AppCAD from HP. Saves you from whipping out the slide rule, the calculator and the formula books too often.

Then GCPreview for checking Gerbers. No "see-through" plane patterns but heck, there is a free version so I won't complain.

Checkout Reg Edward's site, has a lot of RFfy programs,

martin

Joerg, Is there any software that allows to get started with WDF filters? German language is not a problem.

OK, if we're going to branch out into other free software tools:

Another good free Gerber previewer is Viewmate by Pentalogix. Viewing AutoCAD/SolidWorks files: eDrawings More LC filter design: Elsie (AADE is quite good too -- the two are just a little different) Transmission line design: TxLine from AWR Text editor: EditPad Lite

---Joel

In message , dated Mon,

With the focus on the sim results, click Plot Settings -> Save plot settings. When you get a new plot with axes you don't like, click Plot settings -> Reload plot settings. Your preferred axes will be restored. You can use short-cut keys, of course.

I just wish there was a comprehensive manual. But it would be a mammoth task to compile and couldn't be done without the full co-operation of the author. Many of the most valuable features are undocumented; you have to learn them from others.

LTspice

Xilinx WebPak

Appcad

RFSIM99.

FireFox!

John

Hello Nico,

Yes, the program written by Prof.Mildenberger's group at the Fachhochschule Wiesbaden in the late 80's. Unfortunately now that he's retired they dumped his web site. Since the web archive butchered the file one of the fellows in the German NG repaired the zip file:

I have the older licensed version but later it became free.

There is also a large set of routines from TU Delft but they did the usual, it's all Matlab and very incompatible with anything else. Even within Matlab versions. So I could not use that. Had they chosen to create a "real" stand alone program from it the TU in Delft could have gained lots of prestige instead of seeing it all go towards that dusty shelf in the basement.

Hello John,

On the last one it croaked even before that. Right click in the plot window -> manual limits -> entered new right limit. Zilch. Nada. Just sits on the (useless) default.

Yes, but luckily there are some very helpful (and English-speaking) folks like Helmut Sennewald.

The Mathworks has had a compiler for Matlab (that can create standalone applications) for years now, although apparently it's one of those things where often universities have a site license for Matlab itself (and various toolboxes) but often don't end up licensing the compiler.

---Joel

I got recommended eagle by someone, tried it, loved it. works on windows and linux too, so I can use it at home and at the girlfriends place.

Mark

Hello Joel,

Well, universities often just don't finish stuff. If they do then they don't maintain it long term. Even the school where the father of wave digital filters (Professor Fettweis) taught is a sad example. The center of competence around him seems to have disintegrated. The result? About

This is one of the reasons why I decided to take the plunge and not pursue a Ph.D. I firmly believe that one must have a strong sense of long-term commitment and not hop from one interesting project (or job) to another just because something else looked more glamorous.

Hello Mark,

It is a great SW and much lower in cost than OrCAD. Plus they don't force-upgrade people by coming out with new versions all the time which is what finally convinced me to switch from OrCAD to Eagle. Also, they actually communicate with their customers a lot (what a concept...).

But using Eagle at the girlfriend's place is usually frowned upon (by the girlfriend).

Hi Joerg,

There's no funding in making "polished products," as far as I can tell... you get your software/design process/whatever to the point where you can publish a paper or two -- knowing that it has plenty of quirks/bugs/etc., but all ones that you know of so that you can work around -- and then it's on to the next project. :-(

Well... ok, but if commercial entities were had such a strong sense of long-term commitment, don't you think that software would generally be somewhat less buggy than it is today? Most companies seem to spend more effort introducing "glamorous new features" than they do fixing their bugs...

---Joel

It was either that, or haul my workshop round there.

In message , dated Mon,

Yes, see my last sentence. It works, but it's not a terribly good substitute for a manual.

Hello Joel,

Sadly, that's their way of thinking and IMHO it's wrong. If they would collaborate and form competence centers they would gain so much more status and fame with industry. This in turn would lead to more lucrative research contracts and that would bring urgently needed money to universities. But the way it is now, during most industrial projects we don't even consider involving a university. That could be different. But only after we see them follow through with their own things.

A great example of doing it right is the Werkzeugmaschinenlabor (WZL) at RWTH Aachen University (my alma mater). They design and maintain a well rounded set of CAD tools. Consequently, in my days they were quite flush with research funds from companies who needed help in mechanical engineering. I fondly remember the keggers they regularly threw on their front lawn. Money didn't seem to be an issue there.

I've tried to talk to some about it but it is like kicking an oak tree so I stopped.

It exists. Look at Cadsoft Eagle. I have yet to have it do a hard crash and I haven't discovered any tough bug. The only weirdness is that you sometimes have to unselect and select the printer. Kind of like the double-clutching on large trucks. Oh well, if that's all I can live with it.

After a while it becomes obvious which companies can achieve quality and which ones cannot. Just like with cars where there are certain brands or models I would never buy. This can vary even within companies. For example, while IMHO much SW from MS is very sub-par there is other SW such as MS-Works which is and always has been of a solid quality. So I use that for my book-keeping.

Hello John,

True. In the case of (otherwise often excellent) university software it's worse. After a few years the folks that were involved have left and now there is nobody who has much of a clue. Only a handful of people knew how to fly it. It's like building a complicated piece of test equipment and not labeling the front panel because you know what all the buttons do.

ProgeCAD, Circuit Maker.