I’m new to Eagle, so this might be an easy question, but I’ve been fighting it last night and this morning and haven’t figured it out yet.
Normally, to switch from the top layer to the bottom layer you run your trace, click to lock it to a location, switch from the top layer to the bottom layer in the Select Layer dropdown and continue to draw your trace. When you’re done a via shows up. Works great.
I want to put the via inside a pad, so I have the top trace entering the pad, then I click to the bottom layer and continue. This doesn’t work once the trace has locked to the pad.
I’ve tried manually placing a via on the pad but I get a Yellow X. I’ve also tried creating a via next to the pad, which works fine, then moving it over the pad, but again, I get the yellow X.
Can you tell me how to trick Eagle into allowing me to put the via on top of the pad?
I love it when I can answer my own question The trick is, assuming your pad is on top, to start with the bottom trace, route it under the pad, switch to the top and continue. Then it works fine.
We just ran into a major problem with this, requiring us to scrap 100 initial run PCB’s. The problem was the board house couldn’t solder components to a pad that had a via under it, since the via would suck up all the solder. The only solution (since we also had a custom stencil made) was to scrap the boards and re-route them with a 6 mill offset on the via’s that were run in the pads.
Moral of that story, 2 via’s in pads cost us $1000 in production costs.
Strange, I use vias inside thermal pads on QFN devices for a little extra heatsinking. I’ve had 4,500 boards assembled that way with no issues. Of course, I’m talking a pad that’s pretty large compared to the via, not a via in a 0603 pad etc.
Your board house may have used a different soldering method, ours was using a stencil and solder paste, we could have had a thicker stencil made but it would be more expensive than just redoing the boards.
I have met with PCB assembly guys saying that vias on the pad will make it hard for them to solder the components as the vias will suck most of the solder paste. But, I also found out in some PCB technical documents which has some patents on PCB routing shows that vias on pad are best and provide less noise on decoupling capacitors (for example). I guess maybe you have to balance them, if your assembling capability is capable of soldering components with vias on pad, I think that would be the best.
It was just an issue that never came up, I guess my board house was able to see the vias in the solder pads and just deal with it when making the stencil.
Shifted:
We just ran into a major problem with this, requiring us to scrap 100 initial run PCB’s. The problem was the board house couldn’t solder components to a pad that had a via under it, since the via would suck up all the solder. The only solution (since we also had a custom stencil made) was to scrap the boards and re-route them with a 6 mill offset on the via’s that were run in the pads.
Moral of that story, 2 via’s in pads cost us $1000 in production costs.
For only 100 boards, they could have easily touched up the pad in question with an iron and wire solder for a nominal fee so you didn’t have to scrap all the boards. I’d consider that lazy on the board house’s end for not working with you to fix a simple problem.
On the original post: was the via ‘named’ correctly? If you drop a via on it’s own with the ‘via’ tool, it’s assigned its own net name. You have to name it with the name tool the same as the net you want it attached to. This stumps a lot of new Eagle users.
The trick is, assuming your pad is on top, to start with the bottom trace, route it under the pad, switch to the top and continue. Then it works fine.