The short version
- An infrared thermometer will read a hub at 80 C as though it were 35 C if the surface is bright metal.
- That is an emissivity error, it is enormous, and almost nobody accounts for it.
- Point one from three feet away and you are averaging a three inch circle, not reading the hub.
- A chassis ear is only useful if one channel is a control on a corner you trust.
Wheel bearing diagnostic tools are cheap compared with the mistake they prevent. Replacing a healthy bearing costs a few hundred dollars and does not fix the noise.
But two of these tools mislead badly when used casually. The reasons are physical rather than a matter of care.
The four tools
| Tool | Rough cost | What it answers |
|---|---|---|
| Chassis ear | $60 to $120 | Which corner, at road speed |
| Mechanic’s stethoscope | $10 to $30 | Which corner, on a lift |
| Infrared thermometer | $25 to $80 | Is one hub running hot |
| Dial indicator and base | $30 to $70 | How much end play, in numbers |
Each answers a different question. Used in that order, wheel bearing diagnostic tools narrow a noise from four corners to one, then confirm it with a number.
How to run the tests themselves is on the diagnosis page. This page is about the instruments and how they lie.
The infrared thermometer trap
This is the big one, and it explains a lot of failed diagnoses.
An infrared thermometer does not measure temperature. It measures radiated energy and converts it, assuming the surface radiates like a standard one.
How well a surface radiates is its emissivity. Most cheap thermometers are fixed at about 0.95, which suits paint, rubber and rust.
A freshly machined or polished steel face has an emissivity around 0.15. It is a poor radiator and a good reflector. A thermometer expecting 0.95 therefore sees far less energy than it should.
The size of the resulting error is startling. We calculated it for a hub at a genuinely worrying 80 degrees Celsius, in a 20 degree garage.
| Surface | Emissivity | Reading shown | Error |
|---|---|---|---|
| Bright machined hub face | 0.15 | 35 C | 45 degrees low |
| Lightly scaled steel | 0.30 | 46 C | 34 degrees low |
| Rusted or cast iron | 0.90 | 80 C | Essentially correct |
Aim at a shiny hub face and a bearing hot enough to destroy itself shows as barely above room temperature. The tool is working correctly. The assumption behind it is wrong.
How to get a reading you can trust
Three habits fix it entirely, and none of them cost anything.
- Read a dull surface. The cast knuckle or a rusty part of the hub, never a machined face or a shiny rotor hat.
- Read the same feature on both sides. Comparison beats absolute numbers, and only works if both surfaces share a finish.
- Stick masking tape on each side and aim at that. Tape sits near 0.95 emissivity, which is exactly what the tool assumes.
The tape trick is what thermographers do professionally. It costs nothing and removes the problem in one step.
Distance to spot
The second thing that quietly ruins infrared readings.
Every thermometer has a distance to spot ratio, often 8 to 1 or 12 to 1. It describes how the measured circle grows as you step back.
| Distance from the hub | Spot at 8:1 | Spot at 12:1 |
|---|---|---|
| 6 inches | 0.8 inches | 0.5 inches |
| 12 inches | 1.5 inches | 1.0 inches |
| 24 inches | 3.0 inches | 2.0 inches |
| 36 inches | 4.5 inches | 3.0 inches |
Stand back three feet with an 8 to 1 tool and you average a four and a half inch circle. That takes in the rotor, the caliper, the knuckle and open air.
Get within about six inches, and hold the same distance on both sides. The laser dot shows aim, not spot size. That is the detail most people miss.
A dragging caliper heats a corner far more than a bad bearing does. Take readings after steady driving with little braking, not after a stop and go trip.
A hot rotor with a cool hub is a brake problem. A hot hub with a cool rotor points at the bearing.
Getting value from a chassis ear
The most useful tool here, and the one most often used badly.
It is a set of clip-on microphones wired to a receiver you hold while driving. Switch between channels and listen to each corner in turn.
The mistake is clipping every microphone near the corner you already suspect. That only tells you the noise is loud there, which you knew.
Put one microphone on the suspect corner and one at the matching point opposite. Compared directly, a failing bearing is obviously louder.
Use the remaining channels on the other axle. You want a ranking, not one loud reading with nothing to judge it against.
Clip to something solid and structural. A microphone on a heat shield or a plastic liner picks up rattles rather than the bearing.
Route the wires clear of anything hot or turning, and check them before every drive. This is the tool where carelessness costs a wiring loom.
Stethoscope or chassis ear
They answer the same question in different conditions.
A stethoscope is cheap and works with the car raised and the wheel spun by hand. It finds a rough or dry bearing quickly.
Its limit is load. Many bearings are quiet unloaded and only sing under the car’s weight. A spun wheel can sound perfect on a bearing that howls at 50 mph.
A chassis ear works at road speed under real load, so it catches what the stethoscope misses. That is why shops own both.
The dial indicator
The only tool here that produces a number rather than an impression.
Mount the magnetic base on something that does not move with the part you are measuring. The knuckle or the strut, never the rotor or the hub itself.
Set the tip against the hub face, zero it, then rock the wheel and read the movement. Typical limits are a few thousandths of an inch.
The value is that it settles arguments. Play you can feel by hand is already substantial, and the play and vibration page covers the limits.
What none of these can do
Worth saying plainly, because it is where diagnosis usually goes wrong.
No tool here separates a bearing from cupped tires reliably. Tire noise is the most common misdiagnosis and sounds nearly identical.
The test for that is free. Rotate the tires and drive it. If the noise moves with them it was never the bearing, and the impostors page covers the rest.
What to buy
A stethoscope and an infrared thermometer are cheap enough to own outright and useful across many jobs.
A chassis ear is the one worth borrowing or splitting with someone. It gets used rarely and it is the most expensive item here.
A dial indicator earns its place if you already do mechanical work. How the wheel bearing diagnostic tools fit the wider budget is on the tools overview.
Turning a felt looseness into a measured figure is covered under dial indicators.
Common questions
Which wheel bearing diagnostic tools are worth owning?
A mechanic’s stethoscope and an infrared thermometer are inexpensive and useful elsewhere. A chassis ear is better borrowed, since it gets used rarely.
Why does my infrared thermometer show a low temperature?
Almost certainly emissivity. A bright machined surface at 80 C reads about 35 C on a thermometer fixed at 0.95. Aim at masking tape or dull cast metal instead.
How close should I hold an infrared thermometer?
About six inches, and the same distance on both sides. At three feet an 8 to 1 tool averages a four and a half inch circle including the rotor and surrounding air.
How hot is too hot for a wheel hub?
Absolute figures vary with driving and ambient temperature, so compare sides instead. More than about 20 degrees Celsius between the same points is worth investigating.
Where do I clip the chassis ear microphones?
One on the suspect corner and one at the matching point on the opposite side, as a control. Clip to solid structure rather than shields or plastic liners.
Can a stethoscope find a bad bearing?
Sometimes. It works well on a rough or dry bearing but misses those that only make noise under load, which is a large share.
Do I need tools to diagnose this at all?
No. Load transfer on a long curve, and a tire rotation to rule those out, cost nothing and settle most cases. The symptoms overview starts there.