NMR Spectrometer Probe Tube Machining
A working explanation of what makes these parts hard: thin walls, non-magnetic material, tight concentricity and clean surfaces. Written for design engineers and buyers who need to judge whether a probe component should be machined, and where the process limits sit.
What the Probe Tube Actually Does Inside the Magnet
A probe tube sits in the bore of a superconducting magnet, so it lives in a field that never switches off. Any ferromagnetic content in the part distorts the field across the sample volume. That single fact drives most of the material decisions in NMR spectrometer probe tube machining: no free iron, no magnetic stainless, no steel tool marks left on a finished surface.
The tube also has to hold a sample spinner at a known height, pass RF energy from the coil to the sample, and let temperature-controlled gas flow around the sample. Those three jobs fight each other. A wall thick enough to hold concentricity adds thermal mass. A wall thin enough for fast gas exchange is easy to crush during clamping.
Most probe assemblies are not one tube. They are a stack: an outer body, a sample guide insert, a coil former, tuning rods and a set of alignment shoulders that must all share one axis. If that shared axis drifts, the shim set has to work harder and line shape suffers.
So the drawing is rarely the hard part. The hard part is holding the relationship between features across several setups, in a material that is soft, gummy or brittle depending on which grade the physicist specified.
- 1Sample guideSets spinner height and sample position in the coil.
- 2Coil formerCarries the RF coil; wall must be uniform.
- 3Outer bodySeals gas flow and indexes into the magnet bore.
- 4Tuning hardwareNon-magnetic rods and screws, often small threads.
Material Purity Sets the Ceiling on Everything Else
Probe parts are usually machined from non-magnetic austenitic stainless, titanium, brass, beryllium copper, PEEK, PTFE or POM. Each behaves differently at the cutter. 316L and 304 gummy up on light finishing passes and work-harden if the feed is too low. Ti-6Al-4V cuts cleanly but holds heat, so the wall can move after the part cools.
PEEK is the common choice for coil formers and sample guides because it is non-magnetic, machinable to a good finish and stable at cryogenic temperatures. It also has a low thermal expansion compared with most plastics. The trade-off is that PEEK is expensive and abrasive on tooling, and thin sections can stress-crack if the wrong coolant is used.
Cleanliness matters as much as grade. Swarf from a previous steel job, a magnetic chuck, or a steel file used during deburring will show up in the spectrum. Shops that run these parts keep dedicated tooling and separate deburring benches for non-magnetic work. Ask how material is segregated before you award the job.
Beryllium copper is the exception worth a conversation. It machines and conducts well and is used for some RF structures, but beryllium dust is a health hazard. It needs controlled cutting, extraction and cleaning. Not every shop is set up for it.
Wall Thickness, Bore Depth and Where Machining Stops Working
Thin-wall turning is the core operation. On a 6 mm outer diameter with a 0.4 mm wall, radial cutting force pushes the wall away from the tool. The usual fix is to support the bore with a close-fitting mandrel, take light passes, and keep the tool nose radius small. Long bores often need a pilot or a follow rest rather than a long unsupported boring bar.
Concentricity between the outer diameter and the coil seat is what you are really buying. A 0.02 mm runout on a 5 mm coil seat is usually acceptable. Below that, the shop needs a mill-turn center or a single-setup 5-axis process so the datum never changes. GreatLight runs 16 simultaneous 5-axis machining centers and 16 mill-turn centers for exactly this reason.
Deep bores with a depth-to-diameter ratio above roughly 8:1 are where cost climbs. Beyond 12:1, chip evacuation and tool deflection start to dominate, and you should expect to redesign the part or accept a two-piece assembly. Gundrilling is an option for straight bores, but it does not produce an internal shoulder or a cross hole.
Slotting and cross holes in a thin tube are the second risk. A slot milled through one wall releases internal stress and the tube bows. The practical answer is to rough, stress-relieve, then finish, and to keep the slot away from the coil seat.
- 1Under 0.5 mm wallExpect a mandrel and light finishing passes.
- 2Depth-to-diameter over 8:1Cost and lead time both rise.
- 3Cross holes in thin tubeRough, relieve, then finish to size.
