Homebrew PCBs

soffee83 wrote:

>Thanks for the tip! I'll go grab one to check out.
>
>I've finally gotten fairly proficient with the handful of functions 
>that I need from Eagle, but have always been bothered by the limits in 
>the low end of the Eagle family. I really wish they had a 100% "non-
>profit" version with full functionality which was still within a 
>reasonable price range for hobbyists. Last I checked, I think they 
>moved up to one which would allow you a 'slightly' larger board, and 
>then one for limited commercial use or something. The fifty to one 
>hundred dollar range seems good for those who have absolutely no 
>intention of making money off it. Seems like even with more students 
>and hobbyists cutting their teeth on it, they'd get even more business 
>from those who go professional.
>
>Those "freeagle" board sizes can be a PITA, especially if you're doing 
>something that really just needs space (like pots or large parts).
>
>-George (sorry to sound like a cheap-ass) 
>
>  
>


  That isn't just you, their licensing scheme is somewhat brain 
damaged.  Creates a sizable gap between needing hundreds of dollars to 
use it commercially, or only use it for free with no profit..  The sort 
of thing that gets come up with by those who are used to having access 
and the company they work for to buy everything for them, and not quite 
enough thinking put into it to realize others may not fit their narrow 
minded box..  Should definitely have another aspect to the limitations, 
some things simply need space even though the complexity is very low.  
It's like the people writing and selling it have never heard of anyone 
making a PCB that isn't a damn tightly packed Eurocard..

  While I don't know the orginal poster from a the next guy on the 
street, I do share Mike's skepticism on the "I've been dealing with the 
limitations of eagle for making pcbs for a few years now, and finally 
hit a problem I could not resolve."  It is likely to boil down to one of 
three options:

1.  The 'problem' is real but actually tiny, and the OP's idea of an 
'unresolvable' problem takes 5 minutes to figure out with Google.
2.  The 'problem' is imaginary, totally manufactured to stump the OP's 
agenda, paid or not.
3.  The near zero % chance of something that doesn't fall into the two 
above options.


  It's one of the least limited programs I've ever used.  A bit obtuse 
on how to do some things, but that's what Google is for.  And at least 
when you get past your own 'teething' of how to use it, it's potential 
is already largely there.  Not some far off thing you want that may 
never even be programmed in..

Also while I'm posting here's a draft from the Autorouter thread.  Not 
everything but a start on how to make the autorouter work correctly, I 
can make it do SS or SS+jumpers or DS etc with no problems at all, did 
take a month or two of playing in spare time to figure it out though..


Old draft from when Yahoo wasn't letting emails go through, but still 
some may find it useful..


Peter Harrison wrote:

>
>For My own preference, I suspect i do not know how to best direct the 
>autorouter in Eagle - too many options - and the default options are not 
>too useful. I tend to let the auto router do either just some traces 
>then hand route the rest or autoroute a board then tidy it up by hand 
>until it looks nice and satisfies any other requirements I have.
>  
>

  Isolate and identify the settings.  Reroute at each step, and pay 
attention to what the changes do, and learn what's going on when you 
change each setting.  Note that if a setting is say 5, and you change it 
to 4, does it now come right after 3, or right before 5, after all the 
other 4's?  Recognize that you have to spread the values out widely, so 
you can be in charge of what's going on, instead of letting the 
autorouter follow semi-random or arbitrarily preprogrammed preference.  
Pot luck is not conductive to having anything do what you want it to..  
Having groups of settings at 10, 20, 30, etc so you can bump things to 
29 or 31 etc and exert fine control over the weighting is absolutely 
necessary.  1-5 or 1-10 for most settings like it's originally set up 
drastically limits the possible control over the system..

  I thinik this is 95% of most people's problems with the Eagle 
autorouter, they don't mess with it enough to realize you simply have to 
spread the values to get control over the routing, and the default 
values are way too close for good operation.  Sort of a PITA to fix them 
all and figure out what the relative values should be, but necessary.  
Of course once you start seeing the most critical settings and what they 
do and start getting them in correct ranges it goes faster, the other 
settings have effect but 5 or 6 main ones control most of the action..

>It seems to me that parts placement is the real art. I don't think I 
>have adequate skills in that area.
>
>  
>
  Get good at spinning the parts by 90 or 180 degrees by the right 
button, putting the part back down, and hitting the ratsnest button to 
shorten the airwires.  Flip a part where it's practical to straighten 
the wires and you have a bus from one part to another that flips.  Tiny 
changes in placement can still produce major differences in the results, 
but it's hardly a drawback of an autorouter, it can do the same thing 
when you route manually.  You can try 10 or 20 variations in the time 
it'd take to do 2 or 3 by hand, it is not more work it is less, and you 
can get a far better job done..

  Most of my boards are relatively small and tight, and routed by hand 
for shaped routing and other things that an autorouter simply wouldn't 
be able to handle.