Topic open for Discussion. Burnt Out Controller with Regenerative Braking

[TheKid]

This all made sense to me until I was about to install the diode. It occured to me that the regen is built into the controller. If the battery is off, then there is no power to the controller, so pulling the brake lever will do nothing but stop the bike mechanically. Furthermore, the motor acts as a generator whether going forwards or backwards, when no power is feeding into it.
 So if the battery is turned off, it still produces electricity back to the controller. Why aren't all controllers fried when the battery is turned off? I'm assuming there are diodes that prevent this from happening. Which goes back to the regen braking. If there's no battery power to activate the regen, then there's no signal to make the motor reverse itself or stop.
I have a rear 9C motor with a GM magic controller. Regen is activated with the brake lever only with this controller. I propped the wheel up and left the battery off. I operated the pedals by hand, until the wheel was spinning quite fast. I released the pedal and pulled the brake lever for the front brakes, and nothing happened. The rear wheel just kept spinning. I did it again, and pulled the lever for the rear brake. Nothing happened, but the wheel stopped. I put the bike back on the ground and tried using motor only. No problems whatsoever. If there's a problem with the motor frying the controller with the battery off, then there must be a problem with the controller to begin with.
So how would you know if the problem was caused when braking, or while just riding with the battery off? The motor still acts as a generator when it's spinning.

[Gregte]

1. When riding with the battery switch turned off the motor produces a small amount of current.

2. When that current and voltage is sufficient, i.e. wheel speed is fast enough, the controller is now powered, just as if the battery were connected.

3. Now, if you pull the brake to activate regen, the now-powered controller will cause the motor to supply a much larger current and voltage.

4. This larger current and voltage has no load (small load-controller is a load), the battery is missing, and thus the voltage rises.

5. The voltage escalates beyond the rating of the controllers components and they burn out.

All of my experience was with the Golden Motor controller and motor.

The results of applying the brakes with key switch turned off and the bike going fast (didn't look at speedo) were instant and dramatic in my case. Upon returning home I discovered that two FETs were burned out in the shorted mode, in my controller. I replaced them and all was well again. I installed the diode across the key switch and have tried the bike with switch off (open) and the voltage does not rise above 43 volts at the controller or 42.3 volts at the battery.

Install a voltmeter across the power wires but on the controller side of the key switch, not the battery side. Mount the meter on your handlebars. Now go for a ride down a steep hill with the switch OFF and watch your meter from 0 to 35 mph. If it never rises above the normal battery voltage during regenerative braking you don't need a diode and your controller will not likely burn out.

When I do this the meter reads fairly low voltage ( a few volts) until around 10 or 12 mph at which time it instantly rises. This is the speed for me that the controller got enough voltage and current to be normally energized, as though it were connected to the battery. It is now capable of generating full regenerative braking power.

[Gregte]

Another thing to note; I have the Golden Motor motor and controller and it includes a 'fuel gage' on the handle bar. This gage is 3 LEDs that are nothing more than a crude voltmeter.

Anyway, when I merely push the bike at 1/2 mph one of the LEDs will light up. This is with key OFF or even with the battery removed from the bike. This is proof that the motor is generating power to the controller, even at 1/2 mph.

[TheKid]

That's my point. It's not supposed to send current to the controller. You'll get a reading on a voltmeter because the circuit is complete. But the circuit shouldn't be completed by feeding back into the controller. I tried it again, this time actually riding the bike, reaching speeds in excess of 15mph, with the battery turned off. The longest run without having to stop at those speeds was 15 minutes or so. Even when stopping from those speeds, I didn't hear the regen kicking in. I then turned the battery on, and there are no problems. I'm no electronics expert, but I do have years of experience working for electricians. You could spin a generator from now til doomsday, but if there's no load to complete the circuit, no electricity will be generated.
I wondered about this a couple of years ago, and asked on another forum about current flowing to a controller when the battery was off or disconnected. One member experienced in electronics said there were safeguards in the controller to prevent that from happening.  I don't remember exactly what he said, but it was something to the effect of switching off the current between the motor and controller, or disengaging the regen if a problem arises where the current can't get back to the battery. There was also a mention of diodes, but I didn't fully understand what he was saying.

[Gregte]

You could spin a generator from now til doomsday, but if there's no load to complete the circuit, no electricity will be generated.

But there IS a load. The load is small. The load is the controller. Also, this is a permanent magnet DC motor. They will produce a voltage with NO load and a current will flow if there is any load for it to flow into.

. One member experienced in electronics said there were safeguards in the controller to prevent that from happening.

This depends entirely on the particular controller. They are not all made by the same company.

It may very well be that your setup behaves exactly as you say it does. I do not doubt it one bit. But my setup definitely produces power with no battery on the bike. If I pedal it at around 10 or 12 mph with NO battery I can honk the horn and all 3 fuel gage LEDs will lite up. That means that at that speed the controller is getting at least 37 volts from the motor/generator. It has to be 37 volts in order for all 3 LEDs to light. If I then apply the brakes the voltage will rise even higher.

Again I will say this;

1. Remove your battery.
2. Connect a voltmeter to the wires that go to the battery.
3. Ride the bike and watch the voltmeter.

Whatever voltage you read is what the controller is seeing. This obviously assumes your key switch is turned ON unless it is part of the battery you removed.

[Gregte]

I tried it again, this time actually riding the bike, reaching speeds in excess of 15mph, with the battery turned off. The longest run without having to stop at those speeds was 15 minutes or so. Even when stopping from those speeds, I didn't hear the regen kicking in.

