How to Approve an RC Vehicle Sample: Steering Centering Log

Sourcing rules that decide the approval

• Define neutral against a sealed golden sample held on the QC shelf, not against a photo or an email. Buyers who approve only an image pay for it on the second lot.
• Record tool number and revision on every logged unit before observation. A trade-fair or booth sample is often a hero unit, and fair lighting is not a production standard.
• Split the log into confirmed facts, assumptions, and unresolved checks. An observed deviation is a fact; "the servo is weak" is an assumption until you disconnect the linkage and re-test.
• Put the safety gate first. On button-cell versions, screw closure, tool access, and warning wording are checked before any steering function test — a compartment that opens with fingernails does not ship.
• Never write an AQL, tolerance, or sample size into your own approval logic. Those numbers come from the agreed inspection plan on the PO, and every one you do not have yet becomes a question to the supplier.

What goes in the log header before you observe anything?

The header identifies the unit so the reading means something later. Without it, a centering note is an opinion attached to an anonymous piece of plastic.

Minimum header fields: SKU and variant, tool number, tool revision, the sealed golden sample ID this unit is compared against, the date and time of observation, the observer, and the power source used (battery type and state of charge). Add the firmware or receiver/transmitter pairing if the model has one, because a paired transmitter with a trim setting left off-centre will produce a "deviation" that belongs to the test bench, not the tool.

One more header line that buyers skip: unit provenance. Was the sample pulled from a trial shot, a pre-production run, or a finished production lot? A pre-production unit answers tooling questions; it does not answer process-capability questions. Mark it and the log stays honest.

Which observation conditions make a centering reading valid?

A centering reading is only valid if the vehicle is free to move and the steering is not fighting anything.

Test on a flat, hard, level surface with adequate clearance. Perform the steering check while the vehicle is driving forward at low speed, not while stationary — on tracked and chain-driven RC models the steering trim typically has no effect when the machine is not moving, so a stationary reading tells you almost nothing about centering behaviour.

Center the transmitter trim and confirm dual-rate or endpoint settings are at their agreed values before you start. If the wheels move a little but never reach full lock, that is a trim or endpoint condition, not a centering defect, and it belongs in a different column.

Record the surface, the speed band, and whether the body shell and battery were fitted. A reading taken without the shell is not the same reading as one taken at shipping weight.

How do you record neutral position and deviation direction?

Describe the observed neutral position in neutral language: does the vehicle track straight on a level surface without continuous steering correction, and after a small steering input does it return toward centre on its own?

Then record deviation direction as left, right, or none — and the magnitude as a band you define with the supplier, not a number you invent. If no tolerance has been agreed, write "tolerance not yet agreed" and raise it as a supplier question. Do not convert a feel into a millimetre.

Log the failure modes as separate observation fields rather than one verdict: sluggish or slow steering response, no return to centre, abnormal noise such as squeaking or clicking during steering input, heavy or binding steering, and visible leakage where the design includes fluid. Each field gets its own fact/assumption/unresolved tag.

Finally, note whether the unit is a hero unit. A hero sample can pass every observation and still not represent the tool.

Step | What to check | Red flag

Where does the safety gate sit relative to the steering test?

Ahead of it. CarToyFactory are a compartment and warning problem before they are a function problem — screw closure, tool access, and warning wording are checked before anyone discusses how the steering behaves.

This ordering prevents a common approval error: a unit that steers beautifully gets waved through, and the compartment design is only examined after tooling is committed. Once the tool is cut, a compartment fix is expensive.

Log the safety gate as its own pass/fail with its own evidence. If the design relies on a screw closure, confirm the closure and the tool needed to open it. If the warning wording is part of the approved artwork, confirm the wording on the unit, not on the artwork file.

How should a failed centering check be handled and released?

Isolate the lot, record the defect class, and release only after rework. A nonconformance report should state the defect, the lot, the quantity, containment, disposition, and the responsible decision — not just a pass/fail note.

Isolate first, diagnose second. Disconnect the linkage to separate a servo or electronics issue from a chassis or geometry issue, then re-observe. Many centering complaints resolve into a linkage or trim finding once the two systems are separated.

Release criteria come from the agreed inspection plan on the PO. If you want a numeric acceptance threshold, it has to be written there — a catalogue sentence is not an inspection plan. Keep supplier records for materials, production, inspection, change control, and testing long enough to support your market and recall obligations.

