I was going through the documentation of TPMS sensor FXTH87 and I am trying to understand better how the low frequency receiver works. Here is a scheeme of it :

enter image description here

I understand that the incomming signal is modulated, and that we have to get rid of the carrier and so to get just the envelope, we use a rectifier. But why are they 4 rectifiers?

As I understand it, the incoming signal goes trough Amp1 and then there is a buffer (why? to keep the signal intact?), then this signal from one side is rectified and on the other side it goes into a second amplifier (Amp2) and then a second buffer (why?),...

The only way I can understand it is that the signal is rectified 3 different times and summed and averaged to minimize the error. This seems weird since the signal rectified by the second rectifier seems to be signal*Amp1*Amp2 and the signal rectified by the 3rd rectifier seems to be signal*Amp1*Amp2*Amp3.

Also why is the signal clamped at the beginning? Is it this operation that they mean(see image below)?

enter image description here

As for what is within the yellow box I am quite lost. In the data-sheet it says you can select 2 threshold for the detection of the carrier: a high sensitivity and a low sensitivity. It seems for the first sensitivity, the incoming signal in the yellow box will be the one after buff1 and for the other sensitivity, it will be the signal after the buff2. I guess if we select the low sensitivity, they will amplify the signal just once and take the output right at buff1. If we select high sensitivity, it means we want to be able to detect weaker signals so we amplify the signal once more before going into the carrier detector, is that the idea? What is the small arrow that looks like a current source?

Also, why is the entry of the Logic Block 2 is 129KHz? If something, it should be 125KHz in my opinion because this is the frequency of the carrier. Why is this frequency needed to decode the data? Are we somehow mixing the 2 frequencies?

One last thing : 125KHz is a low frequency, it has a wave-length of 2.4km, so as I understand it, we are not transmitting electro-magnetic wave here but we are just modulating the magnetic field. Is that correct? The receiver has some tune circuit that resonate with the magnetic field produced by the tranceiver and this is how we get the data?

Thanks for your help and please bear with me, I don't know much about electronics, I just like to learn it.

  • \$\begingroup\$ Pretty fancy tire pressure transponder. ! From what I can tell at a glance, it appears they want to conserve energy by not hunting with automatic gain control and just cumulate carrier modulation after successive stages in order to avoid AGC and also achieve a 56 dB min dynamic range. almost a thousand to 1. \$\endgroup\$ Commented Jul 11, 2017 at 2:25

2 Answers 2


The input clamp protects the downstream amplifiers, in the case when the keyfob is too close to the transmitter.

Dynamic range of these keyfob 125KHz data links is 100,000X in voltage (100dB).

Regarding the rectifiers: diodes can be used (in which case you likely need to buffer, to provide the diode current); other active (uses power) circuits use transistors to "rectify" and can be much more sensitive......but need current to operate.

Why FOUR amplify/rectify behaviors? to implement an accurate RSSI --- relative signal strength indicator, even as the signal varies 10,000:1 or 1,000:1.

No idea about the 129KHz.


Each Mfg has their own protocol for e.g. Rx 125 kHz carrier (±1%) over a range between 80~210 kHz ( LF reception in tire TPMS decoder and dual-option VHF transmission from the TPMS.

The TPMS Rx uses a 125kHz LF AM-BiΦ channel with a series LC tank circuit and expects a specific burst frame rate in the range ~40~85Hz, ( I suspect to avoid regional AC Line interference and make unique from other car brands.)

The MCU is off until Carrier Detect (CD) is active in Block 1 and data is ready in Block 2. THe carrier is sampled @ 1KHz, fast enough to detect the burst rate of the preamble and checks for correct frequency and sufficient amplitude in the lower part of your figure.

THis is done by selecting either at the output after 1st of 3 gain amps or the 2nd for higher sensitivity, while all 4 stages are fed into 4 diodes of the "discriminator". It uses a summation Op Amp to provide the maximum swing of modulation over a wide dynamic range shown in datasheet as 17.11 LFR Characteristics #1200 Dynamic Range= 56dB min starting input at 3.5mVpp worst case in the 1% range and 14mVpp max in the wide carrier range.

This output feeds the Slicer and compares it to the average DC voltage to make an AC coupled limiter, faster than AGC could operate and is more immune to reactions of AGC to pulse noise. This is critical to obtain optimum symmetry of Bi-Φ clock and data signal which is separated in Block 2.

Block 2 might use 3.096Kbps data rate with data indicated by position of the transition at 1/T or 1T while the other phase (180deg) is an expected polarity of clock and standard means for Bi-Φ SERDES recovery is expected depending on which protocol is used (Mark, Space or Invert).

From this protocol data/clock is separated or self- clocked to detect, knowing the clock interval if using a Rx Data of 3.096 kbps.

Block 2 Data recovery includes using what appears to be a 33x bit clock (129kHz) to generate the various internal signals that detect data patterns of start burst sync (0000...), start byte, Byte clock, word counter and End of frame byte (illegal pattern)

This is a flexible standard way of detecting a wide variety of protocols for Async Bi-Φ from my similar design experience.

THe outputs in Block 2, not shown, would be many but must detect a valid message, correct address then enable MCU power ( wake-up). It would have preset inputs from the MCU to define protocol parameters to match with incoming signal.

Block 1 system power is in the 5uA range, and when Block 2 enable around 12uA .

All clocks are derived from a 30ppm 26 MHz Xtal. and power from a 2.7V battery with an EOL ESR of 60 Ω.

It has sensors for Temperature, 100 kPa~ 1500 kPa tire pressure transducer and optional XZ or Z magnetometer sensing for speed and direction.

It's a pretty fancy and flexible FOB tire pressure transponder. !! for about < $10 OEM is my guess.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.