Topic open for Discussion. Burnt Out Controller with Regenerative Braking

[Barry]

My friend was riding his bike with the key off to save battery power so he would have some left for a big hill he had yet to climb. But before he got to that hill he was going down another hill and applied the brakes. He said that the bike came to a stop very quickly presumably due to the regenerative braking but it was much stronger braking than normal.

To make a long story short, his battery fuse burnt out and the wheel would no longer roll easily until he unplugged the motor wires from the controller. Apparently the controller shorted the three motor power wire terminals together when it burnt out. I suppose all the FETs shrt circuited instead of open circuited.

So now his problem is that the battery fuse blows immediately when he turns on the power switch and the brushless motor hub will only spin freely if unhooked from the controller.

Any ideas what can be done to fix this? It appears that the problem was all caused by applying the brakes with the power switch turned off, while going down a steep hill.

If you have any questions I will get him to sign up for this forum and answer them cause this is all the middle man stuff I am gonna do. ;-)

[Gregte]

If you have any questions I will get him to sign up for this forum and answer them cause this is all the middle man stuff I am gonna do. ;-)

I am now registered Barry so I will take over. Thanks
Gregte

[muzza.au]

G'day Greg,

I'm not sure how to fix the problem. Frankly your controller may be fried. As soon as I read your post I seemed to remember that using regen with the power off is a very bad idea as the power that goes to recharge the battery has no where to go. It seems to be similar to an old thread that was on here: http://goldenmotor.com/SMF/index.php?topic=408.0

Muzza

[TheKid]

I was told to disconnect or disable regen when riding with the battery off. The electricity generated by the motor has no place to go if the battery is turned off. With the Magic controllers, it's simple. disconect the brake wires from the controller. I'm not clear as to how to disable regen if it's activated by the throttle.

[Gregte]

Thanks for the input Muzza.

I am pretty confident that what happened to me is much like what was discussed in the thread you referred me to. When I applied my brakes, which have regenerative braking, and my key switch was turned off, thus disconnecting the battery so it could not absorb the power being generated by the motor, the voltage went very high and burned out diodes and/or FETs.

The thing that really bothers me is this: the whole system could be easily designed so that this cannot happen by doing something so innocent as applying your brakes while going down a hill with the key turned off. There could be a high power zener in place and a fuse that would blow in this situation thus saving the electronics.

At the very least, there should be big WARNING stickers on the controller, the key switch and the owners manual that would warn you to never ride the bike with the key turned off because it could cause an electronic failure if the brakes are applied.

I would still like to hear from anybody who might know which components are likely to be the defective ones so I could replace them. I expect this info would have to come from someone who has had this same failure with this same controller (Golden Motor 36v Cruise Controller).

TIA for any help anyone can provide.

[Gregte]

 I did figure out what was wrong and have fixed my controller. It has three pairs of FETs, one each for the yellow, green and blue motor wires. It turns out that both FETs of one of the pairs was defective in the shorted mode, not open mode. I was able to detect this with all components in circuit. I removed the two bad FETs and replaced them with good ones and now it works perfectly.

But, I had to figure out a way so this failure could not happen again. I cannot count on my memory always have the key ON while riding. I screw up too often for that. Situations will come up where I take off coasting down a hill and have not yet turned the key on. Or perhaps a friend will do it.

While riding the bike this afternoon the solution to the problem hit me, and it is so simple that I cannot believe that it had not been implemented from the factory. The solution is to connect a diode of ample current and voltage capacity across the key switch (on/off switch), but in a direction such that battery current cannot flow through it to the controller when the key is OFF. However, current WILL flow through this diode from the controller to the battery, to charge it, with the key OFF and the bike moving and brakes applied.

The only cost of this fix is the cost of buying and hooking up a diode across your off/on switch. I used one diode of a 25 amp full wave rectifier because that was the only device I had that I felt could carry the potential current that the hub motor might generate.

[e-lmer]

I am not sure that does what you think it does.

The problem is not that there is noplace for it to go, the problem
is that the spinning wheel is generating a large reverse EMF
across the FETs. 

The diode won't change that end of the circuit.

I don't know that this is right, it's just how I think it works.

[Gregte]

Well, you will just have to try it as I have. Think about it, when the hub motor charges the battery through its regenerative braking, what is happening is current is flowing to the battery instead of from the battery. This diode allows that to happen. If the diode were turned around the other way then current could flow from the battery which would serve no purpose more than turning the key switch on.

Anyway, like I say, just try it and you will see that it works. I have to be going 12 or 15 mph for it to generate enough power to operate the controller, and also I am using the Golden Motor Cruise Controller and brushless hub motor. Maybe other controller and motor combinations will not work like this, I don't know.