And when the fix changes a material — resin grade, plasticizer, pigment, coating, or adhesive — treat it as a new chemical review, because those changes can move your REACH conclusions.

Worked example (illustrative, not a real shipment)

A private-label buyer reviews a 1:18-scale RC vehicle sample for two destinations: retail cartons for one market and plain cartons for the other, roughly 2,400 units in the planned first order split across the two pack configurations. One unit arrives, plus a phone photo of the steering sitting straight.

Header: SKU logged, tool number recorded, revision noted, unit marked pre-production, transmitter trim centred, battery freshly charged. Safety gate: non-button-cell build, battery compartment closure checked and passed. Bench: linkage moves freely, gaiters intact.

Low-speed forward test on a flat hard floor: the vehicle drifts right and needs continuous left correction; it does not return to centre after a small left input. Recorded as: observed neutral position — right of centre; deviation direction — right; magnitude — tolerance not yet agreed; noise — none; hero unit — no.

The assumption written next to it was "servo weak." Disconnecting the linkage and re-testing showed free movement at the rack, so the servo was not the fault. The unresolved checks list went to the supplier: agreed neutral-position definition and tolerance, which sealed golden sample and revision governs this SKU, the sample size and inspection plan, the AQL (if any) on the PO, and how a failed lot is isolated and released.

Where this holds the shipment: the sample cannot be approved as production-representative while the neutral-position tolerance is undefined, because the second lot will be measured against nothing. Retail and plain carton variants do not change the steering check, but they do change the inspection plan, so both configurations need their own line in the agreed plan.

What to ask suppliers: RFQ checklist

Ask these in writing and keep the answers with the sample. Every number you do not have yet is a question, not a placeholder.

1. What is the agreed neutral-position definition and tolerance, and where is it written — PO, drawing, or inspection plan? 2. Which sealed golden sample (ID, tool number, revision) is the reference for this SKU? 3. What sample size and inspection plan apply, and what AQL, if any, is stated on the PO? 4. How is a failed lot isolated, how is the defect class recorded, and what releases rework? 5. Is the submitted unit a hero or booth sample, or a production-representative unit? 6. For button-cell versions, what is the approved compartment design — screw closure, tool access, and warning wording? 7. How long are material, production, inspection, change-control, and test records retained? 8. If a resin, plasticizer, pigment, coating, or adhesive changes, what triggers a new chemical review? 9. What is the agreed surface, speed band, and configuration for the centering test?

FAQ

Can I approve an RC sample from a photo of the steering centred?

No. A photo cannot define neutral position or show return-to-centre behaviour, and buyers who approve only an image typically discover the difference on the second lot. Approve against a sealed golden sample with the tool number and revision recorded.

Should I test steering centering with the vehicle stationary or moving?

Moving, at low speed on a flat hard surface. On tracked and chain-driven models the steering trim generally has no effect when the machine is stationary, so a stationary reading does not describe centering behaviour.

What tolerance should I put on neutral position?

There is no universal figure — the tolerance must be agreed with the supplier and written into the PO or inspection plan. Until it is agreed, record the deviation direction and mark magnitude as "tolerance not yet agreed" rather than inventing a number.

How do I know whether a centering deviation is the servo or the chassis?

Disconnect the linkage and re-test. If the rack, knuckles, and linkage move freely, the fault is more likely in the servo, receiver, or settings; if they bind, the chassis or geometry is the stronger suspect. Log both results separately.

Does a button-cell compartment check belong in a steering centering review?

Yes, and it comes first. Screw closure, tool access, and warning wording are checked before any steering function test, because a compartment that opens with fingernails is a non-starter regardless of how well the vehicle steers.

What should a nonconformance report contain for a failed centering sample?

The defect, the lot, the quantity, containment, disposition, and the responsible decision — not a pass/fail note alone. The lot is isolated, the defect class recorded, and release happens only after rework against the agreed plan.

Sources

• Steering Control
• Steering System
• 'Bright Idea' For Short Detection MOTOR
• AS MOTOR AS 1000 OVIS RC - Maintenance: Functional Checks; RC and Steering System Checks

Build the log before the next sample lands

Take the header fields, the eight-step sequence, and the nine supplier questions above and put them into one sheet your team uses on every RC sample. Then send the questions with the RFQ so the answers arrive with the unit, not after the tool is cut.

If you want a second pair of eyes on a centering finding — or a golden sample sealed and logged against your SKU — send us the sample details and the observation record, and we will tell you which readings are facts and which still need a check.