We're back at the probe station, jabbing little metal probes on the wafers. The earlier problems with the station having been straightened out (anomalous readings were the result of crossed cables), we're seeing more reasonable performances from the devices.
"Hey, those look good!" The professor points out the current-voltage curves on the screen. "What wafer was this again?"
"Wafer 4."
"Great. Now let's see if we can find a 10 micron device that works." He scans over the wafer to a teeny-tiny transistor and drops the probes onto it. This transistor has the same threshold voltage as the bigger working devices and the really nice current-voltage curves. "Okay! Wafer 4 is going down in history. Let's see if a 5 micron transistor works."
The 5 micron transistor doesn't work. Its parts don't seem to be aligned quite well enough. We check devices located on different regions of the wafer, but get no joy.
The professor still loves Wafer 4. We move on to another wafer and look for working 5 micron transistors there.
I have to ask. "So what if you find just one 5 micron transistor that works? Are you going to say you have a 5 micron process then?"
"Well, no? We'll see why that one worked and make adjustments to the process to try and get the yield up. We may need to adjust the dimensions and maybe improve the alignment marks."
235 words | April 26, 2005 08:33 PM | Ivory tower