How to Verify a Probe Tube Without Touching the Wrong Surface
You cannot clamp a finished coil seat in a vise and expect it to stay round. Inspection has to work around the same problem the machining had. Coordinate measuring machines with low-force scanning heads handle most external geometry. For bore and wall thickness, ultrasonic or optical methods avoid contact altogether.
Surface finish on the sample path is a functional requirement, not cosmetics. A Ra 0.8–1.6 μm finish is typical on gas passages and sample guides; Ra 0.2–0.8 μm appears on sealing faces and coil seats where RF losses matter. Roughness is measured along the axis of flow, not across it.
Reports should show the datum you care about. A certificate listing every dimension to three decimals is less useful than a short report showing outer-to-inner concentricity, wall thickness at four clock positions, and the finish on the sample path. GreatLight inspects 100% of parts before shipment and issues reports on request.
Keep one finished part as a reference. When a second batch arrives two years later, comparing against a physical master catches drift faster than reading a spreadsheet.
Which Process Fits Which Probe Feature
Use this as a starting filter before requesting a quote.
| Feature | Typical process | Watch out for |
|---|---|---|
| Sample guide tube, thin wall | Mill-turn, mandrel supported | Wall deflection on finishing pass |
| Coil former, PEEK | 3-axis mill, sharp tooling | Stress cracking, abrasive wear |
| Outer body with shoulders | 5-axis, single setup | Datum shift between setups |
| Deep straight bore | Gundrilling then honing | No internal shoulder possible |
| Cross hole or slot | Mill, then stress relieve | Tube bows after material removal |
| Sealing face | Fine turning, Ra 0.2–0.8 μm | Tool marks across the seal path |
| Small non-magnetic threads | Thread mill, low feed | Galling on 316L and titanium |
When to Machine, When to Redesign
If the tube is a single thin-wall part with one coil seat, machine it in one setup on a mill-turn center and inspect the seat, not the whole drawing. If the bore is deeper than 12:1 or the wall is under 0.3 mm, split it into a tube plus an insert, or move to a molded PEEK former. Machining a part that the geometry already beat is the most expensive way to learn this.
Questions Engineers Ask Before Ordering
Can you hold ±0.005 mm on a thin-wall probe tube?
We hold ±0.005 mm on stable features with a solid datum, such as a shoulder or an outer diameter near the chuck.
On a free-standing thin wall with no support, the wall moves more than the tolerance during cutting. In that case we support the bore and quote the achievable concentricity instead, which is usually the dimension that matters.
Which stainless grades are safe for probe work?
Austenitic grades such as 303, 304, 316 and 316L are non-magnetic in the annealed condition and are the usual choice.
Avoid 420, 430, 440C and 17-4PH for parts near the sample, since they are martensitic or precipitation hardening and respond to the field. If a part must be hard, use titanium or a non-magnetic coating on an austenitic base.
How do you prevent cross-contamination from steel jobs?
Non-magnetic work runs on dedicated tool holders and separate deburring benches. Files, brushes and tumblers used on steel never touch a probe part.
Parts are cleaned and bagged individually. If you need documented segregation for an audit, say so on the RFQ.
What finish do you recommend on the sample path?
Ra 0.8–1.6 μm is the working range for sample guides and gas passages. It is smooth enough to avoid trapping particles and cheap enough to inspect reliably.
Go to Ra 0.2–0.8 μm only on sealing faces and coil seats where RF loss or leakage is a real concern. A mirror finish on a gas passage adds cost without changing performance.
Can you machine a one-off prototype and then a 200-piece run?
Yes. There is no minimum order quantity, so a single prototype and a 10,000+ part run use the same process route.
For a probe program, we keep the fixture and the setup sheet from the prototype so the second batch starts from the same datum. Quotation and a DFM analysis come back within 12 hours.
Do you sign an NDA for probe designs?
Yes. Uploads are treated as confidential and an NDA is available on request before you send drawings.
We can also work from a simplified drawing that shows only the functional dimensions if your geometry is sensitive.
Send the Drawing, Get a Machining Verdict
Upload a probe tube or coil former drawing and we will tell you which features are machinable as drawn, which need support, and where the cost sits. Quotation and free DFM analysis within 12 hours.
12-hour quote100% inspectionNDA on request
Follow GreatLight
We publish setup notes, tooling trials and inspection data from the factory floor.
