I understand that the way to go in terms of voltage and amp control is pulsewidth modulation. That being said, not only is it expensive, but its just not as much fun to order a solution online and have it sent to your door. I like to buid things.....SO.... here goes the question
What exactly is the benefit of puse width modulation over more traditional current and voltage limiting circuits, such as potentiometers and resistors in parallel etc. I know it would be very hard to find a potentiometer in the size required, but could you do a sort of electrical transmision? say a set up for half voltage, three quarters, and full, which you could manually control? The big problem I can forsee with this is that perhaps finding resistors with that kind of power rating would be hard, and that too much power would be eaten by the resistors. But again, I dont really know these things. If someone has a good deal of electrical knowlege, please enlighten me, as not only do I want awnsers, but I want to learn the why's as well.
Also, has anybody attempted to build a pulse width modulation circuit? How did you find out what its rating was? where did you find the schematics (I found many, but for low power input and output). My latest attemt is to learn to use computer programs that simulate circuits to build a pulse width modulation for low power, and tinker with the parts to see what I can come up with. If anybody knows of any good programs that wont set me back too much, and are user friendly, I would greatly appreciate that as well.
Thanks for your time guys, and happy ebiking.
Tortuga,
at the EAA-SV meeting last month one of the founding members of the EAA gave a presentation on the history of electric vehicles. He had personally designed and hand built controllers for his first few EVs. He had a very clever series/parallel switching controller that used a diode network to switch from a standstill up to full speed in six steps. He said that for all but two steps the pack was balanced. This might be the way to go for a low cost, time intensive home made controller. I can look up his contact info if you are interested.
-Crusher300
Silver EVT 4000e (60 volt) San Mateo, CA
-Crusher300
Silver EVT 4000e (60 volt) San Mateo, CA
That would be great. Thank you very much for your offer. If nothig else it would be a great framework to either build on, or more accurately build down, lol. Thanks for your time.
Wont catch me ridin dirty
http://www.austinev.org/evalbum/518
http://www.geocities.com/thomassonmj/electric_drive.html
An old Taylor Dunn golf cart I have had a very large resistor bank with something similar to a motor commutator for high-current switching between resistors or full on. It was great because there was no current limit. After modifying the setup and putting a controller in, the performance went down a lot.
An ebay search for taylor dunn might turn up something.
---
Avatar taken from http://www.electricmotorbike.org/
Anyone got one they might want to sell?
My KZ750 Project: here E.T.A. 1 mo
[url=/forum-topic/motorcycles-and-large-scooters/587-my-kz750-electric-motorcycle-project]KZ750 Motorcycle Conversion[/url]
[url=/forum-topic/motorcycles-and-large-scooters/588-fixing-my-chinese-scooter]900 watt scooter[/url]
Pic from http://www.electri
Pro-Tinker,
I built a series of PWM controllers for my first Ebike (about 5 or 6 years ago)...
It was a frustrating experience that went through 4 majour revisions and well over $150 in parts that were fried as it evolved. After about 4 months of weekends & evenings I had a 24V 70Amp nominal controller that could smoke a motor, and stay intact.
The main issues were sifting through the Rated specs of the MOSFet's and determining what was realistic vs what the data sheet said. Two of the 4 revisions were just to learn how important it is to clamp the back EMF of the motor as the commutator rotates past the brushes. The key term to research is "Flywheel Diodes".
I still have the working controller, but really can't be bothered to use it as I never got to the point of implementing any sort of current limiting.
For now I've been buying controllers & parts from tncscooters.com for a fraction of what it would cost me to build a similar unit, nevermind the time to build. I've used thier 36V/40Amp units with no grief, and just recieved a 48V/50A controller with several very nice features like, Lock, Charge, Brake Disable that would add even more complexity to a homebrew.
I still would like to revisit the idea of a scratch-built controller, though it would be in the 48-60V 100-200Amp range, as that seems to be the area that store-bought gets expensive. I've been playing with the PicAxe 08M controller just to get a feel for what thier all about and think that micro-controllers are the way to go, rather than a lot of analog discrete components.
I don't think my comments are very encouraging, but are a reflection of my experience.
FrankG
www.theworkshop.ca
FrankG
www.theworkshop.ca
A resistor limits the current by wasting a large percentage of the power in the circuit. A PWM, on the other hand, minimizes wasted power and runs much cooler than a resistor as a result. I've built several PWM circuits and modified many controllers. I also find it much easier to modify an existing crappy design than to build the whole thing from scratch. Brushed controllers are fairly easy. A PIC would be great, as long as you can figure out the code. That's what the Goped ESR750 uses.
does a voltage deviding circuit work the same way? that is spending residual energy on alternate forms of energy (i.e. heat)?
Wont catch me ridin dirty
Yes, a voltage divider is a resistor network; it wastes energy (turns it into heat) just as any resistor would. One could use a multi-pole switch, adding batteries to the string as required. However, what happens in that case is that the batteries in the front part of the string get exhausted before the others. This is the point that was made earlier, about keeping the batteries balanced.
There are a lot of microcontrollers out there; some are designed specifically for motor control. The latter include full PWM timers, and up to 6 outputs (you don't have to use all of them) so they can control a brushed motor in a simple configuration, full 4 quadrant brushed regenerative braking, or a brushless motor (also called BLDC, PMAC, etc.). You might look at Motorola and Atmel in addition to PIC.
Patrick
Before I'd ever heard of EV's, I built one with mechanical PWM control!
When I was around 15, I ran across plans for an electric RC glider winch. It used a mechanical "pulser" using the contacts from a car starter solenoid. The drive motor mount was hinged and attached to the contacts. The "throttle" pulled the contacts closed against a spring, and the counter-torque from the motor tried to open the contacts. The result was a sparking stutter, with longer pulses when you pressed harder on the throttle.
I installed it in a "soap box derby" glider, and had fun frying starter motors, contacts, and junk batteries. Range wasn't good, but acceleration was impressive.
Mark
Okay, so here is another aproach I scheemed a little while back
can I create a class A amplifier with a darlington pair (of transistors) so that the amp with the main mosfet is also the driver circuit? I know that class a amps are very inefficient as they use 50 percent power (under ideal conditions) when they are using no signal, 0 volts. This is because they are biased at point q on the load line graph so that a sine wave can be used (mostly for audio aplications) This way both negative and positive signals fall within the range of the transistor. Since a PWM signal is only dc, not ac, we dont have to wory about clipping off the bottom of the wave. So can I bias the main mosfet bellow cutoff point, and have its max around the saturation point? If this is correct, the class a amp would just be more of a big on off switch and not use the 50 percent power, and be much more efficient right?
I have read that darlington pairs slow the mosfet response way down, if this is the case, is it too slow for pwm?
Wont catch me ridin dirty
In a PWM, you want the FETs to switch as fast as possible. When they're off, there's no power dissipated. When they're on, the dissipation is a function of the current and the on resistance of the switch, which should be fairly low. Half way in between on and off, the dissipation is sky high, so you don't want the FETs to spend much time there. I would recommend a dedicated gate driver IC to drive the gates. Something like an IR2101 is pretty common.