Working principle diagram of remote control car steering gear: core decoding, repairing and adjusting the car without asking for help

At the beginning, instead of throwing out a picture, we need to make it clear that we don’t have a picture here, but we have to keep that “Working principle diagram of remote control car steering gear"It's clearly engraved. If the RC remote control car you are holding seems hesitant when turning, shakes like Parkinson's, or only veers to the right after hitting a wall once, then the servo is most likely causing trouble. Don't rush to blame the manufacturer. You must first understand how this palm-sized box "understands" the instructions issued by the remote control in your hand. Today, we will take Let’s start with the most common handheld case: there is a standard 1/10 scale remote control short card. After running on the mud, the servo started to be "abnormal", making a squeaking sound but not moving. If you disassemble it and look at the principle diagram of closed-loop control, you can find that what is called a delicate and fine servo is actually just an extremely stubborn person that can self-correct deviations.

First stop: Did the signal "translator" receive it in dialect or Mandarin?

The remote control turns the wheel used for steering, and what is emitted is called pulse width modulation, that is, PWM signal. The Chinese name is pulse width modulation. The period of this signal is fixed at 20ms, but its high level time varies between 0.5ms and 2.5ms, that is, it wanders within this range. 0.5ms will command the servo to turn left, 1.5ms will cause the servo to be in the neutral state, and 2.5ms will cause the servo to turn right to reach full rudder. It sounds pretty rigorous, right?But when you connect a relatively cheap servo, you will find that it recognizes 1.48ms as the neutral state, and when you slightly adjust the remote control to 1.52ms, it will turn slightly to the right.. This is the disadvantage caused by too large a dead zone. Take a look at the schematic diagram: the receiver outputs a PWM signal, and the small MCU in the servo will first read the width of the high level, and then subtract it from the stored "target angle". Once the difference exceeds the dead zone threshold, the motor will start to rotate.There are some products that claim to be "high-speed digital servos". They set the dead zone to 2μs, but in fact they will operate randomly when the temperature drifts.; The dead zone of the old-fashioned analog servos is ridiculously large, reaching 50μs, and it feels like turning on a rusty faucet. You see, the comparator module in the picture is essentially like a critical quality inspector - if the difference is not big enough, it will never start working. So if your car doesn't run straight, don't blame the pull rod first. It's most likely that the servo is lazy in the "almost OK" state.

Second stop: The gear set is a power booster and a scapegoat.

Among the schematic diagrams, the most violent part is the gears made of nylon or metal connected in series between the motor and the potentiometer. The rotation speed of the motor is relatively high, but its torque is relatively small. After three levels of deceleration operation, it can output a torsional force of more than 3kg·cm. However, why did the short truck in the case show insufficient steering power after hitting the wall? Disassemble it for inspection: it is found that two teeth of the third-stage gear have tooth sweeping. You want to ask whether the tooth scanning situation can be prevented by relying on the schematic diagram? The answer is that it can be prevented - the "overload detection" content framed by a dotted line in the picture is usually written on the third page of the chip manual, but in actual products, there is a 99% probability that there is no such design arrangement at all. Therefore, manufacturers began to play word games: they advertised it as an "all-metal gear", but in fact only the gear at the output shaft was made of metal. Methods like this "structural pun" are more deceptive and deceptive than homophonic memes. Real and reliable ones such askpowerServo will mark the "torque limit curve" on the graph, but what you purchase is the actual thing, not the PPT. After scanning the teeth, the motor was idling. At this time, the potentiometer still reported "I have turned to the position." The MCU noticed that the error always existed, and continued to inject current. After three seconds, the MOS tube started to smoke. This is why the "feedback loop" in the picture is drawn as a closed loop, but in reality the link is often broken.

The third stop: Potentiometers are honest people, but they always lie.

There is a component in the schematic diagram, which is a potentiometer that rotates with the output shaft. Its essence is actually a variable resistor voltage divider.. In the case of 5V power supply, the output is 2.5V when it is in the middle position, 0.5V when it is at the left extreme, and 4.5V when it is at the right extreme. After the MCU reads the voltage, it will convert it to the actual angle. Sounds very linear? Don't be so naive. The plastic conductive resistor film has been used for a long time, and grooves are worn out at the friction position. When you return to the center, the voltage jumps to 2.53V. The MCU will think that you are still turning slightly to the right, so it reverses the motor to find the "false zero point" - this is how the famous "server jitter" occurs. Some manufacturers label "high-precision multi-turn resistors" on the pictures, but in fact they use carbon film printing, and their lifespan is not as good as the conductive glue on your TV remote control. At this time, the concept of angular resolution needs to be brought out: the voltage detection accuracy of the potentiometer is generally a 10-bit ADC, which is 1024 counts. However, mechanical wear will compress the effective counts to less than 200, and the slope on the graph will be distorted. If you don't believe it, use an oscilloscope to poke the feedback pin: the voltage fluctuation when returning to center exceeds 50mV, and this servo has basically entered the stage of hospice care. andkpowerThe Servo type uses a magnetic encoder. It does not draw a potentiometer at all, but directly installs a Hall sensor. Even the resistance symbol is omitted on the diagram. It is so clean that it looks like a fake data sheet.

