What a repeater actually does
A TPMS repeater is a second pair of ears. It listens on the same 433.92 MHz channel as the main receiver, decodes the frames sent by the wheel sensors, and transmits those frames again so that the main receiver can hear wheels it could not hear directly.
It does not measure pressure. It does not hold tire data. It has no screen and usually no storage. Its entire job is to take a radio frame that arrived at one place and make it arrive at another place, without changing what the frame means.
That last point is the one that makes repeaters practical. Because a repeater passes the original sensor identity and payload through, the main receiver treats a re-transmitted frame exactly like a direct one. Your controller does not know, and does not need to know, how many hops a reading took. No software change is required on the machine side when a repeater is added.
Why the radio link needs help at all
At 433.92 MHz, radio behaves much like light. It travels well through open air and poorly through dense material. A mining truck, a port crane or a beam carrier is mostly dense material: steel plate, axle housings, hydraulic runs, containers and sometimes cargo.
So the number that matters is not the datasheet range. A sensor and receiver pair may be specified for 100 metres in open air, and on a 20-axle carrier there is very little open air. The chassis sits directly in the path.
You could solve this by making the sensors transmit harder or more often. That is a bad trade, because the sensor is battery powered and the battery is expected to last five to six years for an internal unit, or at least two years for a replaceable CR1632. A repeater, by contrast, is wired to the machine at DC 10-30 V and can listen continuously without any battery constraint.
Getting the radio link to close with a powered relay is almost always cheaper and more reliable than pushing more energy through a coin cell.
The relay, step by step
The sequence inside a repeater is short.
1. A wheel sensor transmits its frame: sensor identity, pressure, temperature, status flags.
2. The repeater's receiver front end picks up the frame, provided the frame is strong enough where the repeater is mounted.
3. The repeater validates the frame, so noise and partial collisions are discarded rather than forwarded.
4. The repeater re-transmits the frame on the same channel, on its own timing so that it does not collide with the original transmission or with another repeater.
5. The main receiver hears both the direct copy, if there is one, and the relayed copy. It de-duplicates by sensor identity, so a wheel heard twice is still one wheel with one pressure value.
That de-duplication is normal receiver behaviour and it is why repeaters do not inflate the wheel count or confuse the position mapping.
What it is not
It is not a signal amplifier in the simple sense. It does not capture the sensor's weak transmission and radiate it louder from the same spot. It decodes and re-transmits, which is why a repeater can be placed where the sensor cannot reach at all, as long as the repeater itself can hear the sensor.
It is not a receiver. It has no display, no CAN output and no need for a protocol document, because it never talks to your controller.
It is not a data logger. If you need historical tire data, that belongs in the receiver's output or in the cloud platform, not in a repeater.
And it is not a substitute for planning. Adding repeaters at random usually produces a system that works in the workshop and misbehaves in service, because the obstruction that matters only appears when the machine is loaded or turned.
Latency, cycle time and the machine's own bus
A repeater adds one hop of delay, measured in milliseconds. Against a tire data cycle of 500 ms, that is invisible. The alarm flags travel inside the same frames, so a rapid-leak condition arrives at the controller with the same freshness whether it came directly or through a relay.
This is also why a repeater does not change your integration work. The CAN frame layout, the register map on a serial output, and the alarm semantics are all identical. The repeater is a radio-layer device, and the radio layer is deliberately the part that stays invisible.
How many repeaters a machine needs
There is no formula. Anyone who quotes "one repeater per N axles" without looking at the machine is guessing, and the guess is usually wrong in both directions.
Repeater count follows geometry: where steel mass sits between the front and the rear, how the wheel positions group into radio zones, and where a mounting point exists that has a clear path in the direction data must travel. That is why the count is quoted per machine rather than per product.
Our own production reference is specific. The 160-wheel beam carrier we delivered is configured with two receivers, twelve repeaters and 160 sensors, and ten of those units are in service. The largest single machine we have instrumented carries 224 wheels across 28 axles. Neither configuration used one repeater per axle, and neither was derived from a rule of thumb.
The multi-axle TPMS hub explains how the zones are worked out, and the beam carrier installation page shows what the delivered configuration looks like on the machine.
What bad repeater planning looks like
The failure is rarely dramatic. It shows up as intermittent data: a wheel that reports most days and disappears occasionally, usually when the machine is parked at a different angle, loaded differently, or when a container happens to sit over an antenna.
Because it is intermittent, it gets misdiagnosed. The sensor gets swapped, then the receiver, and the problem returns. By the time the machine is in service, every return visit is expensive and the customer's confidence is already falling.
The cure is procedural rather than clever: record link quality position by position during commissioning, keep that record, and treat any later gap as evidence about the radio path rather than as a mysterious component failure. A written record turns "the TPMS is unreliable" into "position 47 needs its repeater moved 600 mm", which is a repair job instead of an argument.
Checking a repeater in the field
Five checks cover most field cases:
1. Power. The repeater runs on DC 10-30 V from the machine. A loose connector on a machine that vibrates is a common cause, and the repeater is IP68 while the connector may not be.
2. Mounting and shadowing. Look at what is between the repeater and the zone it serves. Metal added after installation, a toolbox or a mud guard, can recreate the original problem.
3. Configuration. Which wheel zones the repeater is meant to relay is a configuration item; an unconfigured replacement behaves like a dead unit.
4. Swap test. Move a known-good repeater to the suspect position. If the affected wheels recover, the radio path is confirmed as the issue.
5. Link quality readout. Per-position signal reporting during commissioning is the fastest way to see the difference, and it is the reason we record it.
None of these requires special test equipment, but all of them require that the original configuration was documented.
Where we fit
GZVIA designs and builds the sensor, receiver, repeater, display and cloud platform in house in Shenzhen. The same hardware family supports receivers and repeaters from one part number, runs from DC 10-30 V, is sealed to IP68, and uses 433.92 MHz license-free radio, so no licence is needed to operate the system. Pressure accuracy is ±7 kPa and the default CAN data cycle is 500 ms.
If you send the wheel count, the axle layout and how far the rear positions sit from the cab, we will propose a receiver and repeater plan for that machine rather than a generic ratio. Send us the machine layout.