Another hare-brained scheme

[bystander]

This problem doesn't affect me much, but I have been wondering if something can be done about it.

Because of the limited regenerative charging capability of the PT, many users find themselves in reduced performance mode when attempting to go downhill at a good clip.

The reasons for this, as far as I can tell, is to protect the electronics, motor windings and batteries from abuse.

If one had beefier motor windings, more (or higher capacity) FETS driving the motors, and larger capacity battery packs, the problem would take care of itself. This would require a product redesign, and may not happen until sales pick up in a big way.

What I was wondering is, instead of tackling the problem from the inside, what if an external element could be added to the PT to bleed off this extra energy?

Does anyone remember this old post?

http://forums.segwaychat.com/showthread.php?t=13499

Some of you may recall seeing these in action at the 2006 Segfest.



Basically, they are mini-generators directly hooked up to LEDs that blink. When the little wheels are in contact with my PT's wheels and allowed to spin, they light up.

These generators only dissipate a few watts at most. And they're always "on". I find when I use them, my range goes down a few percentage points. This is because they are draining some of the kinetic energy of my turning wheels and converting it to heat and light.

But what if someone came up with something with larger capacity, say 1/2 hp (about 375 watts) or so?

A small wheel, spring loaded against the PT's wheel. The small wheel is connected to an electric motor, say a cordless drill. Instead of having a battery in the drill, a bank of high wattage resistors would dissipate the applied power.

To make things compact, a small fan could be added to the resistor bank to allow higher dissipation in a smaller volume. Sure, an accessory battery could be charged instead, but that's more extra weight and more overcharging to worry about.

The resistor bank would be switched in circuit by a variable amount, perhaps a PWM signal. And there would be some means of determining the duty cycle of the signal via some control method.

The tricky part is to have it automatically engage when the PT is undergoing regenerative braking above a certain limit. And to be completely disengaged when the PT is drawing power from the batteries.

The idea is to get this electronic brake to slow the PT enough to prevent the built-in speed limiting to take effect.

The PT's speed would be retarded when these are engaged, but hopefully by less than the built-in limiting would do. The challenge would be finding ways to engage and disengage them gently at the appropriate times to avoid locking the wheels and faceplanting. Also some means of emergency disengagement to avoid overbraking one wheel compared to the other. (in order to avoid sudden spins which would probably throw the rider)

Haven't quite worked out all the details yet. I still haven't convinced myself it's a practical idea.

Thoughts?

[cmonkey]

I don't think it'll work, as it's still an external kinetic braking system. even if it's controlled by a computer, it's still in response to the Segway's control actions and would cause it to use more energy.

[Five-Flags]

Originally Posted by cmonkey

... and would cause it to use more energy.

If I understand bystander's idea correctly, that's what he wants to accomplish -- to dissipate the energy that regenerative braking is producing, rather than having the Segway limit performance to keep from producing it.

This circuitry would only be enabled when the batteries are at or near full-charge -- otherwise the energy would all go into recharging.

[bystander]

Originally Posted by Five-Flags

If I understand bystander's idea correctly, that's what he wants to accomplish -- to dissipate the energy that regenerative braking is producing, rather than having the Segway limit performance to keep from producing it.

This circuitry would only be enabled when the batteries are at or near full-charge -- otherwise the energy would all go into recharging.

Well, what I was considering was that the PT seems to have a limited threshold for regenerative charging of about 1/2 hp. Flat face seems to appear even with the batteries only part-way charged, if going fast enough down a steep enough slope.

For example, if a PT+rider+cargo that weighs 275 lbs, and an attempt is made to go down a 5° slope, the math says that 1/2 hp must be dissipated when traveling at 8 mph. To go faster would cause more power to be absorbed. Since the system can only absorb so much, the flat face occurs.

By shunting a fraction of the kinetic energy of the turning wheels to a resistor bank, I am proposing that the net 1/2 hp limit would be extended.

It wouldn't be a sharp on/off. The amount of additional resistance would be feathered in and out. What I am proposing would be some kind of sensor that measures how many watts of power are flowing _into_ the batteries. Perhaps a magnetic sensor that would measure how much current is flowing through the PCB traces where the battery connects. I suspect such a sensor may be impractical.

Anyway, _if_ such a sensor _did_ work, when there is say, 50 watts going into the batteries, the external resistor loading would be set low, say about 10 watts. As more wattage was detected going into the battery, proportionally more wattage would be diverted to the external resistor packs.

So at 200 w going into the battery, there would be 100 w going into the resistors. And at 300 w to the batt, 300 w would be sent to the resistors. The rate at which the device would respond to changes would have to be similar to the rate the PT operates, 100 times a second (or so).

These actual wattage amounts would be tweaked for best results if an actual device were built and tested.

If it ever lets the regenerative wattage get up to the 1/2 hp (375 w), flat face would occur, and that's what we're trying to avoid.

The end result would be a PT that has poor regenerative charging (due to being stepped on slightly by the added resistive device), but hopefully allows higher speed on descents.