[G-man]

Hello all:

I was wondering whether it would be better to save the weight of a tranny and go direct drive to the diff? If so, couldnt one split the batt pack in half and run at half voltage in parallel for low speed, and full voltage in series at high speed? Thoughts?

If not, why dont people use an automatic with a manual valve body?

Best regards,

G-man

[TigerNut]

Originally Posted by G-man

Hello all:

I was wondering whether it would be better to save the weight of a tranny and go direct drive to the diff? If so, couldnt one split the batt pack in half and run at half voltage in parallel for low speed, and full voltage in series at high speed? Thoughts?

If not, why dont people use an automatic with a manual valve body?

Best regards,

G-man

Any PWM based motor controller will effectively transform the high battery pack voltage to a lower average motor voltage that is proportional to the motor speed, while also transforming the lower battery current to a high motor current, within its limits of power handling capability. So there is no real need to add the extra complexity of series/parallel pack switching.

Some people do use automatics with a manual valve body... you can find some examples by searching through the build threads. Other folks are also using standalone electronic controllers to replace the function of the ECU in electrically controlled automatic transmissions. Bypassing the torque converter is also done, although that's more advanced than a lot of people are interested in (or capable of) doing.

[dougingraham]

Originally Posted by G-man

I was wondering whether it would be better to save the weight of a tranny and go direct drive to the diff?

I think this would work find for a car that has a top speed of perhaps 35mph. But if you want to go on the highway your final drive ratio needs to be lower which means you need more torque which means a larger motor which means more weight which means even more torque...

Most of the motors don't have a wide enough RPM range with enough torque to eliminate the transmission. The Tesla Roadster is an example where the motor can turn up to 14000 RPM so they picked a differential ratio such that the top speed is 129mph. But this motor has a lot of torque so low speed performance isn't a problem. Brushed DC motors have RPM limits that are usually set by brush issues. Brushes bouncing and timing issues generally limit the max rpm. AC motors are limited by iron losses and the physical point at which the rotor comes apart.

The manual transmission in my car weighs 72 lbs. I would need a much larger motor or two of the motors I am using now to go to direct drive and be happy with my acceleration if I removed the transmission. The theoretical top speed in each gear in my car is

1st 27
2nd 44
3rd 68
4th 97
5th 135

This is based on a 5500 rpm limit and the factory sized tires. The 4th gear is 1:1 so with a final drive ratio of 3.909 you can sort of see that this would not be an unreasonable top speed. However this is too much load on the motor at low speeds and without forced air cooling to the motor I wouldn't be able to drive around the neighborhood without overheating the motor as the internal fans need some rpm to work. Try putting your car in 4th gear and starting out sometime.

For OEM's where they can provide a motor with a large RPM range and adequate cooling, no transmission is doable. For DIY converters trying to convert a car it is more economical and better performing to use the transmission.

[PStechPaul]

One factor that is always present is that torque is proportional to current, while speed is proportional to voltage. And power is RPM*T/5252. But for the purposes of what we need for driving, it comes down to the speed we want to maintain and how fast we need to accelerate (or how steep a hill we need to climb), and this relates to the mass of the vehicle and the force needed where the rubber meets the road.

Most of the losses occur because of resistance in the circuit, which includes the motor, controller, and wiring. A larger motor will tend to have larger windings and less resistance loss at lower speed and high torque (current), so all things being equal, a larger, heavier motor will provide more torque and greater efficiency. But any motor tends to be more efficient, and can provide more horsepower, at its rated speed.

So a transmission will allow the motor to run at these more efficient conditions, and will be smaller and lighter which reduces the required torque and/or provides more room for batteries. But you must balance that with the size and weight of the tranny. I think manual transmissions are about 95-98% efficient so that is not a major factor.

I do have a question, however, about the clutch. An EV would not really need a clutch because the motor can start from zero RPM. But does a clutch help with shifting gears?

