Has anyone used the sky130_fd_sc_hvl library for something? It looks like it might be incomplete and missing essential tap cells. Also I created a small group of cells in a rail structure and magic DRC complains about HV Nwell to HV Nwell spacing which would prevent any real use of the cells.

Is the HV nwell spacing error between neighboring cells in the same row, or between rows?

This what i ran the DRC check on. So the error exist between the rows, since there are no neighbors.

I guess you'll need to increase row spacing

I just need to flip the middle one, such that the nwells align.

Oh, right

Hmm actually that does not work either. Now i get HVNTM spacing errors.

Can you show the HVNTM only?

message has been deleted

You didn't abut them correctly I think.

Your right Sylvain lets try to do this right.

FWIW I have had to manually patch up HVNTM in (analog) arrays before, due to magic generating some boxes too small. Shouldn't be an issue with standard cells though

Yes i know that issue i had it too with analog NMOS devices. But here it was only me not putting it together correctly. Now that everything aligns correctly i get a spacing of 0.7um from HVNTM to HVNTM which will satisfy DRC.

Check the width of HVNTM in the section. I would have expected it at the bottom, but the screenshot shows a DRC error in the upper-right

I used 3 stacked sky130_fd_sc_hvl__dfrbp cells. So there is a 16.87um wide and 1.29um high strip across the entire cell.

I would expect the bottom HVNTM strip to be too short (probably 0.645um)

I'm not sure if I follow you. Here is a more clear picture of the same layout as shown above in magic.

Make sure that this match the layer in the original GDS and wasn't regenerated by magic.

I did use klayout to build the test structure and only opened the gds in magic to run drc.

Ah, I misunderstood

This is a .gds file that i opened with magic so no .mag at all.