Ultra-Capacitor Powered Car

Thread Starter

GrantRitchie

Joined Nov 9, 2015
6
I'm working on a project to design and build a 1/10th scaled ultra-capacitor powered electric car. I'm a Mechanical Engineer & I really want to focus on the the building and testing of the vehicle - not the electronics, which is why I need your help!

As I have a limited electronics knowledge I was hoping that I could run my initial control system design past some of you in the hope you'd be able to confirm my ideas and maybe point me in the direction of some suitable hardware. Or help correct my errors.

Here is a basic modular outline of my design:


The plan is to have 3 ultra-capacitors connected in series using the Maxwell Integration Kit. I know I need a Buck-Boost converter with the voltage and current rating of the motor. Ideally I'd like something off the shelf - I was investigating these modules from Linear Technology (http://www.linear.com/solutions/Supercap,_Capacitor_*_Battery_Back-Up_Solutions). Would these do the job? Or is there something better out there without creating one from scratch?

The motor can then be decided based on these parameters of the rest of the control system but I require a minimum motor torque of around 0.2Nm. I'll probably go for a DC brushless based on efficiency.

The vehicle is intentionally simple and the target is to drive a 50m straight journey at a constant speed.

Thanks in advance for your time and please let me know what you think.
 

#12

Joined Nov 30, 2010
18,224
Numerous previous discussions have arrived at, "Why you want to complicate this so much when a battery will work better and cheaper?"
 

Thread Starter

GrantRitchie

Joined Nov 9, 2015
6
Numerous previous discussions have arrived at, "Why you want to complicate this so much when a battery will work better and cheaper?"
I'm wanting, on a small scale, to prove that caps are viable for electric vehicle power storage once the energy density matches that of batteries (i.e. after future work on graphene based ultra-capacitors etc) & it's more fun!!
 

#12

Joined Nov 30, 2010
18,224
You can prove exactly what the future technology will need as its specifications, with nothing but math. Proving something you don't have by a physical demonstration is impossible, so I think the real answer is, "fun".

I have no objection to fun. In fact, I prefer silliness, but you go ahead and have all the fun you want, especially if it's educational. I was mostly worried about you lying to yourself about your goals and methods.
 

Thread Starter

GrantRitchie

Joined Nov 9, 2015
6
You can prove exactly what the future technology will need as its specifications, with nothing but math. Proving something you don't have by a physical demonstration is impossible, so I think the real answer is, "fun".

I have no objection to fun. In fact, I prefer silliness, but you go ahead and have all the fun you want, especially if it's educational. I was mostly worried about you lying to yourself about your goals and methods.
Yeah it most definitely is about fun! It's nice to have some sort of link to future applications but it's more interesting dealing with ultra-caps than with batteries. With that in mind, any idea about the type of buck-boost I require and if I can get something off the shelf?
 

blocco a spirale

Joined Jun 18, 2008
1,546
I'm wanting, on a small scale, to prove that caps are viable for electric vehicle power storage once the energy density matches that of batteries (i.e. after future work on graphene based ultra-capacitors etc) & it's more fun!!
That assumes battery technology will remain at the current level. The problem you have with a capacitor is that as soon as you put a load on it, the voltage falls and continues falling and in order to maintain the same power output you need to draw more current which increases losses when you really want to be reducing them. You don't have much voltage to work with so you need to find a DC-DC converter with the lowest possible working input voltage. I think you may just end up proving why capacitors aren't suitable as the primary energy store in an electric vehicle but it's still a worthwhile challenge.
 

Thread Starter

GrantRitchie

Joined Nov 9, 2015
6
That assumes battery technology will remain at the current level. The problem you have with a capacitor is that as soon as you put a load on it, the voltage falls and continues falling and in order to maintain the same power output you need to draw more current which increases losses when you really want to be reducing them. You don't have much voltage to work with so you need to find a DC-DC converter with the lowest possible working input voltage. I think you may just end up proving why capacitors aren't suitable as the primary energy store in an electric vehicle but it's still a worthwhile challenge.
Yes I understand that the voltage will fall on the capacitors but I'm only looking to run the vehicle for around a minute anyway. I understand that the technology is currently not viable for the industry but that's why I'm experimenting and testing the performance of such a vehicle. Could you direct me towards a buck-boost that would meet the requirements? I'm a beginner with this sort of thing and I'm struggling to locate a buck-boost that meet the requirements as you mentioned.

