-----------------------------------------------------------
The "Panamax" Padframe
-----------------------------------------------------------
General-purpose 129-pin padframe in SkyWater sky130
-----------------------------------------------------------
Created by Tim Edwards, Efabless Corporation
-----------------------------------------------------------

The Panamax padframe design is an update of the padframe used for the
Caravel and Caravan harness chips, also known as "caravel_openframe"
(although it does not have its own repository but exists in the caravel
design at https://github.com/efabless/caravel).

The padframe was designed to accommodate a wider range of I/O types and
a much larger number of GPIO.  There are a total of 9 banks of 8 GPIOs
each, for a total of 72 GPIOs, 16 of which are overvoltage-tolerant
pads.  In addition, there is one differential SIO cell which can be
configured for use as two independent GPIOs, one dedicated reset pin,
and one additional GPIO that is not part of a bank.  The power supplies
are similar to the original Caravel design with two 1.8V domains and two
3.3V domains ("user" domains in the Caravel and Frigate harness chip
designs), one 1.8V core domain and one 3.3V ESD I/O domain, but with
two core analog domains (one more than Caravel/Caravan).  In addition,
there are six analog pads such that there is no need for separate
analog and digital use cases.

See the documentation in the doc/ directory for more details.

Note:  "Panamax" refers to the definition of the largest ship that
can fit through the Panama Canal.  The "Panamax" padframe does not
have the absolute maximum number of pads that could be fit in the
size of a standard "Caravel" harness chip area, but it is close to
it.

--------------------------------------------------------------

Issues list (as of Feb. 2025):

1. The verilog generation script is unfinished and does not generate a
   valid verilog module---The final verilog edits have been done by hand.

2. Magic needs development work to fix the issue with extracting devices
   with 45-degree angles on devices, which happens in the flanged-gate
   ESD structures in the I/O pads.  Without that, the GPIO OVT and SIO
   cells will not LVS correctly.

3. The DEF file has incorrect placement of subcells.  This was generated
   by magic and needs to be fixed;  there is some issue with the cell
   origin translation between the magic database and DEF.

--------------------------------------------------------------