[hondo]

OK guys, I had this idea and I would like to know if it might work or if it is not possible. I understand the power end of a controller with the mosfets, free wheeling diodes and capacitors but, is it possible to drive the mosfets with a mechanical switch? What I was thinking was kind of like a set of points on an old distributor. A v8 motor that was running at 5000 rpm opened and closed the points 20,000 times per minute (5,000rpm = 2,500 rpm distributor shaft x 8 cylinders = 20,000). Is this fast enough for the mosfets? If this would work, you could construct a multi-lobed cam driven by a small dc motor with a movable switch that varied the duration (dwell) time. Any thoughts?

Hondo

[dimitri]

Or you could attach a hamster running inside of a wheel to switch MOSFETs and depending on how much you zap the hamster in the butt, he will run faster and this would control switching frequency.

These ideas remind me of "crazy contraption" project that my 10 year old kid gets in 5th grade, but even 10 year olds know its not for real...

sorry, I don't mean to offend you, but it just makes no sense at all....

[dimitri]

I felt bad for making fun of you and not explaining why your idea makes no sense.

Mechanical distributor would distribute control pulses among many MOSFETs, one at a time, which is not what you need in PWM controller. You need all MOSFETs to switch on and off at same exact time, that is the real issue. Even $1 IC 555 timer can do a better job at creating control pulses than your mechanical device, that's the easy part.

The hard part is how to syncronize 2 dozed FETs to the point of few microseconds, because that is how long it takes to burn one if it switched on out of sync from all others.

Again, I am sorry if I have offended you with my first reply, I didn't mean it

[hondo]

Thanks for your reply Demitri. As you can tell, I am just learning about electronics, and I tend to use my mechanical experience instead of learning what I should. I have been following Paul's open source controller thread over on the ecomodder forum and Paul uses one driver to drive 10 mosfets. Thats what I would be doing with the points. A single switch would turn all the mosfets on or off at the same time. This is where I am ignorant. What would be the difference between my mechanical switch and Paul's driver? I just want to get rid of my resistive/voltage control I have been using in my Fiero for the last year without spending $1500 for a "real" controller.

Hondo

[dimitri]

I think if you have comfort level to build the power portion of Paul's controller, then you should be able to build the whole thing, without any modification. Then you just spend money on parts and that is much much less then $1500.

I followed Paul's progress as well and I would be able to build a copy of his controller, but I just don't trust MOSFETs driving my car, its just a matter of time before they will fail. Afterall, Paul's controller has not passed the test of time yet, has it? I think its a great design for scooter or bike, but not for 3000 lb EV, there are just different power levels involved, which greatly increases risk of burning MOSFETs, since you need more of them to keep in sync.

If there was a DIY design similar to Paul's of single IGBT based controller, I would probably try it out, just to see if I can do it. It would be a good excersise while I am waiting for my BMF-750

[dimitri]

Originally Posted by hondo

What would be the difference between my mechanical switch and Paul's driver?

Hondo

The difference is that in mechanical switch you can't control the width of the pulse, only the frequency of it. PWM works by changing the width of the pulse relative to the width of the gap between pulses, which controls average amount of power the motor is seeing. The frequency in PWM does not need to change ( for the simplicity of this discussion, in reality there are reasons to change the frequency in some situations ).

To draw a parallel to mechanical distributor, you would have to make the contacts inside of distributor wider or narrower, depending on the throttle position, but it would be spinning at same speed.

Hope this helps

[hondo]

[quote=dimitri;115899]The difference is that in mechanical switch you can't control the width of the pulse, only the frequency of it.

Isn't the width of the pulse how long the switch is on verses off. This can be done by moving a spring loaded switch closer to or farther away from the cam the same way we used to adjust dwell in the old distributors. If the switch was away from the lobes- no juice. As soon as you move it to where it first touches the lobes, you would have a very short "on" time and a long "off" time. The farther you move it in, the longer the "on" time and the shorter the "off" time.

PS Thanks for your resonses, this is how we learn.

Hondo