This tutorial attempts to use the Betaflight Blackbox curve to determine whether the gyroscope is abnormal, but the release notes contain a critical errata: the 0.012 ms in the video is actually 12 ms, and the final acceleration result should be reduced by 1000 times. Calculated based on a change of about 170°/s in the slice, a radius of 0.08 m, and 12 ms, the tangential acceleration is about 20 m/s², about 2 g, not 2000 g; therefore this value alone does not prove that the gyroscope is damaged.
First find the abnormal vibration fragments from Blackbox
If the aircraft retains a certain vibration no matter how the PID is adjusted, you can record a Blackbox log of a safe flight and locate the time period when the vibration is obvious in the analysis tool. The video simplifies the view into the gyro and motor outputs on the same axis, temporarily hides the PID, FeedForward and other curves, and then uses markers to mark the starting and ending points of the gyroscope's rapid changes.
When the gyro reports a fast error, the PID commands the motor to follow, so it's not surprising that the motor output changes drastically at the same time as the gyro. If the signal is noise, the controller may continuously accelerate and decelerate without producing effective attitude changes, and the energy will eventually turn into motor heat; however, this correlation only shows that the flight controller is responding to the signal and cannot independently distinguish between bad gyroscopes, mechanical vibrations, or electrical interference.
Correctly handle time units and acceleration estimation
The video uses the angular velocity change to estimate the tangential acceleration at the end of the arm: first divide the angular velocity difference by 360, then multiply it by 2π and the radius 0.08 m to get the tangential velocity change; finally divide it by the change time. The key is that the unit of 0.012 of the marker is seconds, which is 12 ms, not 0.012 ms.
Substituting approximately 170°/s, the tangential velocity changes by approximately 0.24 m/s; divided by 0.012 s, the result is approximately 20 m/s², or approximately 2 g. This order of magnitude is not obviously impossible in flight vibrations, so the inference from the original video of "about 2000 g, so the gyro is not telling the truth" does not hold. Unit inspection must precede hardware diagnostics.
How to determine the source of the problem more reliably
First check the most common mechanical sources: blade damage or incorrect installation, motor shaft and bearings, loose screws, frame cracks, wire knocking on the flight control, flight control soft installation that is too loose or too tight. Then compare the logs under different throttles, different shafts and static conditions with the propeller removed; if the peak value moves with the motor speed, it is more like mechanical or power system noise.
Recheck the electrical and sensor environment, including power supply noise, grounding, distance between ESC and high current lines, flight control design and gyroscope batch. Make sure filtering and log settings are consistent to avoid misinterpreting other axes, interpolations, or scaling as anomalies. Only after troubleshooting installation and power problems, the same flight controller still produces unexplained jumps when stationary or in multiple configurations, it will be more convincing to replace the flight controller with a known good one for A/B comparison.
Avoid having PID chase false signals
Before the root cause is clear, do not simply increase P or D to suppress the vibration. If gyro noise is amplified by D-term, the motor may heat up quickly or even run out of control. After each test flight, check the temperature after landing. If there is sustained high-frequency vibration, squealing or abnormal heat, the flight should be stopped immediately. All propellers are dismantled when analyzing and replacing hardware.
Operation steps
- Record reproducible Blackbox logs
Capture vibrations in safe conditions, noting throttle, motion, filtering, firmware and airframe status.
- Only display the gyro and motor key curves
First select the same axis and hide irrelevant traces, and use markers to measure the amplitude and time of abnormal changes.
- Correctly read 0.012 as 12 ms
Use seconds to substitute the formula uniformly; the example in the film is about 20 m/s², about 2 g, not 2000 g.
- Troubleshooting machinery and noise that changes with rotational speed
Check the propellers, motors, frames, screws, wires and flight control installation, and compare the frequency characteristics under different throttles.
- Check electrical environment and sensors
Check the power supply, grounding, wiring harness, flight control design and log scaling, and consider replacing the flight control after eliminating interference.
- A/B comparison using known good hardware
Keep other conditions unchanged and replace the flight controller to confirm whether the abnormality changes with the sensor hardware.
FAQ
Is 0.012 in the video milliseconds or seconds?
The publisher has corrected it to 12 ms, which is 0.012 s; the original video regarded it as 0.012 ms, causing an error of 1000 times.
Does the correction of about 2 g prove that the gyroscope is broken?
No. About 2 g is not obviously impossible in flight and vibration environments and must be combined with spectrum, mechanical inspection, electrical inspection and replacement comparison.
What does it mean when the gyro and motor curves jump randomly at the same time?
It means that the controller is responding to the gyroscope signal, but the source of the signal may be real motion, mechanical vibration, electrical noise, or sensor problems. It cannot be determined based on correlation alone.
Why can’t D be directly increased to suppress vibration?
