Hello Everyone Today tried to measure some structures. Even structures worked, but structures going over angles didn't. We consulted with Henry who told us, that we'd not be the first one who had this very problem. So tomorrow we prepare some more wafers. This time, we don't use plasma etching to etch the trenches but instead TMAH, which should (based on the lateral etching side effect) have nicer slopes on which the polysilicon and metal can nicely settle on.
Stay tuned.
Cheers, David
On Monday, 14 January 2019 7:13:02 AM HKT Éger Ferenc wrote:
Hello Everyone,
As per the last mumble session, my recommendations for debugging are the following:
Test subject: a native NMOS from the nmos table (no isolation/HV/etc., as only the basic layers were done). Specific measurements:
- Diode characteristization (V-I curve) between source-bulk, while all other
terminals left floating. Aim: check metal1 and contact continuity, check doping. Expected result: diode-like behavior. Possible failures: linear I-V curve (faulty doping), no current flows at all even when source is more negative than bulk (broken metal1, broken contact) - Diode characteristization (V-I curve) between drain-bulk, while all other terminals left floating. same as above. - V-I curve between source-drain, while gate and bulk tied to source (not to be performed if poly proves to be broken). Aim: check if end of the poly gate is indeed extending over fox edge (concern due to unexpected fox over-etch). Expected result: no current flows while drain is more positive than source. Possible failures: linear I-V curve (not enough overlap causes "bridging" between S-D).
The same may be done on a PMOS as well (just with opposite polarities).
Regards, Ferenc