Another way to look at it is like this.... during normal operation current flows from the battery to the controller and then the motor. But during regenerative braking current flow just the opposite. Current does NOT flow from the battery to the controller and also from the controller to the battery at the same time, that is impossible and nonsense to think.

So, since we have established that current can flow from the controller to the battery then it can do it through a diode instead of a key switch. It is DC, not AC and it only flows one direction at a time.... that is, it is either charging or discharging the battery and when it is charging the battery all it needs is the diode as pictured. And we know that it can charge the battery since this model has "regenerative braking".

[TheKid]

Excuse my ignorance, but I'm new to this sort of thing. It seems that in effect, this diode is an electrical version of a flow check valve that's used in plumbing, which allows water to flow in only one direction. Is that correct? Do all diodes act this way, or do we need a certain type? Your solution makes perfect sense to me.

[Gregte]

Yes, your assumption is correct.

Some diodes have more specific functions but they all have the one way flow to some degree. A zener diode will pass current both directions but in one direction the 'zener voltage' must be reached before it will pass current in that direction.

An LED is a Light Emitting Diode whose main purpose is to emit light but it also passes current in one direction only.

Diodes can be configured in a manner to turn AC (alternating current) into DC (direct current). In this capacity they are generally referred to as a rectifier, but they are performing the same task of allowing current to flow in one direction only.

[TheKid]

Thanks. That's a good idea you have. The only time I don't turn the key is when I forget, so I'll install one just in case.

[Brandnew]

I am trying to find a diode as described for my kit. What exactly should I ask for? I tried to find one with a web search but not sure what I need. It should be 25 amps but what voltage? My kit is 36 volts so should the diode be rated for 36 volts? As you can see I am not an electronics person. Will Radio Shack carry these or where should I be looking? Thanks from Lagrange, Georgia USA

[GoldenMotor]

Hi guys, I hope this picture is self explanatory, thanks greg!

[stl_recum]

Has anyone tried this yet,was it sucessful.

[Gregte]

I have tried it and yes, it does work. My method for trying it was to get at the top of a steep hill and go down fast with the key turned OFF. Then I hit the brakes. The voltage went as high as 43 volts (I have a voltmeter permanently mounted on my handlebar) but no damage was done to the controller. I tried it several times with the same result. The battery itself only sees 42.3 volts because there is a 0.7 volt drop going thru a silicon diode.

For the diode, I used a 35 amp, 200 PIV (Peak Inverse Voltage) full wave rectifier. There is no need for the full wave part since that means the device contains four diodes, but it is the simplest and cheapest solution I could find. It cost $1.25. It has male spade terminals. This rectifier is normally used for changing AC to DC but we are just using one diode in it. The others just come along for the ride.

You just use two of the four terminals. The four spade terminals will be the green dots on the picture. Never mind the other 3 diodes, just pick one and wire it according to the picture above.

I got my rectifier at:
http://www.bgmicro.com/index.asp?PageAction=VIEWPROD&ProdID=11660

Note the + marking and the AC marking on the body of the rectifier. They can be used just like the terminals in the picture on this post with the + and AC being the same thing. Use the + and either one of the AC terminals.

If you connect the + and AC terminals to your key switch backwards what will happen is your key switch will not turn off your power. No harm will result but your power will not turn off and your controller will NOT be protected from regen braking with key off.

Conversely, if you connect to two terminals of this rectifier and your key switch does turn the power off then you have connected it correctly. The exception to this would be if you use two spade terminals that are diagonally opposite. If you use one of the possible diagonal combinations your system will work except there will be a slightly higher voltage drop between the controller and battery. Again, no problem but do it the right way just to keep things straight forward and sensible.

If it helps, think of each of the 4 diodes as being one way water valves and water can only flow the direction that the arrow points. That is the direction that the positive electrical current will flow. For the purists, electricity actually flows from neg. to pos. but it makes no difference to think of it as flowing from pos. to neg. This way the arrow used in schematic symbols of diodes and transistors have an intuitive sensibility as to direction of current flow.

ONE LAST THING: if your key switch is in the Negative (-) line of your system instead of the Positive (+) side then your diode needs to point the other direction than the above picture. Just remember what it is you are trying to do; you are trying to allow electrical current to flow from the controller to the battery when the key switch is open (off). And as I said, as a test to see if you are wired correctly all you need to do is see to it that your key switch still turns your power off. If it doesn't then reverse the wires going to the diode.

One more one last thing   If you use too small of a diode it will burn out and not do any good at all. So be sure you know what the rating is of your diode.