Proper Speed Controller Circuit for 24 VDC motor : electric bycycle

Hi, I'm an American living in Bangkok, Thailand. I'm new to this forum and an absolute novice with basic electronics, I was hoping someone here can assist me with a few basic questions.

I'm planning to convert my bycycle electric in the near future (after some electrical enlightenment of course). I've basically narrowed the motor down to a 24 VDC 250 watt motor. (Though 350 and 500 watt are also options)

Using Ohm's law (I think), I calculated that such a motor's current draw would be aprox. 10.416 amps (250 watts / 24 volt = 10.416 amps), which I assume from reading other threads here, is what the motor would draw an one hour under load. (Please don't hesitate to correct me--elaborate)

As for power supply, I've decided that I want to rig six 12 volt 10 amp/hour batteries in series-paralell for a total of 24 volts, 30 amp/hour power supply. This I figure should give me roughly two hours of motor work between charges...

Assuming my calculations are correct, my next task is to figure out the necesarry (speed) controller. I've browsed the many schematics/theories for DC motor controllers and am lost. I'm assuming a Pulse Width Modulator circuit is what I'm looking for (rigged with a handle grip motorcycle throttle set-up for precise control). So, then, I would like to know is how to decide which particular circuit will suffice for the amp and volt consumption required for my particular project.

Most every PWM circuit/schematics I've come across on the web are for low amp consuming loads. As I mentioned, the motor will be drawing atleast 10 amps. What is the theory and/or calculations for amp consumption in DC motor controllers?

Ideally, a universal controller that would work for 250, 350, and 500 watt

24 VDC motor with minimal power loss would be best. Ofcourse, I would like to know what determines/controls the current load/consumption in a control circuit. There is so much to decipher among---MOSFET, other capacitors, diodes, that I have still yet to learn their functions, purposes.

While we're at it, how does adding more amp/hours (placing more cells in paralell) to the power supply (battery pack) affect the control circuit? In other words, how must the control circuit be modified to accomadate the extra current without overloading/overheating the circuit, load, etc.?

Thanks in advance,

Steven

This message was sent using the sci.electronics.basics web interface on

That calculation has nothing to do with time. If the motor is rated at 250 watts and 24 volts then the only way it can consume 250 watts is if it passes 10.4 amperes. How long it does that depends on how large the battery is (ampere hour rating). You will also probably not ask the motor to produce that power continuously, but may run it at lower current (torque) much of the time.

I think you have this right.

I think the best drivability may come from an adjustable torque control (current control). This is because you feel the torque as acceleration, regardless of what speed you are traveling. So I would look for a PWM controller that included current feedback, and that used the throttle as a torque (current) set point.

Current control also protects the motor from overload in the event that you come to a slope too steep for it to handle. If the setpoint can reach only the rated motor current (or slightly above), stalling the motor will not overheat it.

That is just a matter of the current rating of the PWM controller.

No. it just provides power for a longer time. The supply voltage affects the top speed.

10 amp hours is the size of a motorcycle battery. I hope you can get deep cycle type batteries in that size. Deep cycle batteries are what you need. A deep cycle is designed to drive a steady load for a long time and thus will last much longer than a starter battery.

The elements that actually carry the heavy current that the motor runs on will be either mosfets or a big igbt or igbts. These are modern transistors that do not require much power on the gate to turn them on and off, which simplifies the gate drive requirements. Thus the brains of the controller will consume a very, very small proportion of the total power. The major power concerns lie with the big semiconductors. If you are going to build this thing yourself you will need to know the differences between mosfets and igbts. Mosfets behave like resistors, and can be paralleled to carry heavier currents than a single mosfet can handle. On the other hand, igbts have a voltage drop across them instead of a resistance which means that power dissipation doesn't increase as much in an igbt at higher current levels as it does in a mosfet. In simple terms, power dissipation in an igbt is proportional to current (P=VI), while the mosfet's power dissipation increases with the square of the current (P=I^2 R). You should do some internet browsing of datasheets, app notes etc. What you're really looking for is a device designed for the purpose. There might be a motor-drive igbt on the market. There may design requirements like voltage clamping built in to it so that you don't have to design them into your circuit. If you have the perseverance to do the necessary research on the power side, you might even be able to do the PWM too.

Why? IGBTS don't even enter the the picture according to

Mosfets behave like

What the hell does that paragraph mean? Mosfets behave like resistors? IGBTS have a voltage drop across them instead of a resistance? Does "On the other hand" imply Mosfets don't have a voltage drop across them? The equations P=VI and P=I^2R are identical. Why do you think they are different?

Whatever it is you have in mind isn't coming across.

Ed

You should do some

Hello Steven I used to mess with this sort of thing some years back. In fact there is a 24V bicyle rusting away under the porch as we speak. Let me get back home tonight and I will dig through some of my old designs and see what I can modify to make suitable for your application. Regards Robert

A kid's electric scooter has a 20A 24V controller. About the size of a pack of cigarettes. They're made by the zillion, so expect that they're probably much cheaper than anything you could build. Do some research on the electric scooter sites. Don't know anything about Thailand, but here in the USA, you can buy the whole scooter at a garage sale under $10. Batteries will be dead, but...

I'd rethink the batteries. If you're on flat land, most of the time you'll be using much less than the maximum power.

You should shock mount the batteries somewhat. Bikes are very hard on stuff.

There are also gearing issues. My scooter maxes out at about 10mph. And it still doesn't have enough torque to make it up my driveway. The slightest hill slows it down considerably.

