Nintendo Style Speed Booster

[varcsscotty]

ok so we've all played the classic nintendo racing game where you hit the "Z" button and you get a booster, or whatever. What if this could be implemented in a robot.
Say I have a 'dead-band' around my joystick inputs so that they can only go so fast for extra control. But there are those times when you want to get across the field as fast as possible. Alas, The Speed Booster!
Quite simple actually to program, but it may be hard on the machine, I dunno. The program would just include the dead-band code but with a variable for the top joystick input. When the button is pushed, max out the variable. Simple, eh?
Basically the joystick would have some sort of desensitizing function on it but when the button is pressed it is undesensitized...that's not a word is it...

PS I don't know if there is another thread about this topic because the javascript pop-up menus appear to be broken for FF.

[Cory]

People do this all the time.

If you had a super fast robot and wanted to slow it down for precise movement, I could see using it.

It seems like a waste to limit the power of your motors when using this option as if it were a psuedo 2 speed transmission, however.

[Hieb]

Originally Posted by Cory

People do this all the time.

If you had a super fast robot and wanted to slow it down for precise movement, I could see using it.

It seems like a waste to limit the power of your motors when using this option as if it were a psuedo 2 speed transmission, however.

While it may be a waste, for teams with limited resources (money or experience) it is sometimes the only way to achieve 2 speeds.

During our rookie season, we had no driver practice prior to shipping, so it was during our first practice round that we discovered that it was very difficult to control while turning. We used software to scale down our top speed to 1/3 during normal operation, and tied the thumb button on the joystick to full speed while travelling straight. Very inefficient, yes, but given the time constraints and lack of experience I think it was a good solution to the problem. And it was kind of fun to see the robot just kind of crawling along, and then suddenly take off at full speed.

[Denman]

We actually had a total mechanism for this. If you look on our robot its 4 wheel drive but with a twist. There is a mechanism that is controlled by a motor that pushes a pair of castors onto the ground at the front, thus enabling much more manourveablility, at the expense of power. The idea was that you use 4wd to get the tetra to the goal and use the castors to align it more accurately. However the welds got smashed off during a nasty bump in one of the later rounds.
This pic is the best i can find.

If you look there there is a V shaped piece, which was later attached to the black motor. We then added a little flag that went up when the castors were down.

[Pat McCarthy]

For our 2005 robot, it was crazy fast and jerky when trying to get the precision needed to cap well. So we made the trigger button on the joystick decrease the robot's top speed by 2/3 or so.

[Greg Needel]

Originally Posted by Cory

It seems like a waste to limit the power of your motors when using this option as if it were a psuedo 2 speed transmission, however.

it really isn't limiting the power of the motors. since the IFI setup uses pwm(pulse width modulation) the motors are still receiving maximum voltage, it is just the pulse length that is shorter. and since
Torque(transmitted to the shaft) = (Horsepower*63000)/rpm

and since 1 HP = 745.699872 watts

and watts= voltage^2/ohms <--constant in this case because resistance of the motor doesn't change and neither does the voltage.

the total torque transmitted at a given rpm is the same, all that is being changed is your speed. This is the same thing that happens when you push the joysticks only a little bit forward.

we did this this past year and it works very well for precise movements on the field.


although to make a "speed boost" button you would have to be limiting your pwm the whole time except when you want to boost, which would be a waste.

[Gdeaver]

While not the same as programing 2 response curves, you could use an exponential response function to give a larger low speed area on the joy stick. Thats what our team did last year.

[Dave Scheck]

Originally Posted by varcsscotty

ok so we've all played the classic nintendo racing game where you hit the "Z" button and you get a booster, or whatever. What if this could be implemented in a robot.

First off, shame on you for making me feel old ...classic nintendo turbo will always be the "B" Button

We've actually been using such a system for years. We usually use the top button on the joystick as our "turbo" button. The trick is that when the button is not pressed, we scale the full range of the joystick so that it maps to +- half speed. When the button is pressed, we take this scaling away and the full joystick maps to the full output.

[Matt Leese]

Originally Posted by Greg Needel

it really isn't limiting the power of the motors. since the IFI setup uses pwm(pulse width modulation) the motors are still receiving maximum voltage, it is just the pulse length that is shorter. and since
Torque(transmitted to the shaft) = (Horsepower*63000)/rpm

and since 1 HP = 745.699872 watts

and watts= voltage^2/ohms <--constant in this case because resistance of the motor doesn't change and neither does the voltage.

the total torque transmitted at a given rpm is the same, all that is being changed is your speed. This is the same thing that happens when you push the joysticks only a little bit forward.

we did this this past year and it works very well for precise movements on the field.


although to make a "speed boost" button you would have to be limiting your pwm the whole time except when you want to boost, which would be a waste.

