Kicad is starting to act nice.

Submitted by Ed_B on Thu, 10/08/2009 - 03:53

Printer-friendly versionPrinter-friendly version

It's 3:30am and I'm tired. But there's still time to post some pix of where the board is at.

The circuit got a new ferrite noise chomper on the filtered +5. The only filter on the first board is a tiny SMT thing that would be really hard to hand solder. On the Analog board, I think it's necessary because of the hash the comms lines make. The RS485 tranciever chips are supposed to be slew-rate limited, but the data wires are big antennas. There's no getting around that. The biggest change is that I moved most of the parts around to try to get a nice fit that doesn't have impossibly crowded corners. Along the way, I re-defined the i/o pin mapping onto the end-user connections. Line names (pin 1, pin 2, etc.) are defined in software anyway, so I rearranged everything to make the traces come off the chip straight without crossing over each other. This is a detail showing the new pin mapping.

 

 

The configuration of the components is right, but the spacing needs a little work. The clock crystal interfers with the MCU socket, and other things like that, but overall it looks good.

 

It's really hard to tell what it's going to look like with parts in it, but keycad has this 3-d viewer that is fairly helpful. The bugger of the thing is that the component drawings in the CAD parts library are pretty sparse. I know it's open source, and if you want componant models in the library, then go ahead and make them, and all that. But not today.

 

That's enough now.