D amplifies high-frequency noise, which may cause motor heating and more severe oscillation. You should first find the root cause and monitor the temperature.
Full timeline transcript
Transcripts are arranged according to video time, making it easy to quickly locate the explanation content. Transcript language: Simplified Chinese.
Hello hello everyone, this is SquareRank, and it’s time for our Betaflight trivia again. Today we come to pay homage to this gyroscope problem. If you find during the challenge that no matter how you adjust these parameters, the aircraft always has some vibrations that cannot be eliminated no matter what. At this time you should wonder if there is something wrong with your gyroscope. So how do we check whether there is any problem with this gyroscope?
At this time, you can record a flight log and then use the medium setter to open this log. For example, here I have a log that I have adjusted the aircraft. We open it and then extend it to a place where the vibration is relatively large. For example, the vibration here is relatively large. Then in the trace, we select this debarcdebarcplastictrace. What can we see? One is
You can see the Tolo instrument, and there is another one where you can see the motor, and then this person is all in Roll. We remove these unnecessary signals, such as this PID and this FeedForward, we don't need to look at them. What do we only need to look at at this time? We only need to look at the Toro instrument. Now we can leave this motor alone. You see, when the Toro instrument vibrates, the motor will also go crazy.
So how do we check whether there is any problem with the Gyro 1 signal? We are going to use this marker tool here. We first set a section of Gyro 1, this place is a little bit, we press m, and then we see how much it is here, is it -80 degrees here, right, and then how many degrees has it changed here, is it positive 90 degrees, right? It is about positive 90 degrees, that is to say, what is the time gradually? It is 0.012 milliseconds.
That is to say, it changes from an angular velocity of -80 degrees per second to an angular velocity of -90 degrees per second in 0.012 milliseconds. We will calculate whether this angular velocity reflects a real situation. We know the time is 0.012 milliseconds, right? Check it out. Yes, it is 0.012 milliseconds, right? Then what is the change in our angular velocity?
How much acceleration is required to complete such a large angle change in 0.012 seconds?
Let’s calculate the place where the motor of the aircraft collides. Assume that our aircraft is rotating on the spot. At the motor location, how much acceleration does it have to withstand in order to achieve this change of 170 degrees per second in 0.012 milliseconds? Let’s play and calculate the radius of our 5-inch machine, for example, 80 meters. That means R2 is equal to 0.08 meters, right?
0.0 seconds, 0.23 seconds, 0.24 seconds
The acceleration is 0.24 divided by 0.012 because this is milliseconds, so we have to multiply it by 1000. How much is this worth to you? This value is equal to exactly or an integer equal to 20,000. What should its unit be? You per square second, right? You per square second. We know what our gravitational acceleration is. What is the gravitational acceleration?
The subtotal neutral acceleration number is 9.8 meters per square second. How many meters are there here? This is equivalent to 2000 meters. This is probably because we assume that if we take 10 meters per square second, there will be 2000 neutral accelerations. What is the concept? This should be impossible. My plane cannot reach it anyway. So after our estimation just now, we know that Gyro Silver may not be telling the truth in this place.
Maybe the angle here can only be from compound time to timing at most. As a result, he understood it too much, which caused unnecessary vibrations of our aircraft. Once the signal of the gyroscope becomes too large, the error will be very large. The error will be very large. The PID will be very large. Once the PID is used, the motor will start to rotate wildly. However, it keeps accelerating and decelerating. It keeps accelerating and decelerating. In fact,
Is it right to move with the left hand and the right hand? You are working with yourself. There is no way to generate any inferences. It all becomes lively. So when this happens, you will find that your motor is very hot. So why does the gyro silver sometimes not tell us the truth? There are various reasons. Let us first take a look at the internal structure of the gyro silver used in the flight control. Generally speaking, there are some fixed ones.
A heavy object connected by a small copper piece and an oil spring. But when heavy objects move in these copper sheets, will they produce various electrical reactions, voltages, and electrical signals? We then detect the current horniness by detecting these electrical signals. However, when these chips are produced, there are various reasons due to process or quality issues.
When the new film comes out, it may not be very, very perfect. There is another aspect, of course, when designing this flight control, the isolation of electronic noise may not have been done well. We know that the electrical environment on our aircraft is actually very, very harsh. These four motors are now constantly going up and down, accelerating and decelerating, and cannot accelerate and decelerate. And sometimes our motors are also very bad and unbalanced in terms of the axis of the motor itself.
So this electrical environment is very bad. If you do not do a good job in shielding the electronics, it will also cause the signal we dragged the machine to be interfered with. As for these reasons, the signal we towed to the Yicai machine was not clean. After it was unclean, the plane amplified the signal without knowing it, creating a vicious cycle.
Okay, today this video will share such a little knowledge. If you think this video is helpful to you, please don’t forget to click three times. I am Kuran. See you in the next video.
