On the kart, in the kit: the JOES tire gauge (every pressure number in the ledger), the pyrometer (coming), and the Sniper laser alignment tool (every toe number). Each one is a machine that turns something you can't see into a number. Knowing what's inside is what tells you when the number is lying.
The principle in one sentence
An instrument converts the thing you want into something that moves or glows, and every conversion has a range where it's honest and a range where it isn't.
The tire gauge: a bent tube that straightens
Inside an analog gauge is a Bourdon tube: a flattened metal tube bent into a C. Air pressure inside it tries to make its cross-section round, which makes the C straighten slightly, and a little lever and gear (M01) turn that tiny motion into the needle's sweep.
needle angle ~ pressure (over the design range)"Over the design range" is the catch. The tube straightens in proportion to pressure only in the middle of its travel; near zero and near full scale the gear train's slack and the tube's stiffness make the relationship bend. That is why gauge accuracy is quoted as a percentage of FULL SCALE, not of the reading:
JOES, 0-30 psi, 2.5% of full scale = 0.025 x 30 = 0.75 psiAt 12 psi, 0.75 psi is 6% of the reading. A 0-60 psi gauge of the same grade would be off by 1.5 psi at 12: worse, because the same percentage applies to a bigger scale. This is the whole argument for buying a gauge whose full scale is close to the pressures you run, and for the 0-30 over the 0-60.
Math: percent of what? "2.5%" alone is meaningless; percent of WHAT is the question. Percent of full scale (30 psi) gives 0.75 psi at every reading. Percent of reading would give 0.3 psi at 12 psi. Manufacturers quote full scale because it's the honest description of a mechanism whose error is fixed in size, not proportional. When you read a spec, find the "of."
The digital Intercomp uses a strain gauge (a thin metal foil whose electrical resistance changes when it's stretched by pressure on a diaphragm) and reads to 0.1 psi. Its full-scale rule is the same; its full scale is just bigger and its percentage smaller (0.1%).
The pyrometer: two metals and a voltage
A thermocouple is two wires of different metals joined at the tip. Where two different metals touch, a tiny voltage appears, and the size of that voltage depends on the temperature of the junction: about 41 millionths of a volt per degree C for the common "K" type (typical). A 100 C rise makes 4.1 thousandths of a volt; the meter measures that and prints a temperature.
voltage ~ (temperature at tip) - (temperature at the meter)The second term is why the meter has its own thermometer inside: it measures its own temperature and adds it back. Leave the meter in the sun and its "cold end" is hot, and if its internal thermometer lags, the reading is off. Keep it in the shade.
Probe vs infrared: the needle probe pushes a millimeter into the rubber and reads the temperature under the surface, which is where the tire's working heat lives and which cools slowly. An infrared gun reads the surface only, which cools in seconds off the track and reads low unless you're fast. Same tire, two different numbers, both right about what they measure. Our "within a minute of stopping" rule is about the surface cooling; the probe is more forgiving.
The Sniper: a laser is a straight line
The Sniper's whole trick is that a laser beam is straight for as far as you need. Clamp a laser to a stub axle so the beam is exactly parallel to the axle (the unit is calibrated for this), shine it at a grid on the other side of the kart, and the dot lands where the axle is pointing. If the axle is toed in by a small angle, the dot lands off center by:
offset = angle x distance (angle in radians, small angles)Math: small angles. For small angles, the sideways offset at a distance is just the angle (in radians) times the distance. A radian is the angle where the offset equals the distance; there are 57.3 degrees in one. So 1 mm of offset at 1200 mm is 1/1200 = 0.00083 rad = 0.048 degrees. Toe angles on a kart are a few hundredths of a degree per side; the Sniper turns them into millimeters you can see by using a long distance as a lever (M01 again, in disguise).
The manual says each grid line is 2 mm of toe per side, the mid-line dot 1 mm. That "2 mm" is the offset that the same angle would produce across the tire's own diameter, so the grid is pre-scaled for a kart tire. The unit can't be wrong about straightness; it can be wrong about parallel (calibration), level (the kart on an uneven floor tips both units the same way, which shows as camber not toe), and the steering not being centered, which puts toe-in on one side and toe-out on the other. Read both grids; the sum is toe, the difference is steering off-center.
On our kart, in numbers
The right-rear change on 9/19. 12 to 11.5 psi on a 0.75 psi gauge: the change was smaller than the instrument's rated error. Lesson 12 says what to do about that (repeat readings; the gauge is more repeatable than it is accurate).
Toe on 9/19. 1 mm per side on the Sniper = 2 mm total. On a 10 in (254 mm) tire, 1 mm of offset is 1/254 rad = 0.23 degrees per side. The tie rods were adjusted to make that number; nobody could have seen 0.23 degrees by eye.
Two things to notice
- Every instrument has a "cold end." The gauge's is atmospheric pressure (it reads gauge pressure, lesson 1 Q5); the thermocouple's is the meter's own temperature; the Sniper's is the calibration of beam-to-axle. The number you read is a difference from that reference, and when the reference drifts, so does every reading.
- Repeatability is a property of the instrument; accuracy is a property of the instrument AND the calibration. A gauge that reads 0.5 low every time is perfectly good for comparing this session to the last. It's only wrong the day you compare with someone else's gauge.
Problems
1. Percent of what. A 0-100 psi digital gauge rated 1% of full scale, and a 0-30 analog rated 2.5% of full scale: which is more accurate at 12 psi, and by how many psi?
2. The cold end. The pyrometer meter is left on the black kart stand in the sun and its body warms to 45 C while its internal thermometer still thinks it's 30 C. You measure a tire that is really 60 C. Does the reading come out high or low, and by how much? (The voltage measures tip minus meter; the meter adds back what it THINKS its own temperature is.)
3. Sniper geometry. The units clamp to the two stub axles, so the laser-to-grid distance is our front track width, about 1200 mm (typical; measure ours) - and the grid's "1 mm per side" markings were calibrated for a distance in that range. If a kart with a 1500 mm front track used the same grid, what would 1 mm on the grid actually mean in tire-scale mm? (Offset scales with distance.) Is that a reason to re-check the grid's calibration whenever the front track changes, for example after moving spacers (lesson 8)?
Go look: take the gauge apart in your head by reading it at zero (does the needle rest on the pin, or above it?), then press the bleed button on a tire and watch the needle come down: does it move smoothly or in steps? Open the Sniper's manual to the calibration page. Hold the pyrometer meter in your hand for a minute and watch its reading drift with your body heat: that's the cold end.