Cheers.
 

blocco a spirale

Joined Jun 18, 2008
1,546
Yes I understand that the voltage will fall on the capacitors but I'm only looking to run the vehicle for around a minute anyway. I understand that the technology is currently not viable for the industry but that's why I'm experimenting and testing the performance of such a vehicle. Could you direct me towards a buck-boost that would meet the requirements? I'm a beginner with this sort of thing and I'm struggling to locate a buck-boost that meet the requirements as you mentioned.

Cheers.
Start by going through manufacturers datasheets to find whatever comes closest to your requirements.
 

wayneh

Joined Sep 9, 2010
17,496
... it's more interesting dealing with ultra-caps than with batteries.
You might seek professional help for that bias. A battery has a natural and huge advantage for this application - a near constant voltage as its energy is drawn down. Why insist on using the wrong technology?

That said, you just want a DC-DC converter that can take an input of 0-9V and output your motor voltage, all at the current draw of your motor. Obviously, you're not going to find a converter that can use 0V, but the closer the better, since this will allow you to access more of the capacitor's stored energy. Capacitors have an advantage over batteries in this sense - they can go to full discharge without damage. Batteries can't do that.

I'd be tempted to use a 9V motor so that your converter is always in boost mode, or a low voltage motor (assuming there is such a thing) so that it's always in buck mode. I really don't know, but I'm guessing that a buck or a boost converter is cheaper than a combination buck/boost.
 
Last edited:

tcmtech

Joined Nov 4, 2013
2,867
I'm wanting, on a small scale, to prove that caps are viable for electric vehicle power storage once the energy density matches that of batteries (i.e. after future work on graphene based ultra-capacitors etc) & it's more fun!!
I don't get that statement at all. You already know the best high end capacitors still come up way short of batteries on a equal size and mass comparisons to stored energy ratios so what exactly is there to show that's not already self evident and well known? o_O
 

blocco a spirale

Joined Jun 18, 2008
1,546
The challenge here is to make the most efficient use of the energy stored in the capacitor so you need to have very precise control over the rate of acceleration and the speed of the vehicle, exactly as you would drive a car to minimise fuel consumption.

The problem you have with a DC-DC converter, aside from conversion loses, is that it will have a minimum input voltage, let's say that threshold is 1V, below which the output drops to zero and the car stops but there is still energy remaining in the capacitor. But, a low-voltage brushless motor could still produce usable torque at 1V or below (the car is now rolling so requires less energy) so why not drive the motor directly from the capacitor. The key is to use a micro-controller (powered from a dedicated battery) to provide the control pulses to the motor. The micro-controller could also measure speed and capacitor voltage/ rate of discharge and adjust drive to the motor to optimise efficiency.

This reminded me of The Great Egg Race TV series where contestants had to transport an egg using only the energy stored in a rubber band. They often used pulleys with complex profiles in an attempt to compensate for the continuously reducing energy available and the reducing energy requirement of the vehicle as it reached its top speed. You could also use a fixed control profile, it might not be as efficient but would be much easier to implement.
 
Last edited:

Thread Starter

GrantRitchie

Joined Nov 9, 2015
6
You might seek professional help for that bias. A battery has a natural and huge advantage for this application - a near constant voltage as its energy is drawn down. Why insist on using the wrong technology?

That said, you just want a DC-DC converter that can take an input of 0-9V and output your motor voltage, all at the current draw of your motor. Obviously, you're not going to find a converter that can use 0V, but the closer the better, since this will allow you to access more of the capacitor's stored energy. Capacitors have an advantage over batteries in this sense - they can go to full discharge without damage. Batteries can't do that.

