wafer.space Community - TT GF180 padframe

6:42 p.m.

Thanks a lot Sylvain! Just updated the project template.

6:43 p.m.

The padring now looks like this. Once you have yours we need to double check the bondpad positions. I have updated the padring script to also support pads with arbitrary widths, but for this case it should behave the same as before.

Can you post the gds ?

Sure! Without fill I assume.

Yes.

6:56 p.m.

Something looks wrong from the picture though.

6:57 p.m.

Wait, the orientation doesn't match.

6:57 p.m.

It's rotated 180 deg.

Is it? Pin 0 should be bottom left, but let's see... I'm almost at StreamOut.

Bond pad 17, the first on the right edge at the bottom should be power.

7:00 p.m.

Ok, I just realized, in the text column I wrote "left" but it's actually "right" ...

I see

Then let me check again

7:02 p.m.

chip_top.gds.gz

4.77 MB

7:02 p.m.

Here is the GDS anyways, just for the pad positions

7:05 p.m.

Alright, rerunning!

7:11 p.m.

7:12 p.m.

chip_top.gds.gz

4.77 MB

7:13 p.m.

Should be right this time!

Yeah the position of power pads match. The labels (input / bidir / clk / rst) don't but that doesn't matter, all those should be broken out 1:1 without any special consideration. It's important for the breakout that :

The PWR VDD AUX are not connected to power rings, but just broken out to pins like other , just with decoupling caps
The IO / Core VDD pins are correctly mapped to independent rings with appropriate decoupling.

Might be worth posting in the #📦-cob channel :)

Pad position is fine, it's distributed equally, so I can make sure to match that ( if for some reason my script doesn't match it already ... ). (edited)

Great! After the script rewrite I had to triple check that this is still the case

But I have yet to try a cell with a different width.