[piotrsko]

Originally Posted by PStechPaul

I do have a question, however, about the clutch. An EV would not really need a clutch because the motor can start from zero RPM. But does a clutch help with shifting gears?

the flywheel and low resistance motor takes about 4 seconds in my vehicle to slow enough to get to the next higher gear. A very low mass flywheel, or none at all still would take time to slow because the armature still has significant mass. The clutch decouples all that motor /flywheel mass from the transmission and allows a normal speed shift beause I don't have to wait for rpm speeds to match to re engage the clutch. Please note that I have a constant mesh trans, synchros might be faster. If this helps the shift, then........

[G-man]

Originally Posted by TigerNut

Any PWM based motor controller will effectively transform the high battery pack voltage to a lower average motor voltage that is proportional to the motor speed, while also transforming the lower battery current to a high motor current, within its limits of power handling capability.

But wouldn't the controller work better with a lower input voltage, when operating the motor at a lower speed range? Wouldn't the battery pack have double the ampacity, and could that provide less sag and allow the motor to develop more torque?

[G-man]

Originally Posted by dougingraham

For DIY converters trying to convert a car it is more economical and better performing to use the transmission.

Thanks Doug, that makes sense. It seems like a diyer should size their motor first, based on the expected donor weight, final drive ratio, and desired top speed, and then pick their pack based on room, required voltage, and range considerations, right?

[TigerNut]

Originally Posted by G-man

But wouldn't the controller work better with a lower input voltage, when operating the motor at a lower speed range? Wouldn't the battery pack have double the ampacity, and could that provide less sag and allow the motor to develop more torque?

For any controller where the input current ripple is substantially handled by the input capacitors, I don't think that running at half voltage will help much. In fact you would have to make the parallel to series (and also in the opposite direction) switch very carefully because the capacitors would have to be precharged to the new voltage level before closing the contactors and re- engaging the drive current.
You'd also have to size all the input wiring on the controller for double the current... And a contactor screwup would short half your battery pack. There is a lot of extra complexity and safety implications with series/parallel switched batteries, for negligible gains.

[dougingraham]

Originally Posted by G-man

Thanks Doug, that makes sense. It seems like a diyer should size their motor first, based on the expected donor weight, final drive ratio, and desired top speed, and then pick their pack based on room, required voltage, and range considerations, right?

There is probably a guide in the Wiki section that can help with selecting these things. The way I did it was to decide on the range I needed first. Turned out this was only 20 miles. But a pack that would provide only 20 miles would not be able to provide the current I wanted to motivate the car in a spirited manner. So I bumpped the pack size up to give me the performance I wanted to see. As it turned out I ended up with an overall system that weighs about the same as the stuff I took out. I ended up with an 80% DOD range of about 60 miles. Performance should be better than the original car. If I was building a race car I would do things differently. I have plenty of room for more batteries but I don't want it to get heavy. Selecting the motor was not a struggle either. I could tell from looking at other projects that a 9 inch motor would give plenty of power if the car is light enough. I could also have used an 11 inch motor but it would more than likely be overkill and added unnecessary weight. The motors that are intended for use in EV's all seem to have similar characteristics. If you scrounge up a motor from a fork lift it might be intended for low voltage use and be inappropriate for use with high voltage. This could limit the power output greatly.

If I was building a street racer I would go for an 11" motor and run A123 pouch cells or Headways because you get more power for the weight although the cost is quite a bit higher. I would size the pack to keep the vehicle light weight but capable of maxing the motor/controller out.

I hope that helps.

[G-man]

Originally Posted by TigerNut

For any controller where the input current ripple is substantially handled by the input capacitors, I don't think that running at half voltage will help much.

You mean sag, right? A DC battery doesn't really have ripple.

Originally Posted by TigerNut

In fact you would have to make the parallel to series (and also in the opposite direction) switch very carefully because the capacitors would have to be precharged to the new voltage level before closing the contactors and re- engaging the drive current. You'd also have to size all the input wiring on the controller for double the current... And a contactor screwup would short half your battery pack.

Let's assume for the sake of investigation we had lockout contactors, and separate controllers, properly sized wiring, etc. Under this (theoretical) scenario, would there be an advantage?