This is wrong but I don't think I should delete it, just didn't want anyone to pay attention to it:

No they're not. To control the speed of a DC motor, you vary the voltage applied to it. PWM is used solely as a communication mechanism between the RC and the speed controller. It's not used for controlling the actual motor (you'd probably break the motor if you tried). The actual inside of the speed controller consists of an H-Bridge which allows for variable voltage control. So, by limiting the speed, you're also limiting the torque output from the motor. However, you may want to do this as it can allow finer grained control over the robot.

Matt

[Alan Anderson]

Originally Posted by Matt Leese

To control the speed of a DC motor, you vary the voltage applied to it. PWM is used solely as a communication mechanism between the RC and the speed controller. It's not used for controlling the actual motor (you'd probably break the motor if you tried). The actual inside of the speed controller consists of an H-Bridge which allows for variable voltage control.

The Victor speed controllers do use pulse width modulation of the power to the motor. The H-switch inside them is a digital device, either completely off or completely on in one direction or the other. The average voltage to the motors is controlled in approximately 1% steps, but the peak is always full battery voltage.

[jdiwnab]

Originally Posted by Dave Scheck

First off, shame on you for making me feel old ...classic nintendo turbo will always be the "B" Button

lol I was going to say the same thing. "classic" stuff doesn't have "Z". However, you might be able to make a case that they may have A, B, X, Y, L, and R.

[Dave Scheck]

Originally Posted by jdiwnab

lol I was going to say the same thing. "classic" stuff doesn't have "Z". However, you might be able to make a case that they may have A, B, X, Y, L, and R.

Go back one more generation and you only have A and B...ahh, those were the days.

[Matt Leese]

Originally Posted by Alan Anderson

The Victor speed controllers do use pulse width modulation of the power to the motor. The H-switch inside them is a digital device, either completely off or completely on in one direction or the other. The average voltage to the motors is controlled in approximately 1% steps, but the peak is always full battery voltage.

You're right about the PWM control although I'd say the H-bridge is also an analog device but that's beside the point. I really need to think through before I post.

While peak-voltage is still the same, the voltage changes over time. What that means is that the power is changing over time. But it's happening fast enough that it's really not noticable. Instead, you'd be looking at the average voltage for power calculations, not peak power.

Matt

[Manoel]

Originally Posted by Greg Needel

it really isn't limiting the power of the motors. since the IFI setup uses pwm(pulse width modulation) the motors are still receiving maximum voltage, it is just the pulse length that is shorter. and since
Torque(transmitted to the shaft) = (Horsepower*63000)/rpm

and since 1 HP = 745.699872 watts

and watts= voltage^2/ohms <--constant in this case because resistance of the motor doesn't change and neither does the voltage.

the total torque transmitted at a given rpm is the same, all that is being changed is your speed. This is the same thing that happens when you push the joysticks only a little bit forward.

we did this this past year and it works very well for precise movements on the field.


although to make a "speed boost" button you would have to be limiting your pwm the whole time except when you want to boost, which would be a waste.

First, you should check your formulas. "Watts" = Voltage^2/R (and note that "Watts" should really be Power) is the electrical power to the motor. Considering it is not perfect, there are lots of losses and the mechanical power available at the motor output is omega times torque. They're not the same and the electrical power (supplied) will always be higher.
Second, the PWM does change the voltage seen by the motor. Even though you are pulsing it between 12V and 0 V, the motor can be seen as an RL (resistor-inductor) series association and thus this pulsed voltage is smoothed and you get an average voltage that is proportional to the PWM duty cycle. I don't want to extend and there are lots of old posts about it, good information can be found there.
When you change the voltage to the motor, you can simply scale its speed/torque curve by a factor of (applied voltage/test voltage), so you are actually changing both its speed and torque (and power, as a consequence).

Concerning the speed boost, we used that one year. It sucked.
The only possible use I see is to make a very fast robot a bit more controllable, but then again your gearing is probably wrong and you are (again) wasting power.
To me, the "coolness factor" of this solution can really hinder your performance.

If your opponent is coming at you with 4 CIMs full power, why in the world would you want to get in a shoving match with your robot at 70%?

[Madison]

Originally Posted by Hieb

While it may be a waste, for teams with limited resources (money or experience) it is sometimes the only way to achieve 2 speeds.

Remember, however, that such software limitations in no way approximate the functionality of a multiple-speed, mechanical transmission. They're simply a convenient way of limiting your students' lead feet.

Our team met last night to implement 'follow-wheels' on our 2005 robot and did so with great success -- or so I'm told; I was busy doing something else. These unpowered wheels, coupled to encoders, remove from our drivers consideration any notion of voltage or power and instead make them concerned only with position. The joystick input no longer controls average voltage, but instead indicates desired position or speed. The feedback loops do the rest in making sure the robot does what it's meant to do.