On this page
The electrical curve on a throttle card has meaning only after its angular origin is connected to the shaft. Artwork can be accurate relative to substrate fiducials while the substrate seats differently in the housing, or the shaft axis can shift relative to the intended track centre. Angular phase error moves the entire curve; centre error changes effective radius and contact tracking around the sweep. Treating both as one alignment tolerance hides different risks and inspection methods.
Key design decisions
- Define angular zero from a functional shaft feature or controlled stop.
- Separate print phase, card mounting and shaft-centre errors.
- Verify contact position and connected output against an independent shaft coordinate.
Connect artwork coordinates to the throttle shaft
Identify print fiducials, substrate outline or holes, card seats, housing datums, shaft axis, keyed flat or lever feature and mechanical stops. State which features clock the card and which locate its centre. Clearance may allow several seated states; assembly force and fastener sequence determine the actual one. A substrate edge is not a functional angular datum unless the drawing controls its relationship to the printed origin and assembly clocking surface.
Add signed angular contributions at the shaft coordinate
Convert print-to-card, card-to-housing and housing-to-shaft clocking offsets into one angular coordinate. Centre displacement projected tangentially creates a position-dependent apparent phase and should not be inserted as a constant.
Δθ_phase=Δθ_print+Δθ_mount+Δθ_shaft; ΔV≈(dV/dθ)Δθ_phase
- Δθ_print locates the electrical pattern from the card datum.
- Δθ_mount is card clocking within the housing.
- Δθ_shaft relates functional shaft zero to the housing.
- dV/dθ is local connected-output slope.
Small angular offsets, a defined seated state and no centre eccentricity in the constant-phase term.
Convert a registration stack into local voltage shift
In an illustrative stack, print phase is +0.18 degree, card mounting is −0.10 degree and shaft reference is +0.22 degree, producing +0.30 degree total. At a local curve slope of 20 mV/degree, predicted output shift is +6 mV. Arithmetic tolerance limits must be applied with their signed adverse directions, not to these nominal offsets alone. The values explain the calculation and do not state manufacturing or throttle accuracy.
Analyze eccentricity as a rotating vector
Represent shaft-centre displacement from track centre as x and y components. Project the vector onto radial and tangential directions around the arc. Radial error can move a finite contact footprint toward a track edge; tangential projection changes apparent coordinate with angle. Plot both through the valid sweep. A phase correction cannot repair loss of lateral overlap. Contact-arm geometry and runout may add their own rotating vectors and should retain separate identities.
| Contributor | Primary effect | Verification |
|---|---|---|
| Print clocking | Constant curve phase | Optical fiducials to card datum |
| Card seating | Assembly phase and translation | Assembled datum measurement |
| Shaft centre | Angle-dependent radial and tangential error | Independent axis location |
| Contact footprint | Track overlap and spatial averaging | Contact-path inspection |
Use fiducials and functional nesting appropriate to each operation
Screen printing, substrate singulation and assembly may reference different features. Create a documented transfer chain and verify that inspection measures the same functional relationship. A camera can report excellent print registration to a fiducial that is poorly located to the assembly hole. Nest wear, debris and substrate edge variation can shift seating. Preserve lot and fixture identity so a phase trend can be connected to the responsible operation rather than corrected indiscriminately in artwork.
Distinguish phase, eccentricity and electrical loading patterns
A nearly uniform angular shift in forward and reverse sweeps supports phase error. A sinusoidal residual or changing track-edge margin suggests centre offset or runout. Direction separation implicates backlash or seating, while receiver-dependent voltage changes point toward circuit loading. Capture shaft angle, contact or carrier position, excitation and output together. Software recalibration can hide a phase offset but does not restore physical overlap or prove adequate endpoint geometry.
Validate registration at mechanical and electrical endpoints
Measure print datums before assembly, then locate shaft axis, card and contact in the final seated state. Sweep slowly from both directions through endpoints and interior stations using an independent angular reference. Compare output residual with the phase model and inspect contact-path margin. Repeat assembly to reveal seating reproducibility. Any work on powered throttle hardware requires application-owned safe procedures and acceptance; card-level evidence alone cannot authorize system use.
Release clocking, centre and output allocations separately
Control print origin and centre relative to card features, card seating relative to housing, shaft axis and zero, contact footprint, valid travel and local curve. Allocate angular phase and radial overlap independently. ChipSimple can review and manufacture drawing-defined ceramic thick-film geometry; the throttle assembly owner controls shaft, housing and functional acceptance. Reopen registration after artwork, datum, substrate, mount, shaft, contact, receiver or calibration changes.
Close the remaining implementation and validation risks
Review closed and open throttle ends independently. A positive phase shift can increase margin at one end while consuming it at the other. Confirm that the contact remains on its intended region before each mechanical stop and that stop load does not bend the card or carrier. Software clipping can conceal physical overtravel, so preserve raw controlled-test output beyond the declared interval where safe. Stack stop tolerance, footprint, print boundary and phase in the same angular coordinate. Monitor print-to-card phase optically and shaft-to-card relationship in assembly. Functional voltage at several angles detects combined shift but cannot alone separate centre error from curve shape. Trend fiducial phase, eccentricity components and endpoint output by lot and fixture, with a reference part able to expose nest or camera drift.
Complete the page-specific release closure
When a correction is proposed, evaluate both geometry and process capability. Moving artwork can offset a stable mean phase but does not reduce housing variation, shaft eccentricity or seating scatter. Tightening print registration cannot correct a poorly located shaft. Use the sensitivity stack to assign action to the contributor with evidence. Confirm that revised endpoints preserve conductor overlap, glaze boundaries and contact wear path. Maintain a golden coordinate record that links optical fiducials, assembled angular reference and electrical curve so future tooling or software changes can be compared without rebuilding the datum logic.
At final review, compare nominal alignment, arithmetic tolerance extremes and measured assembly distribution without substituting one for another. Confirm that the drawing defines inspection temperature and seated state. If the shaft datum is inaccessible after assembly, establish a validated transfer feature rather than measuring a nearby cosmetic surface. These steps make subsequent production checks traceable to the original functional coordinate.
Provide the throttle shaft-to-artwork datum chain
Registration review requires the mechanical zero and printed curve in one angular coordinate.
- Card artwork, fiducials, outline and print-registration tolerances.
- Housing, shaft axis, zero feature, stops and seating details.
- Contact footprint, path, local curve slope and connected circuit.
- Phase and overlap allocations, assembly sampling and validation owner.
The drawing-upload form loads as you reach this section.

