Hi Zach, thanks for the videos, they are really helpful! I have noticed in this video that you haven't used teardrops. Is there a special reason for it or is it something that you do at the end of routing process? What is your general opinion on teardrops and do they really increase reliability that much? What is, in your opinion, the point where you say "I will not use teardrops, its not worth it..."? Thanks for the feedback. Cheers!
Teardrops are not a requirement if the pad size is appropriately large. Some fabrication houses will demand the teardrops anyways on Class 3 designs just as an insurance policy. And in any case, it is something I always wait to do until the end of routing and review, that way I don't have to do them over if the routing needs to be changed.
I have a room defined around my FPGA (XC7). When I "import changes from schematic" it wants to delete the room. How do I define or sync the room on the schematic?
Is there a video anywhere that addresses the considerations when attaching ground to a ball vs a signal? I noticed that there were balls that seemed to be connected to the ground pour. I thought this was dangerous as the ground pour will not heat up at the same rate as the signal, which may cause the solder to not connect to the pour because the pour never gets as hot as the trace's pad?
I have not addressed this in a video, but with an appropriate reflow profile this is a non-issue. If you look you will find evaluation products with fine-pitch BGAs that use pad attachments directly to copper pour, and these are produced at volume. For controlled impedance lines, it is important to maintain an acceptable spacing between the line and the pour, which I have discussed in another video, and that spacing may eliminate your ability to use copper pour to connect to the balls.
Hi Zach, thanks for the videos, they are really helpful! I have noticed in this video that you haven't used teardrops. Is there a special reason for it or is it something that you do at the end of routing process? What is your general opinion on teardrops and do they really increase reliability that much? What is, in your opinion, the point where you say "I will not use teardrops, its not worth it..."? Thanks for the feedback. Cheers!
Teardrops are not a requirement if the pad size is appropriately large. Some fabrication houses will demand the teardrops anyways on Class 3 designs just as an insurance policy. And in any case, it is something I always wait to do until the end of routing and review, that way I don't have to do them over if the routing needs to be changed.
I have a room defined around my FPGA (XC7).
When I "import changes from schematic" it wants to delete the room.
How do I define or sync the room on the schematic?
Is there a video anywhere that addresses the considerations when attaching ground to a ball vs a signal? I noticed that there were balls that seemed to be connected to the ground pour. I thought this was dangerous as the ground pour will not heat up at the same rate as the signal, which may cause the solder to not connect to the pour because the pour never gets as hot as the trace's pad?
I have not addressed this in a video, but with an appropriate reflow profile this is a non-issue. If you look you will find evaluation products with fine-pitch BGAs that use pad attachments directly to copper pour, and these are produced at volume. For controlled impedance lines, it is important to maintain an acceptable spacing between the line and the pour, which I have discussed in another video, and that spacing may eliminate your ability to use copper pour to connect to the balls.
Don’t waste your time watching this if you want to know why you’d use a SMD pad over a NSMD pad
Hi