Kai-Martin Knaak <[email protected]> wrote: > That is, you already did the initial gsch2pcb run?
I did the uEDA equivalent thereof. I use uEDA rather than gEDA, but the output PCB elements and PCB netlist format are exactly the same. All nets are meaningfully named. > My rule of thumb for an economical size, that is still no pain to layout > is twice the combined area of all components. Would having 2 extra layers make it less pain to fit into a smaller size or not? Here is my issue, and it's a sentimental one: my SDSL modem design is the functional equivalent of a CSU/DSU (sync serial on the other end, no Ethernet), whereas all competing SDSL CPE devices are full routers. One particular SDSL CPE router that has caused me an untold amount of grief in the past was the (in)Efficient Networks 5851, and my project's goal has basically been to replace that abomination. The PCB inside EN 5851 has 6 layers (I can tell by the layer stackup markings on the side), and its functional section is approximately 130x165 mm. I have previously thought about fitting my OSDCU board into that form factor and putting it into the enclosure from a gutted EN, but I have subsequently rejected that plan for a number of reasons. I now want to make it a simple rectangular PCB, but I'm keeping the 130x165 mm dimensions. (EN's design probably wasn't metric and their PCB was an odd non-rectangular shape anyway, but I'm approximating and I like metrics.) I suppose I could make the PCB bigger than 130x165 mm if that makes the layout labor cost much cheaper, but then we'll have a device with CSU/DSU functionality that is physically larger than a competing device which is a full router, and that somehow feels "wrong" psychologically. > > Again because the project is non-commercial, mass production is not a > > consideration. > > Maybe small series :-) There is one ISP I am currently in negotiations with, they say they have some existing customers with an SDSL-to-V.35 arrangement, but those customers are served from some very very old DSLAMs. The ISP says they would like to migrate those customers to a newer DSLAM platform (still old, but not as old as the other one), but apparently their existing SDSL-to-V.35 CPE won't work with the other DSLAM. I am trying to convince the ISP to fund the NRE cost of my OSDCU (basically the cost of the layout labor I'm soliciting right now) so they can use it as their new SDSL-to-V.35 CPE solution that works with all DSLAM flavors. The ISP's response so far is that *if* they are interested (it's still an if), they will want "maybe 100 units". Well, 100 units is not a low quantity in my book - but again the whole deal is still hypothetical. I still consider my project to be non-commercial though because I want to build it regardless of whether or not that 100 unit ISP deal comes through, just because *I* want to have the toy for my own use on my own SDSL line. MS _______________________________________________ geda-user mailing list [email protected] http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
