Failure description: As the speed of the rear wheel reached 80km/h the speedometer would suddenly display 0 km/h.
Probing the signal from the rear hub sensor with an oscilloscope showed a signal that increased in frequency as the speed of the rear wheel increased. There was no sudden dropout at 80km/h (~120Hz). We did note that the signal amplitude was only around 300mV p-p.
Tracing the signal up to the speedometer head, we noted that the sensor fed into the "Moto Meter" board, via the BE pin. Measuring at this location confirmed the same 300mVp-p signals, which would increase in frequency as the speed of the wheel increased. No cut outs at this point.
Probing the Output of the Moto Meter board (BA pin), saw a very large scale clean signal (at least 5V p-p, didn't actually note the exact voltage). However, at 80km/h (~120Hz) this signal started to stutter and then completely quit. It would return when the speed decreased below 80km/h.
The Moto Meter board utilizes a dual op-amp LM2904N chip. After a bit of probing and measurement, view the approximate schematic in the next post.
The first stage is configured as a comparator. Every time the BE input signal swings greater than 0V, the first op amp outputs a high amplitude signal. Note however, that the input signal is first passed through a Low Pass Filter (LPF) before reaching the + input of the op amp. (The LPF is the series resistor and capacitor to ground.) A capacitor acts as a short circuit to ground as the frequency increases. The suspicion here is that as the frequency increases, the LPF is filtering too much of our already small signal from the rear hub away, and at some point, the comparator no longer sees it. The 5k / 4.7uF combination makes for a 6Hz cut-off frequency, which seems very low considering that the hub sensor is sending a frequency around 150Hz at 100km/h.
Suspecting that perhaps the hub sensor normally outputs more than 300mVp-p, or perhaps the 4.7uF cap is starting to go bad. So, removed the cap and replaced it with a 0.5uF ceramic capacitor (what we had laying around). See photo in next post.
Bench tested with a home made signal generator (using a ESP32 micropython board, that's a whole other story!) injected 300mVp-p on the bench at various frequencies. Recreated the original failure with the original cap (LPF output dropped to 16mVp-p at 300Hz). Confirmed that with the new capacitor value the LPF output only dropped to 45mVp-p at 300Hz. More importantly, the output of the Moto Board on the BA pin remained strong and active the entire time. 300Hz is around 200km/h, so this seemed like more than enough operation range as a fix.
Gave the circuit and speedometer back to Suzuki12. And he confirmed that it fixed the problem! So, this was either a bad cap, or a beginning-to-fail rear hub sensor that we are squeezing a little more life out of with the lower capacitor value