I'm finding that riding my bike is less hassle than riding the electric scooter. It's very easy to pedal on any slope that the electric can handle.

mike

--
Wanted, Serial cable for Dell Axim X5 PDA.
Return address is VALID but some sites block emails
with links.  Delete this sig when replying.
FS 500MHz Tek DSOscilloscope TDS540 Make Offer
Bunch of stuff For Sale and Wanted at the link below.
MAKE THE OBVIOUS CHANGES TO THE LINK
http://www.geocities.com/SiliconValley/Monitor/4710/

Mosfets -- Rds (drain-source resistance). IGBTs -- collector-emitter on voltage.

If you double the current going through a mosfet, you will quadruple the power dissipation. P = I^2 Rds

If you double the current going through an IGBT, you won't quadruple the power dissipation. The power dissipation will increase to a lesser extent than in a mosfet because the voltage across an IGBT is not proportional to current.

I oversimplified a little in my previous post with the relation P = IV suggesting that power dissipation in an IGBT rises in direct proportion to current. Collector-emitter voltage would have to remain constant for power dissipation to rise in direct proportion with the current, while in fact the collector-emitter voltage rises somewhat with increasing collector current. Still, the power dissipation at higher currents does increase much less in an IGBT than it does in a mosfet.

Thanks! Now I understand what you were saying.

Unfortunately, I don't think the OP can make use of IGBTs in this case. IGBTs don't become practical until the voltage is

250 or higher, per the information at the International Rectifier url posted earlier. The voltage is way too low to get into the benefit of the portion of the Vce drop curve that is relatively flat. In fact, I have no idea what the curve would look like at 24 volts.

The smallest IGBT Vce drop I have seen in the datasheets is 1.2 volts. Even if it applied at 24 volts, it's 12 watts wasted at

10 amps. An 80 amp 60V P-channel mosfet such as an Infineon SPB80P06P with an RDSon of .023 would drop only .23 volts, and waste only 2.3 watts, and an N-channel such as the STP80NF10 with an RDSon of .012 would drop only .12 volts and waste about 1.2 watts. Those, or equivalent, would be a good choice for the OP if he can get them. He definitely needs *way more* than 10 amp mosfets to accomodate a possible stall condition. As you pointed out, he could parallel mosfets and get lower RDSon and higher current capability.

Ed

Hello Steven, I've now put some stuff on the website (go to the last item on the index page). First thing to look at is the motor you intend to use, as this will determin things like gear ratio of the bicycle. The bike I mentioned previously had a 750W motor. Regards Robert

Hello Steven, I've now put some stuff on the website (go to the last item on the index page). First thing to look at is the motor you intend to use, as this will determin things like gear ratio of the bicycle. The bike I mentioned previously had a 750W motor. Regards Robert

Thank you everyone for input advice. This is all quite overwhelming at the moment---lots to grasp, learn, and consider in what I thought would be a basic application in learning about circuits. Hopefully this thread will be accessible for a while because I think it will take me a while before I'm to the level of comprehending the relevant electrical components and how they work together in a circuit.

Sure, I can easily buy a used bicycle speed controller/drive system as someone suggesteed. There was one available for the equivalent of 30 USD here. However, it was a used part imported from Japan or China. Even if it did work, I wouldn't understand the power consumption, or gain any electrical knowledge for that matter. That's why I've adapted the goal to learn about how the speed control circuit works from the positive terminal and everything in between, the actual path of the current..

There are a few manufacturers of electric bicycles/scooters around here, but unfortunately in Thailand, the market is so limited, so they won't sell individual circuits/parts/controllers as I'm guessing there cautious of other competitors stealing the designs (which they most likely did from Japanase/Chinese makes anyway).

Can someone suggest a particular schematic of a controller / drive system circuit that would be ideal for a 24 volt motor between 250 to 500 watts? If you're confident enough, explain to me the path of electrical current from positive to negative (or the other way around) i.e. from the positive terminal the current comes to a ??k resister which .... then at ??amps ??volts comes to a diode which...etc. etc.

What about this PWM circuit :

Is it sufficient for my application. If not, which values will need to be adjusted?

This would benefit me (and a lot others) greatly, if anyone has the resources / capability to articulate, interprate the controller schematic, particularly it's current path.

I hope this isn't too much, but imagine it's just a short cut request for my self-guided learning...

Cheers This message was sent using the sci.electronics.basics web interface on

I worked up a PWM controller, recently for someone who wanted to drive an 8A 15V peltier cooler. A modification of that could be made into a 24 volt 10 to 20A current regulator for your motor. But it would take a bit of rearranging and component substitution. It has the advantage of using two alternating phases (two PWM pulsers that take turns something like the way a two cylinder engine shares the load between two pistons) that reduce the ripple current drawn from the battery (reducing its self heating) and also reduces the ripple current to the motor, so a lower pulsing frequency can be used for lower switching losses. It also divides the current between twin drivers, so each can be made with smaller parts. I'll think a bit about what changes would be needed.

It stinks. The PWM generator section is okay, but instead of having the throttle just changing the duty cycle (effectively just changing the fraction of the battery voltage the motor sees), I think you need to measure the motor current, somehow, and have the duty cycle automatically hold that current to the value set by the throttle pot. This, in effect, controls the motor torque, which is the force you feel. It also controls the current to a safe value (whatever the full throttle setting is) even if the motor is overloaded and the bike doesn't speed up. This will prevent blown fuses and burnt up motors.

You have to do *some* of the work! If you read the information on the site, you can answer the question "is it sufficient for my application" and the question "If not, which values need to be adjusted?"

The site does a nice job explaining things. If you have specific questions after reading it, ask them.

Ed