You shouldn't get any affects of regen braking without a battery in circuit, or at least it should be very minor if detectable at all. In order for regen braking to slow your bike the regen system has to take on a lot of energy. If there is not battery (a substantial load) there is no place for the energy to go and you will not be slowed. You are driving a generator that has a very small load to drive. It is not hard to turn that generator. But attach a heavy load to that generator (a battery) and it will not turn so easily. It will slow your bike.

[kostelacj]

My two cents, as I have been following this thread. I, too, have a GM motor and controller. The GM motor generates electricity any time it turns, going backward or forward. I have "lit" the power meter LEDs on the throttle and honked my horn. After hearing about what happened to Gregte I quit doing that. If I do more than walk the bike from one parking spot to another I turn the key on, thus completing the circuit to the battery, jic. After monitoring this for some time I went to Radio Shack and got a diode. I put it on. I sleep well.

And, as for regen, it is awesome. Not so much from a power saved perspective as from a "it feels so cool" perspective. When I am hauling a$$ up to a stop light and pull my brakes it feels like the scifi scenes where ships are coming out of hyperspace. I go from near 20mph to NOT fast. Of course,  any observers nearby do a double take. It is most satisfying.

I expect that different combinations of motors and controllers behave similarly or very differently depending on the design, efficiency, quality of the parts and such. As for me, I really enjoy my toy.

John
Bike E Recumbent e- Bike
Driven by a GM 36V 750W 20" rear
Powered by a GM 36V 16Ah LiMn battery
Ridden by a MadMan

[TheKid]

You'll get a voltage reading because the voltmeter completes the circuit. But there should be no current going through the controller with the battery off. It should have safeguards to protect it. With the battery turned off, I get nothing to the LED's on the throttle, so in that regard it's working correctly. However, the regen works for a second, then the wheel keeps turning. At any speed, there is no difference between braking with the battery on or off. It's the same even if I don't connect the motor to the controller at all.
I bought this controller because it had features that the Infineon controller that came with my 9C kit doesn't have, particularly reverse and regen. Reverse woks perfectly, as does cruise control. But I'd really like to be able to stop using regen only, as the motor on my delta trike does. My trikes are quite heavy, and so am I. Since adding the motor to my delta, I haven't had to adjust or replace the disc pads yet. I had to adjust them almost every week before that, and replace them every few months.
My point is that these GM controllers seem to vary from one to another. I just found out my Infineon can be modded so I can get all the same features I wanted in the GM. I ws going to order the Magic Pie when it comes out, but now I think I'll go with another brand for my next trike build.

[Gregte]

After monitoring this for some time I went to Radio Shack and got a diode.

Just be sure the diode is rated to carry the current (unknown by me but must be several amps) and a high enough PIV (again, unknown by me but may see 100 volts, maybe more, just a guess).

And, as for regen, it is awesome. Not so much from a power saved perspective as from a "it feels so cool" perspective.

Ditto!  The power saving is not worth much for my road riding but when trail riding it is. I am always going up a steep hill or down. When going down I most always need to apply brakes, so the regen is doing some real good.

I often stop and enjoy scenery at the top of a hill. I then may jump back on the bike and start coasting down, forgetting to turn on the key. If my speed is not too great the first time I hit the brakes, no problem. But when going fast the diode saves the day (or FETs rather).

[Gregte]

With the battery turned off, I get nothing to the LED's on the throttle, so in that regard it's working correctly.

Where is your key switch?

[TheKid]

The key switch is on the battery pack

[Gregte]

I guess your motor is enough different than my Golden Motor motor to make the difference.

Also, I am using the Golden Motor "Cruise Controller". I assumed you are also since you have the cruise function.

[TheKid]

Yes. Magic controller, 9C motor. But I don't think it's the difference in the motors. It seems to me that there are differences between the controllers. Maybe programming, or maybe components that vary from one controller to another.
I'm learning more about Infineon controllers, since that's the one that came with my kit. I just found a MS Word document that explains how to install jumpers and switches to enable regen, cruise, and reverse. It also explains how to program the regen for different size wheels and/or degree of regen, as well as speed parameters for both forward and reverse.
As I said, I only bought the GM controller because I didn't know my Infineon could be modded with the features that the Magic controller has. But I have no idea how to get the regen to work the way I want it to in the GM, but I now know how to do it with the Infineon. I'll keep the GM as a backup.
I had no problems dealing with GM, they sent everything I ordered very quickly, within 6 days, except for the battery, which was out of stock. But they gave me notification without having to ask them when the battery was shipped. I received it 5 days later, which was a total of 17 days. They also gave me the upgraded battery at no extra charge. I ordered the older battery that didn't slide on a rack.
From what I'm seeing on the Net, it seems there are various problems with the Magic controllers, or none at all, which leads me to believe there are inconsistencies in the manufacturing of these controllers. That makes me leery of ordering the Magic Pie at the early stages.

[Gregte]

Maybe this has nothing to do with your situation but are you aware that the GM controller can be configured to work with non-GM motors, i.e. with motors with different phase angles than the GM motor? The following is a copy/paste from a GM pdf.
-------------------------------------------------------------------------
The controller can work brushless motors with different phase angles other than defaulted 120 degree. Below are procedures to let controller match the motor phases:
 
•Cut-off power to controller and motor

•Grip and hold power brake lever, turn throttle to maximum speed position, depress the cruise control button, turn on the power switch (using battery key)
 
•Connect battery power to the system, release throttle, release brake lever, release cruise control button. 

•Listen 2 beep sound, turn the motor wheel forward and reverse once

•If hear 1 beep sound, motor phase calibrating successfully completed. Otherwise, swapping any two of three motor phase wires, and repeat above steps. 
 

[TheKid]

That's the first thing I did. It calibrated on the first shot. I set the controls on some tubing so I could hold the brake lever, cruise button, and throttle with one hand, while turning the key switch with the other.