Reaching specific speed with RC car

Hi everyone (=

I want to conduct a series of experiments determining the effects of a handful of parameters of an RC car on its breaking performance. In order to do so I need to get the car up to a certain speed and trigger the breaks at a specific point.

For the latter I want to use a light barrier which transmits the breaking command wirelessly to an Arduino in the car which controls the ESC via PWM.

However, I am stuck at reaching and maintaining a certain speed with the car. I can’t really measure the RPM of the drive shaft because of how the car is built. So I am stuck with measuring from the outside. Could you please recommend ways I could achieve my goal?

Any help is highly appreciated (=

MarkGoingToSpace: Hi everyone (=

I want to conduct a series of experiments determining the effects of a handful of parameters of an RC car on its breaking performance. In order to do so I need to get the car up to a certain speed and trigger the breaks at a specific point.

For the latter I want to use a light barrier which transmits the breaking command wirelessly to an Arduino in the car which controls the ESC via PWM.

However, I am stuck at reaching and maintaining a certain speed with the car. I can't really measure the RPM of the drive shaft because of how the car is built. So I am stuck with measuring from the outside. Could you please recommend ways I could achieve my goal?

Any help is highly appreciated (=

Do you want to load your car down with sensing and transmitting equipment, or are you looking for stationary equipment to then measure speed?

Paul

Paul_KD7HB:
Do you want to load your car down with sensing and transmitting equipment, or are you looking for stationary equipment to then measure speed?

Well I want to use an accelerometer in the car for evaluating breaking performance anyway, but I figured that adding up acceleration in order to determine speed would be pretty inaccurate. Would you disagree?

Most would meter the speed via the wheels. Can’t see adding for example, a hall effect, an Arduino pro mini and a wireless tx would add enough “weight” to make any difference.

MarkGoingToSpace: Well I want to use an accelerometer in the car for evaluating breaking performance anyway, but I figured that adding up acceleration in order to determine speed would be pretty inaccurate. Would you disagree?

That is true. You could measure the motor rpm using a hall sensor or by reading in one of the 3 motor phases.

Hi,
OPs pic,
be5476f4d1f02abf585a580df1623d28d6f4084e.jpg

Tom… :slight_smile:

MarkGoingToSpace:
Well I want to use an accelerometer in the car for evaluating breaking performance anyway, but I figured that adding up acceleration in order to determine speed would be pretty inaccurate. Would you disagree?

If you really want to test breaking performance, throw it off a cliff and see how it survives. I think we’re talking about braking here, which is a whole different thing :slight_smile:

Accellerometers are not all too accurate, and you’ll have a terribly noisy signal as you’ll measure the vibrations of the motor and every tiny bump in the surface you’re running your car on.

As this is for brake testing, you’ll have to do it at the same track over and over again. Therefore controlling the current to your motor should be a pretty good way of controlling your speed.

Other than that, magnets on a wheel (two of them on opposite sides to balance the wheel) and a hall effect sensor would be the most accurate.

The QRE1113 reflective optical sensor is very small but is also easy to attach as it has long "legs" and just needs two resistors. It can detect a spot of white paint on a shaft or wheel. Sparkfun make a breakout board for them but it is not necessary to use a breakout board if you need to squeeze it into a small space. The circuit diagram for Sparkfun's analog board is the one to use to drive an interrupt pin.

...R

bluejets: Most would meter the speed via the wheels. Can't see adding for example, a hall effect, an Arduino pro mini and a wireless tx would add enough "weight" to make any difference.

couka: You could measure the motor rpm using a hall sensor or by reading in one of the 3 motor phases.

wvmarle: If you really want to test breaking performance, throw it off a cliff and see how it survives. I think we're talking about braking here, which is a whole different thing :-)

Other than that, magnets on a wheel (two of them on opposite sides to balance the wheel) and a hall effect sensor would be the most accurate.

Sorry for the wrong spelling (=

And yes I agree with you all. Tbh I didn't know that hall effect sensors were that small until I googled them just now. I will go this route.

Would you say that using a Kalman filter for combining the hall effect sensor with the accelerometer data would be worth while? Or is the hall effect sensor good enough on its own?

Thank you for your numerous and great replies!!

The hall sensor will do just fine. It measures each rotation of the wheel, so as long as you have the circumference measured accurately you will get an accurate speed out of that. An accelerometer will never be able to come near.

wvmarle: The hall sensor will do just fine. It measures each rotation of the wheel, so as long as you have the circumference measured accurately you will get an accurate speed out of that.

. . . as long as you don't allow too much wheel-spin

Don't attach it to a powered wheel, should help a lot.

AWOL: . . . as long as you don't allow too much wheel-spin

I can take a long distance to reach the desired speed. So I don't need to accelerate hard which means that there shouldn't be much wheel spin at all.

wvmarle: Don't attach it to a powered wheel, should help a lot.

It is 4WD so that's not an option.

Add a fifth wheel for measuring then.

I use Alegro A1120 as it switches rail to rail.

AWOL: . . . as long as you don't allow too much wheel-spin

At a constant speed there is no acceleration, so in theorie no force is needed on a flat plane => no wheel spin. All assuming that drag is negligible.

couka: At a constant speed there is no acceleration, so in theorie no force is needed on a flat plane => no wheel spin. All assuming that drag is negligible.

Theories are great, aren't they?