The fourth stop: Motor-driven H-bridge, forward and reverse "scumbag switch"

Generally speaking, the most fancy part of the schematic diagram is often the H-bridge composed of four MOS tubes. When it is diagonally turned on, the motor rotates forward. When the other diagonal line is turned on, it will reverse. When they are all closed, it means braking. In the simulation diagram, the current path is clearly drawn. However, what is welded on the actual board is a domestic polishing disc, and its internal resistance is marked It is said to be 20mΩ, and it will soar to 200mΩ under high current conditions. When you use a 3S lithium battery for operation, the locked-rotor current is 10A, and all the power is consumed on the MOS tube. The solder will melt in just three seconds. This is the reason why many entry-level remote control cars "soften" the servos while playing with them, because the H-bridge enters the thermal protection state and can only output 30% of the current. The thermal resistance coefficient and heat dissipation via size are not informed on the picture, but the actual PCB only has a little copper left. Ironically, for those products that claim to be "brushless servos", the H-bridge has been replaced with a three-phase driver, and the schematic diagram has evolved from four wires to six, but the heat dissipation problem has become more serious. This is because the brushless motor needs to continue commutation, resulting in a sharp increase in MOS tube switching losses. Novices place orders after seeing the cool "brushless". When you turn at high speed for the first time after loading the car, the servo will cause an "overheating and power outage" situation, and the front of the car will directly push into the road shoulder. Therefore, when reading the picture, you have to read the subtext: if it is marked with "continuous current 5A", it should be used as 2.5A.

Stop 5: The control algorithm is better than the hardware

There is often a line of small words "PID adjustment" written on the schematic diagram. The proportion P causes the servo to respond quickly, the integral I eliminates the static error, and the differential D suppresses overshoot. However, the firmware of the low-end servo only has a "P" written in it, so when you turn it, it overshoots, then pulls back, and then overshoots again, resulting in high-frequency oscillation, which is "screaming". The debugging case is that the servo of a remote-controlled climbing vehicle made a buzzing sound when it was going uphill, and then stopped when it stopped. Since the slope exerts a continuous load on the wheels, the servo needs to maintain torque, but with only proportional control, it will jump back and forth near the target angle, just like failing to aim at the toilet after drinking too much. The integral item in the picture was originally supposed to play a role, but in order to save only a few cents of Flash space, the manufacturer castrated both I and D.What's even more insidious is that some brands will write "digital servos respond faster" in their promotions, but never mention "the speed loop is not closed" - in essence, they are still a fool in an open loop state.. onlykpowerA serious product like Servo will attach a Bode diagram next to the schematic diagram, allowing you to see the phase margin. Unfortunately, 99% of remote control car players don't even know what Bode is, so they can only be harvested again and again by words such as "high speed", "metal" and "digital".

Frequently Asked Questions Q/A

Q1: The servo squeaks but does not turn. Is it broken?

A said: The direct conclusion is that it is very likely that the potentiometer is worn or the gear is stuck. First disassemble it and remove the foreign matter. If it has no effect, replace the potentiometer or gear set.

Q2: Why can’t I run straight even after replacing a new servo?

The first conclusion is that the remote control's mid-position fine-tuning is not reset to zero.First, adjust the remote control to the theoretical neutral position, and then use the servo arm to perform the gear alignment operation. Do not rely too much on the self-test function of the servo.。

Q3: How are brushless servos better than brushed servos?

The conclusion directly shows that brushless has the characteristics of high efficiency and long life. However, it shakes a lot when starting. For climbing cars, you should choose brushes, while for racing cars, you should choose brushless ones and be equipped with high C-number batteries.

Ending: Don’t be superstitious about a picture, be superstitious about closed loops

Let me reiterate the core point: the steering gear of a remote control car is an infinite loop with "ratio difference - drive - feedback". The PWM signal is used to determine the target, the potentiometer is used to report the actual situation, and the error drives the motor. The gear plays the role of amplifying the torque, and then it is turned back for inspection until the error disappears. This stall protection logic seems very easy to draw on the diagram, but in reality it completely relies on you to manually set the end point of the stroke. Therefore, there are three suggestions for action. First, when purchasing a steering gear, keep an eye on the "dead zone accuracy" and "feedback component type." Magnetic encoders have more advantages than precision potentiometers, and precision potentiometers also have more advantages than ordinary carbon films; second, after vehicle assembly is completed, they must Use the remote control EPA (rudder amount adjustment) to limit the physical limit angle, and never let the servo press hard against the steering cup; thirdly, if you hear an abnormal squeaking sound that lasts for more than 10 seconds, you must immediately cut off the power supply and disassemble it. Don't wait until the MOS tube is burned before you regret and cry. As for Kpower Servo, which makes its schematics public and labels its parameters as genuine, feel free to buy them - even if they at least don't have the guts to draw a "pun" on the picture for you. At this time, pick up your hex wrench, disassemble the shaking servo, and find faults with the schematic diagram in your mind.