My ebike wont go anymore! - Controller problems

[TMaster]

I just got done modding my controller to handle more voltage, I put in 975 ohms for the 82V hot voltage I have. Everything was working fine I put a few miles on the bike, then I was leaving a friends house and there was no throttle. Totally nothing. Fuse is not blown. I put a different throttle on and that one doesnt work either. Before I crack open the controller, could the Hall sensor be bad? If so how do I test it without taking the motor off.

Symptoms:

IF the controller is powered on the motor shutters really bad when I am pedaling.

IF the controller is off the motor is smooth up to about 8mph pedalling, but if I go higher then 9mph the motor starts to shutter.

I remember running out of juice one day earlier this year, and had to pedal all the way back, the motor did not shutter like this at all.

I'm kind of puzzled becuase all I did was ad in 3.3V more voltage to the config, and change out the power resistors from 870 to 975 total.

 

[fechter]

Could be the hall sensors. Toward the end of this thread there is a diagram for a hall tester: http://endless-sphere.com/forums/viewtopic.php?f=16&t=3484
Also hall testing tips here: http://endless-sphere.com/forums/viewtopic.php?f=16&t=3720

You could try disconnecting the hall sensors and see if you get the same behavior. With the halls disconnected, there should not be any shuttering. If there is, then I would suspect a shorted FET somewhere.
FET testing instructions here: http://endless-sphere.com/forums/viewtopic.php?f=16&t=8969&start=45

 

[TMaster]

Thanks fechter, I just did the test on your "When things go bang" thread. I have some results,

The fets, all measure the same, although they measure 3 connections measure 10.01 K ohms exactly, the other three measure 10 and start slowing going up into the 11's. No where near your 1.6 or 1.9M ohms. But I think they are all ok...

The G B Y hall wires are all good( .025v, 4.8v rotating) except, I am not getting 8 to 12V on the Red wire. I am only getting 4.3V. ?????

The LM317T is stepping down the voltage to around 18V then going to the 7805, which is putting out 5.4V

The throttle measures .8V no throttle, and 3.5V full throttle.

One thing I found is the controller will no longer program through the USB. It does nothing when I transmit. Its the 116 chip controller.

I will look through the other threads you posted, and see what I can find. let me know if the 4.3V on the hall is a bad thing.

 

[fechter]

A lot of controllers use 5v on the hall line, but 4.3v seems too low. It sounds like something is overloading the 5v line.

Some of these controllers have a LED inside on the board that will blink status codes. You might need to open up the controller and take a look inside. Many times there is just an end plate you can take off to see the LED.

If you have a milliammeter, you could try measuring the current going into the power wire (the small wire that connects to battery+ to turn on the controller). Normally, this should be under 30mA or so. The 5v regulator will go into limiting at around 100mA. It's real easy to blow up your meter in the mA mode, so be careful not to put the probes across the battery and but the probe back into the volt socket immediately after testing.

You could also try measuring the 5v at the throttle and disconnect the hall sensors to see if it goes back up to 5v. If it does, then you have a bad hall. If not, then something inside the controller probably blew. I can remember one that blew the processor and did something like this.

 

[TMaster]

No LED on the inside, this is the latest infineon 12 fet. I did check the throttle voltage after disconnecting the halls, and it was 4.3V and went up to 4.5V. I also hooked up my old shenzen 15fet controller and the motor runs fine. So I belive I blew something in the infineon controller. I also mentioned that the infineon wont program anymore.

Since I have a Golden motor, I had the yellow and green phase and yellow and green halls swapped for the infineon. I also have been using the throttle that came with the golden motor package, the one with the red on and off button.

I read that some people with the golden motor swap the Blue and yellow phase and halls, Not the Green/yellow. ?????

I also read that someone mentioned that the type of throttle I have doesnt work right with the infineon. Although it does have the standard 3 wires. ?????

I have been running the bike this way for the last 350 miles with no problems.

I checked the LM317T and it is putting out 12.69V now, and the 7805 is putting out 5.0V now.

 

[TMaster]

After talking with Lyen, he suspects the 116 MCU has gone bad because the controller will not program. So he's going to ship me a couple and I will replace it! I hope that fixes the issue!
I have already removed the old 116 chip with my desoldering skillssssss, Luckly I did not lift any of the tiny pads.




I a still a little puzzled on what phase needs to go to what, Or is there more then 1 possible working condition? I see people running yellow and blue swapped, and I see myself and others running the yellow and green swapped, with the same swapped halls. But I have the pics of my old shenzen 15fet, vs the new infineon here before I modified the internals.

Going from left to right, the Shenzen is B, Y, G, inside the controller then going to the Golden motor. The infineon is Y, G, B from left to right inside. Does that mean theoretically it needs to be this way? And what about the Halls?

Infineon To GM Motor
Y - B
G - Y
B - G

 

[fechter]

Nice de-soldering job!

There seems to be no rhyme or reason to the phase and hall wire colors between models of controllers and motors. If it runs without drawing excessive current or skipping, then it's the right combination.

 

[TMaster]

OK thanks! I will report back with results on the chip change.

 

[TMaster]

The bad news is in the controller still wont run, not a peep from the motor. Good news it Will now program though the USB with the MCU change.

So, I tested the 3 gate transistors, and they all acted the same and the voltage went though the right ways doing diode check test with my Fluke meter.

