We got no information from SkyWater concerning ele...
# magic
t
We got no information from SkyWater concerning electromigration limits, but based on the material (Aluminum) I expect that the limits should be as follows: Metals 1 and 2, 0.65 mA/um; Metals 3 and 4, 1.5mA/um, and Metal 5, 2.3mA/um.
k
Hi @Tim Edwards, can you also tell about the electromigration limits of local interconnect (li).
t
I don't have experience with titanium nitride to know what a likely value would be.
j
@Alfonso Cortés
a
Hello @Tim Edwards, have you updated these current limit estimates? You recently mentioned that the maximum current for 143um of metal3 was 150mA (on channel #ieee-sscs-dc-22, about the analog power pads of the Caravan harness)
t
The 1.5mA/um number should be correct (and therefore the maximum current for 143um of metal3 should be 215mA). According to Weston in the #ieee-sscs-dc-22 discussion, the gold bond wire should not take more than 150mA, so I didn't want to state anything above that number, since it won't do any good for the pad to survive the current if the bond wire doesn't.
a
Ok, I understand. Thank you
All of these are electromigration limits, right? How much higher can we expect the destructive current limits to be? (if we do not plan to use the chip continuously and want to go over electromigration limits). According to Weston, the fusion limit for the bond wires is just above 300mA. For the metal layers would it be also around twice the electromigration limit?
t
Not sure. That's not my area of expertise; I'd need to look up what are typical values for Aluminum, which is what I did for the electromigration, since SkyWater didn't provide any values anywhere that I could find.
a
@Tim Edwards are the values above at ambient temperature or at 110 degC? Thank you!
t
@Andrea Mifsud: The values are at ambient temperature.
👍 1