open-source-silicon.dev

Hi,
I was wondering if/how the breakdown voltage of the diodes in the pdk scales with area. How can we determine the breakdown voltage of a diode we instantiate? I am hoping to add ESD protection to the gates of a few transistors that are connected to input pads and am not sure which diode in the pdk would be most suitable for this, or how they should be sized. Also, what is the difference between the sky130_fd_pr__diode_pd2nw_05v5 and the sky130_fd_pr__diode_pw2nd_05v5? Thanks

`pd2nw` is a diode between P-diffusion and N-well.  `pw2nd` is a diode between N-diffusion and P-well (or P-substrate).

The geometry will affect the breakdown somewhat because of voltage gradients across the device, especially if you don't have it properly contacted, but otherwise you can expect that roughly the same voltage occurs over the whole area of the diode, so area will only affect the amount of current through the diode, not the breakdown voltage.

Breakdown voltages are in the documentation;  see <https://skywater-pdk.readthedocs.io/en/main/rules/device-details.html#id74> section Diodes.  There is a table of breakdown voltages there.  Most breakdown voltages are in the 10-12V range.  This should also match breakdown voltages in the SPICE models which is, I think, `VB`;  e.g., for a standard installation of open_pdks, you could find such information in `/usr/local/share/pdk/sky130A/libs.ref/sky130_fd_pr/spice/sky130_fd_pr__diode_pw2nd_05v5.model.spice` (which has `vb = 11.7` , which matches `VBN` or N+ breakdown voltage in the table referenced above).

<@U016EM8L91B> since the device details page states that the models are only valid between 0 and 5V, we’ll never really be able to simulate reverse breakdown right? 

The model takes reverse breakdown into consideration, so it would simulate _something_, but probably not particularly accurate.

Has anyone measured the reverse breakdown voltage of the sky130_fd_pr diode_pw2nd_05v5 diode on a test chip they got back? I tried to simulate the reverse breakdown with a DC simulation in xschem but I was not able to measure any reverse current.