So I found 1 of the 2N5551 transistors was letting the voltage go through it in every combination from my meter probes. The other 2N5551 wasnt doing that, so I thought ok, that one is bad. I then found an M7 diode and an S4 diode in that area that were doing the same thing. So I thought ok, those must be bad too. Well I took them out today, and away from the board they are still good, I probed with with the diode check on the Fluke, and they all are working properly letting the voltage only go one way. So, I'm a little puzzled now. I dont know what else to do now. I can't find anyting wrong inside and nothing burnt either. Could something be bad inside that is making the polarity of those components go both ways? D4, DB5, Q6

Any ideas?

 

[kjparker]

Probably a stupid question, but have you tried antother controller on your wheel? At least that will confirm it to be a controller problem.....

 

[TMaster]

yep, 5th post from the top. I also hooked up my old shenzen 15fet controller and the motor runs fine.

 

[fechter]

What is the 5v line looking like now?

Can you measure the current draw of the keyswitch line?

 

[TMaster]

Tested the Amps between the main power and the controller and its .06 amps, full throttle and no throttle.

the 5V regulator is putting out 4.97v, the throttle red wire hooked up is 4.3v still.

 

[dnmun]

60mA is good for idle, usually around 45mA but that is ok on the big scale.

does the voltage on the hall signal return on the throttle change from .3V-4.7V when you rotate the throttle? it may be grey or green. sometimes the ground is white on the throttle. i just reread and saw you did show different voltages at the throttle return.

curious that the shenzen works and not this one. you have gotta be skilled if you resoldered that mcu without shorting the leads.

 

[TMaster]

Thanks! My eyes are really good up close, and a steady hand! Here is the replacement chip soldered in. I actually did it twice, I was frustrated today and thought, well maybe the first MCU I replaced was bad, so I took it off again and put another one on! LOL. The next one programed fine too. But still no motor go.

 

[fechter]

The current sounds like it's in the ballpark, but possibly a little high. At least nothing on the 5v line is shorted.

It is weird that the throttle only has 4.3v. You could try unplugging the throttle and measure the 5v coming out of the controller to see if it goes up when the throttle is disconnected. Not sure what that would indicate though. As long as the throttle signal can reach 4.0v, it should work normally. It might be related to a safety feature that kills output if a throttle connection fails.

Too bad it doesn't have the status LED. Some boards have a spot for the LED and resistor, but they're not installed. Of course you need to have the documentation that tells you what the codes mean before it's useful.

On similar units, there are a number of faults that will inhibit output. The LED can tell you which one. One is an invalid hall signal configuration. This could happen if a hall wire or sensor is bad. This would also happen if the 60deg/120deg jumper is not set correctly. Most bike motors are 120deg. Out of range or high throttle on startup is another one that can cause shutdown. Under and over voltage can also do it. Make sure you know what the settings are. It's possible the settings may have changed at some point.

 

[TMaster]

The throttle voltage is 4.35v and when i disconnect it it goes to 4.98V

There are no jumpers on the board, I know in the programming you can program the phase degrees, 60, 120 or compatible. Compatible is sopposed to work for both types, and I've always ran it in that mode. I never changed anything. Everytime i program the controller it resets the controller settings to what i have picked.

 

[dnmun]

but that throttle worked on the shenzen controller though. sounds like it could be a bad connection or maybe the leads are in the wrong place in the plug.

 

[TMaster]

I tried another throttle I have and it didn't work with the bad controller, but it did work with the shenzen controller. PLus I know they are working because I measure .8v no throttle, and 3.5v full throttle on the middle wire.

 

[dnmun]

when you used this other throttle on this bad controller, it would not work. correct? but it worked with the shenzen.

the current throttle also worked on the shenzen, but not your bad controller, right?

next you might want to see if the hall signal from the throttle gets to the inside of the controller. maybe a broken connection or a lifted trace or trace was torn or cut when you had it open or maybe the hall signal wire of the throttle is pulled off the spot where it is soldered to the board.

 

[TMaster]

I'm getting .8v to 3.5v right at the board trace where the signal wire is. No throttle, full throttle.

I checked all the resistors in that area and they are fine. One of the 102 (1K) is reading a little low, .77k, but that shouldn't be cause for any alarm I dont think. I wish I had the schematic, so I could trace the routes of things and then see where the problem is. I guess I could do voltage and connectivity checks all over the board and try to see the route manually. ugh!!!

The throttle is sending the voltage to the board, but for some reason nothing is activating inside to tell the controller to power the motor. I may have to dig into the forum here and see if theres any infineon schematics anyone has posted. Also I think I may take the extra wires off the board that I don't use and get them out of the way. I dont use the CA connection, ebrake, or the blue single wire.

A few more days and I think I will need to get a new controller, very sad. I need two ebikes incase one breaks down!

 

[dnmun]

so now, maybe the controller LVC may have shut down the output drivers so maybe next is to check the voltage to the gates on the output mosfets. then you will have an idea of how the mcu is working to drive the output. the LVC is programmed into it i think.

 

[TMaster]

Well I F**K*D it up!!! I measured 11.9V off the legs of the gate transistors, at no throttle and full throttle. I decided to check the voltage on last mosfet at the end of the controller, and I must have slipped the probe and one of the mosfets exploded about 1 foot or so away from my face, I quickly closed my eyes and held my breath. I would say the fireball was about the size of a softball. I think its time to give it up and contact lyen for a new controller. I think I will salvage what I have and sell the good parts or work something out with lyen.

 

[dnmun]

i think fechter will be supplying you with the coveted fried guinea pig award.

 

[fechter]

Ha Ha, that old boy hasn't been out lately! Here you go for your effort:

It's happend to the best of us....

If there wasn't something else wrong to begin with I'd say try fixing it.