After resolution, it’s the question every desktop SEM buyer asks: will my parts fit? It’s a better question than most people realize, because the answer has three different numbers in it — and vendors have been known to quote whichever one flatters. Here’s the honest geometry of a benchtop SEM chamber, what the numbers mean in practice, and the mounting craft that gets awkward real-world parts under the beam.
The three numbers that matter
1. What fits in the chamber. The SNE-Alpha accepts samples up to 70 mm in diameter — a poker chip, a small gear, a fastener head, a packaged device, most coupons and stubs-full of parts, a 2-inch wafer with room to spare.
2. What the stage can show you. The motorized stage travels 40 × 40 mm in X and Y. That’s the window you can navigate under the beam in a single loading. A 70 mm part fits physically, but you’ll survey a 40 × 40 mm region of it per load — so when the region of interest is known, mount the part with that region centered, and when it isn’t, plan a reposition-and-repump step. At 90 seconds to vacuum, repositioning costs less coffee than it used to.
3. How tall the sample can be. Height trades directly against working distance — the space between the final lens and the sample surface. The Alpha’s working distance runs 5–40 mm with 40 mm of motorized Z travel: a tall sample eats into that budget, and the highest-resolution imaging happens at short working distance. Practical translation: low, flat mounting is always the right instinct, and genuinely tall parts should prompt a quick conversation (or a saw) before purchase, not after.
Add the rest of the 5-axis stage — tilt from −45° to +90° and full 360° rotation — and a surprising range of geometry becomes imageable: sidewalls, bore edges, fracture faces at raking angles, threads in profile.
Mounting is a craft, and it’s learnable
Most “will it fit” problems are really “how do I hold it” problems. The standard toolkit:
- Pin stubs and carbon tape for powders, fragments, filters, and anything light — fast and disposable, with the tape doubling as the conductive path.
- Clamp and vise holders for machined parts, PCBs, and anything you’ll want back unmarked — mechanical grip beats adhesive for heavy or precious pieces, and grounds them properly.
- Cross-section pucks for mounted-and-polished metallography samples — the workhorse format for coatings, welds, and interfaces, and the same puck moves between the SEM and your optical scopes.
- Multi-stub holders to batch several samples per pumpdown — with a 90-second cycle and a motorized stage, a morning’s QC queue goes through in one session.
- Silver paint and grounding straps wherever tape can’t guarantee a path — charging from a poorly grounded mount is misdiagnosed as “bad instrument” weekly, somewhere in the world. (If the sample itself can’t be made conductive, that’s what low-vacuum mode is for.)
When the part is bigger than the chamber
It happens: a full engine component, a 300 mm wafer, a weldment the size of a bread box. Your options, in the order we’d suggest them:
Section intelligently. Most failure and QC questions live in a small, identifiable region — the fracture face, the discolored zone, the weld toe. A bandsaw or precision cutoff wheel, used far enough from the region of interest to leave its story untouched, turns a too-big part into a routine sample. The skill is deciding where the evidence is before cutting; when in doubt, photograph everything first and cut generous.
Replicate. For surfaces that can’t be cut — a die in production, a huge mold, an aircraft component — replica tape or silicone compounds capture surface topography faithfully down to well below a micron. You image the replica instead of the part. It’s an old trick from the field-metallography world and it remains underused.
Use witness coupons. For coating and process QC, a small coupon that rode through the same process run answers the question without ever touching the production part.
Or admit you need a big chamber. Whole-wafer inspection at 150–300 mm, complete assemblies, aerospace panels imaged intact — that’s large-chamber floor-system territory, with the facilities and budget that come with it (we covered that math in the cost-of-ownership post). If your daily work is whole large parts, buy the instrument that swallows them. If it’s occasionally a big part and mostly normal samples, the section-replicate-coupon toolkit covers the gap for a fraction of the cost.
A quick reality check from the applications world
It’s easy to over-index on maximum capacity. Scan the sample lists in our materials science and metallurgy application notes and notice how few samples are large: powders, particles, fibers, fracture fragments, polished sections, filters, fasteners, electronic components. The 70 mm chamber isn’t the compromise it sounds like — it’s sized for what actually crosses an SEM stage, which is why the benchtop category works at all.
Frequently asked questions
Can I image one corner of a part that’s larger than 40 × 40 mm?
Yes — anything up to the 70 mm chamber diameter loads, and you center the corner you care about in the stage’s 40 × 40 mm travel. Larger surveys are a mount-image-reposition cycle.
How many samples can I run per pumpdown?
A multi-stub holder carries a batch of standard pin stubs, and the motorized stage drives between them — the practical throughput limit is usually your prep bench, not the microscope.
What about heavy parts?
The 5-axis motorized stage handles routine machined parts comfortably; for genuinely heavy pieces, ask us before assuming either way — mass, balance, and tilt interact, and the answer is geometry-specific.
Do irregular, non-flat samples image well?
Usually beautifully — enormous depth of field is the SEM’s party trick, and the tilt/rotation axes let you present the surface to the beam. The care point is grounding and, for insulators, charging management.
Not sure whether your parts fit?
Send us dimensions and a phone photo — that’s genuinely all it takes for a yes, a no, or a mounting suggestion. And if it’s a maybe, we’ll load your actual part on an SNE-Alpha during a demo and settle it empirically. Get in touch.