Alloy Junction 1952 Place the Emitter Disk

# Place the Emitter Disk: Placing Blind Against a Reference You Cannot See

The collector went on a face anyone could watch directly; the emitter goes on the opposite face of the same wafer, where the collector it has to land opposite is now hidden behind the full thickness of opaque germanium. This step cannot place the emitter relative to the collector's actual position the way an operator might instinctively want to, because that position simply isn't visible from this side — placement here depends entirely on the fixture's own mechanical registration surviving the flip from one face to the other, trusting that the fixture's known geometry, not direct sight, correctly relates where this disk lands to where the collector already is.

## 1. The Target Position Is a Mirror Image, Reached Blind

$$\vec{r}_E \overset{\text{target}}{=} T(\vec{r}_C)$$

The emitter's ideal position $\vec{r}_E$ is the mirror image of wherever the collector actually landed, $\vec{r}_C$, transformed through $T$ — the physical flip the fixture performs between faces. This equation looks simple, but notice what it depends on: not the wafer's nominal center, and not where the collector was supposed to go, but $\vec{r}_C$, the collector's real, already-placed position from step thirteen, carrying whatever placement error that step actually had. This step is aiming at a target it cannot observe, defined by a quantity it cannot directly measure, using nothing but the fixture's trustworthiness to bridge the two faces.

## 2. Real Diagram: Placing Opposite a Position You're Trusting, Not Seeing

The Fixture's Flip Is the Only Link Between the Two Faces nothing about this face lets you see where the collector actually sits wafer, viewed edge-on collector, step 13 visible when placed, now hidden mirrored target, T(r_C) emitter, actually placed here gap this gap is the thing step seventeen exists to measure and correct

## 3. Two Independent Placement Errors Compound, They Don't Cancel

$$\epsilon_{\text{total}} = \sqrt{\epsilon_C^2 + \epsilon_E^2}$$

The collector's own placement error $\epsilon_C$ from step thirteen and this step's independent emitter placement error $\epsilon_E$ don't offset each other — for two independent sources of random error, they combine as a root-sum-square, which means the total misalignment between the two disks is always larger than either error considered alone, even when both individual placements were well within their own tolerances. This is precisely why alignment can't be treated as automatic just because each disk was placed carefully on its own face: carefulness at each step only bounds that step's own contribution, and the two contributions still add up on the finished assembly.

Combined Error Exceeds Either Individual Contribution root-sum-square, not simple addition, but still always larger than one error alone error magnitude → ε_C collector alone ε_E emitter alone ε_total combined, what step 17 sees neither individual bar predicts the third bar's height on its own

## Place the Emitter Disk's Place in the Process Lineage

Placing the emitter disk is step sixteen of RCA's forty-two-step alloy-junction manufacturing sequence — immediately after the first assembly was cooled, and before the two disks are ever formally aligned. It is the step that introduces this process's one genuinely blind placement, aiming at a mirrored target it cannot see and compounding its own placement error with step thirteen's already-established one rather than resolving it. Step seventeen, aligning the opposing disks, is the step built specifically to measure and correct the combined gap this step's blind placement necessarily leaves behind.

Take alloy junction 1952 place the emitter disk further

Ask the copilot about this term, or have our engineers assess it against your process.

Ask CFSGPT about this → Request a technical assessment → Open the CFS app