I'd be tempted to use a 9V motor so that your converter is always in boost mode, or a low voltage motor (assuming there is such a thing) so that it's always in buck mode. I really don't know, but I'm guessing that a buck or a boost converter is cheaper than a combination buck/boost.
Thanks Wayneh, that's useful! Batteries are very well investigated and tested - it's more interesting and educational using the caps. I'm not interested in trying to prove that they are better than batteries, I'm just interested in testing them in this context and consolidating the data to describe their performance. That can then lay the ground work for further research on the possibility of using Caps for electric vehicle power in the future.

The main reason for using the buck-boost would be to utilise the boost mode when running the motor and then the buck mode when charging the caps (charging will be done from a DC power supply) So choosing a 9V motor would mean that the 3x2.85V caps would always require boost to run the motor. The issue is trying to find a DC-DC Converter that means this ~0-9V range and a usable current for a 9V motor. I've looked at these http://www.linear.com/product/LTC3350 but they seem to be designed for back-up applications and I'm unsure of they can be used in this context. Do you think it's something along these lines that I would need? Or would it be simpler to implement this in a motor controller? Thanks.

The challenge here is to make the most efficient use of the energy stored in the capacitor so you need to have very precise control over the rate of acceleration and the speed of the vehicle, exactly as you would drive a car to minimise fuel consumption.

The problem you have with a DC-DC converter, aside from conversion loses, is that it will have a minimum input voltage, let's say that threshold is 1V, below which the output drops to zero and the car stops but there is still energy remaining in the capacitor. But, a low-voltage brushless motor could still produce usable torque at 1V or below (the car is now rolling so requires less energy) so why not drive the motor directly from the capacitor. The key is to use a micro-controller (powered from a dedicated battery) to provide the control pulses to the motor. The micro-controller could also measure speed and capacitor voltage/ rate of discharge and adjust drive to the motor to optimise efficiency.

This reminded me of The Great Egg Race TV series where contestants had to transport an egg using only the energy stored in a rubber band. They often used pulleys with complex profiles in an attempt to compensate for the continuously reducing energy available and the reducing energy requirement of the vehicle as it reached its top speed. You could also use a fixed control profile, it might not be as efficient but would be much easier to implement.
Thanks for your reply, it's really helpful! I understand that it's difficult to implement a control system that will allow the full energy of the caps to be used. The goal of the vehicle however is not to travel as far as possible on one charge but to travel 50m. Therefore I do not indent on completely draining the caps before the opportunity for recharge. I know this is not useful in terms of real word application but it will be enough to provide the data I want to extract and provide an opportunity for further improvement work in the future.

Excuse my lack of knowledge, but how technical is the process of programming a micro-controller to control the motor? (I have a basic understanding of C++ coding). Is it more sensible to buy something of the shelf or start from scratch?

Could you briefly elaborate on a fixed control profile within this context? Thanks
 

blocco a spirale

Joined Jun 18, 2008
1,546
By fixed, I mean that your motor control algorithm follows a fixed/blind routine rather than taking input from a speed sensor and measuring capacitor voltage etc.

Small micro-controllers are usually programmed in assembly language or C. There are some systems, such as Arduino, that use something else, I don't know what exactly, that might be easier to get started with. It shouldn't be too difficult to program if you start with a routine that provides sequenced fixed pulse widths to the motor coils. You then need to write a routine that slowly increases the speed of the motor to whatever you decide is optimal and then hold it there using just enough power to maintain that speed. You might also want to shorten the pulse widths at this point to reduce power even further.

Two stage control like this could also be done using hardware only;There are chips available that take care of the commutation for BLDC motors and require just a voltage or PWM input to set he speed but they may not provide the level of control that you need.
 
