Mail Archives: geda-user/2011/10/27/01:03:45
On 10/26/2011 07:48 PM, Gabriel Paubert wrote:
> On Thu, Oct 27, 2011 at 12:07:57AM +0100, Peter Clifton wrote:
>> On Wed, 2011-10-26 at 10:27 +0200, Gabriel Paubert wrote:
>>
>>>> Personally, I've noticed a big improvement in snapping to the ends of
>>>> off grid lines (thanks Peter!), which is something I often need to do.
>> (Thanks!)
> And thank you very much for fixing the insert tool! Just pulled and
> compiled, it works fine.
>
>>> Indeed a few things are easier to do, so there is definitely progress
>>> in this area. But I've also noticed sometimes problems with microscopic
>>> examination of the result.
>> If you can reproduce it, and have a test-case with current git HEAD,
>> please send it along. It may be as simple as a computation which doesn't
>> take rounding into account properly.
> I shall try when I hit it again. I have also a polygon dicer (at least I believe
> it is) problem somewhere which I want to submit when I have more details.
>
>>> I'm designing right now with PCB some very high resolution circuit on alumina,
>>> where design rules can be down to 10µm/0.4mil minimal width and spacing,
>>> although I stay at a more reasonable 50µm/2mil for my first thin film circuit.
>> What is the application?
> Switching of microwave signals. Actually there are two different circuits,
> one based on RFMD's RFSW2041D and another one based on Hittite's HMC975.
> Both circuits have a square shape whose side is 8.89mm (350 mil).
>
> Actually the design of the (milled aluminum) enclosure is (at least for me)
> much more difficult than the PCB, in which everything is at an unusual
> scale, but otherwise very simple (the most complex of the 2 circuits
> has less than 10 components).
>
> Regards,
> Gabriel
>
>
Speaking of microwave design, I wonder, if one wants to do stripline
design and has to use open-shunt stubs, can the ends of the traces be
straight instead of oval? If not, could one workaround be, perhaps,
creating patch (one-pin footprint) to terminate an open shunt? The
width of such pseudo-component would be the same as the width of the
trace it "terminates" and the total length of the two would be the
calculated length of the shunt.
I am not sure how this trick would work for microstrip, though, as the
terminating "pseudo-component" would be positioned between planes. This
case is probably very rare.
I use the version of PCB that came packaged with Ubuntu 11.10.
Regards,
Gus Fantanas
- Raw text -