Last edited:

wayneh

Joined Sep 9, 2010
17,496
If you don't care how far the caps are drawn down, then just run a ~6V motor off the caps directly. The motor will survive the ~9V peak voltage - because it won't see that for a significant length of time - and it will quit turning when the voltage on the caps falls to 2V or so. Done.
 

ronv

Joined Nov 12, 2008
3,770
Now might be a good time to try and pick a motor so you can understand how much power is needed. I think you have more than enough if you don't want to set any speed records for the 50 meters and the car is not loaded with rocks.:D
A PWM speed controller and a motor is much like a buck converter so I don't think you will need the boost converter, just a low voltage motor.
 

Thread Starter

GrantRitchie

Joined Nov 9, 2015
6
I think the idea of using a motor controller instead of a designated buck-boost converter seems sensible. But it would only need to be simple as I do not require a great amount of control, just basic speed control would suffice. However as the caps have a current of >100A the motor controller would have to limit the current flowing to the motor.

I see two options now:

1. Use something like an Arduino to programme a basic fixed routine based on fixed pulse widths to the motor coils as suggested by blocco a spirale.

2. Use an off-the-shelf PWM speed controller and modify accordingly. (This is preferred if I can find something suitable)

The question remains - can I limit the current through the controller?

Now might be a good time to try and pick a motor so you can understand how much power is needed. I think you have more than enough if you don't want to set any speed records for the 50 meters and the car is not loaded with rocks.:D
A PWM speed controller and a motor is much like a buck converter so I don't think you will need the boost converter, just a low voltage motor.
I would like to be able to choose a motor but I need to decide on the control system first. If I go for just a buck system then my motor will be <9V but if I go for a boost system my motor must be >9V. Suitable candidates of both voltages would provide the 0.25Nm.
 

kubeek

Joined Sep 20, 2005
5,794
I think what you need is a buck-boost controller. You need the buck in order to not spin the wheels at the start, and the boost to get most of the remaining charge out of the caps.
The caps will store roughly 14KJ of energy, which say at 50W will last you some 270 seconds (minus losses, unrecoverable charge etc.)
But note that if you want 50W out then when your caps get to just 2 volts you will need the controller to provide 25A average.
 

ronv

Joined Nov 12, 2008
3,770
I think the idea of using a motor controller instead of a designated buck-boost converter seems sensible. But it would only need to be simple as I do not require a great amount of control, just basic speed control would suffice. However as the caps have a current of >100A the motor controller would have to limit the current flowing to the motor.

I see two options now:

1. Use something like an Arduino to programme a basic fixed routine based on fixed pulse widths to the motor coils as suggested by blocco a spirale.

2. Use an off-the-shelf PWM speed controller and modify accordingly. (This is preferred if I can find something suitable)

The question remains - can I limit the current through the controller?



I would like to be able to choose a motor but I need to decide on the control system first. If I go for just a buck system then my motor will be <9V but if I go for a boost system my motor must be >9V. Suitable candidates of both voltages would provide the 0.25Nm.
You don't need to limit the current to the motor if you keep the voltage to it within spec. If you think about a 12 volt motor in a car tied to a 12 volt battery there is no problem because the motor cannot draw to much current by design.
The micro might be a good choice if you know how to program it. They are easy to get pwm outputs.
I suggest the 3 volt motor because when you try to boost the voltage you loose some power in the conversion process and most boost converters don't like to run below 2 or 3 volts.
With say a 3 volt motor you could set the pulse width to say 30% to start and as the voltage of the cap goes down increase it until it is 100%. It would still run down to less than a volt.
Finish up the calculations on the motor and wheels. My napkin says a 2" diameter wheel and a 180 RPM motor would get you across the finish line in under a minute. Would that work?
 

wayneh

Joined Sep 9, 2010
17,496
I wager it will be very hard to beat the strategy I proposed in #15, that uses no controller, converter or computer. I guess I'm a sucker for simplicity.

If your goal is to solve the general problem, to build a system that outputs a constant voltage while fed a decreasing supply voltage, then you can entertain these more complex strategies. But to move a model car across the finish line, complexity is overkill.
 
Top