MAC motor suddenly stopped working

I have been using my MAC 8T with cellmans 48V A123 battery and the 9 fet 3077 controller on my folding bike for two months without problems. However, it just stopped working today when I was going home from work. When I twist the throttle now, it makes a strange sound, see the video below.

https://vimeo.com/49707996

Initially I thought the gears might have been worn out, but I opened the motor, and the gears were just fine. Now I'm totally clueless what might be the problem. Anyone has any ideas?

Cheers

Gears stripped ? has to be that or the clutch id imagine. Gears are pretty easy to replace at least
and quite a few people sell them.

usually you'd have some type of indicator though when it happens so weird nothing sticks out.

The gears are fine, I opened the motor and checked. I also don't think the clutch is broken, as the freewheel works fine. Or could the clutch be broken without affecting the freewheel?

efold said:

The gears are fine, I opened the motor and checked. I also don't think the clutch is broken, as the freewheel works fine. Or could the clutch be broken without affecting the freewheel?

Click to expand...

Exactly the same thing happened to my 10T recently.
One of my halls went bad. Not sure yours has the same problem, but worth checking halls.
Check the connector block first to make sure a pin hasn't backed out of the socket.
If that's all ok and you still have a problem, then test the halls.
If you are unsure on how to do this, then I will explain how I tested mine :wink:

P.S. Also double check the phase connectors. They look like Andersons, but appear to be clones. They can melt with heavy current.
Worth replacing with either genuine Andersons, or a suitable robust connector if you discover any damage.

I checked all the connectors, and they seem fine. When checking the halls, I know I have to apply a voltage of at least 5V over the red and black wires, and check for a voltage change on the other wires. But over which other cable do I measure the voltage? The red or the black?

I also noticed that the phase wires become hot very quickly when I twist the throttle, even though the motor is hardly running and there is no load.

efold said:

I checked all the connectors, and they seem fine. When checking the halls, I know I have to apply a voltage of at least 5V over the red and black wires, and check for a voltage change on the other wires. But over which other cable do I measure the voltage? The red or the black?

I also noticed that the phase wires become hot very quickly when I twist the throttle, even though the motor is hardly running and there is no load.

Click to expand...

Check that you can pull the phase wire connectors apart. One of mine had obviously overheated, coz the plastic cover had melted.
The other two looked fine, but the contacts had damage inside.
I decided to replace them all with original Andersons.

This is how I checked my halls, using a multimeter with a diode/continuity function. Mine beeps to indicate continuity.
This is carried out without the phase wire or hall connector block connected on motor and controller.
I had a spare length of 2 core red/black power cable handy, so I connected the controller hall sensor power supply to the red/black wires on the motor hall connector block.
My controller provides just over 4.5 volts to power the sensors.
Then using my multimeter, I connected the black/neg probe to the black/neg on the block and the red/pos probe to the yellow on the hall block.
Then turning the wheel very slowly, the multimeter should beep on and off as the corresponding hall sensor switches on and off.
Make sure you turn the wheel the right way, as the wrong way just freewheels and doesn't turn the motor.
Do this for the green and blue too.
If any of them doesn't beep, then it could indicate a bad hall, or bad wiring to that particular sensor.

Sounds like a broken one way bearing to me, does it freewheel forwards and backwards ok? easy way to check if its the halls or the clutch is to take the motor apart and run it so you can see the gears spinning you should be able to tell straight away. You can of course use an ebike tester or a multimeter to test the halls are firing, the red and black wires should have 5V across them and the other cable cores should pulse at 5V (between the black negative wire and the colours) when the motor turns.

The freewheel going doesnt account for the phase wires getting hot though so it may be halls or you may be a phase down on the controller, try spinning the wheel forward and then try the throttle, does it still do it? you can also measure the no load amps, simply connect your multimeter to read amps and put it inline with one of the battery leads, you should see only 2.5 or so amps when powering the motor, any more and its indicative of controller or hall issues.

If it does turn out to the freehweel then you can check my thread here on how to change and diagnose them.

http://endless-sphere.com/forums/viewtopic.php?f=2&t=38710&hilit=ilia+one+way+bearing

If the phase wires are getting hot, its a hall sensor problem, which also makes exactly that noise. You have to test them. If you're lucky, you can get your voltmeter probes on the connector while it's connected. Check between the black and each of blue, yellow and green while someone turns the wheel slowly with it switched on. You should be able to see the 5v going on and off. You might want to check that you've got 5v between the red and black before you start.

A tester is an excellent bit of gear for this type of problem....it does sound like a hall but you'll need to test first.

e.g.

http://www.ebay.com.au/itm/24V-36V-...588?pt=LH_DefaultDomain_0&hash=item4163f46e14

Thanks for the helo with diagnosing my MAC. There seems to be an issue with one of the halls. When measuring the voltage over the blue and black wires, the voltage is constant at 5V, while measuring over the other hall sensors produces an alternating voltage between 0 and 5V, as expected.

Now I'll just have to figure out how to replace the hall sensors. Will I have to remove both clutches to get to the hall sensors?

efold said:

Now I'll just have to figure out how to replace the hall sensors. Will I have to remove both clutches to get to the hall sensors?

Click to expand...

It's one serious PITA, unless you have the skill and patience of a surgeon.
The pcb that the sensors are attached to is epoxyed to high heaven, and the actual sensors are stuck in fast too.
It requires a fair bit of dismantling of the motor assembly also.
I tried it recently, and it's unlikely that you will achieve a sensor replacement without destroying either sensors or the pcb.
Even Cellman says it's highly unlikely it can be done successfully.
Best bet is to order a replacement pcb and sensors from Cellman, which would make the job a lot easier.

Just out of interest, do you think that sensor failed because of overheating?

At least the 2 8t mac's I messed with look like below.



In my experience there was so much epoxy on the thing I couldn't get the old one's out. I got a
spare board from someone and used that.

Just out of interest, do you think that sensor failed because of overheating?

Click to expand...

I don't think so. I wasn't driving very aggressively, and the motor didn't fail while driving.

I've emailed Paul about my problem, and that it's likely a failed hall sensor. Waiting to see what he says.

If you can get him to send you one of those boards it's a lot easier to fix. Just take lots of pics as you take the motor apart
so you can get all the little parts back together the same.

Thanks for the advice! I'll make a thorough documentation with lots of pictures, and if they turn out to be useful I'll post them here.

For the benefit of Efold and anyone reading this thread, when your motor runs rough, quits running or makes noise and runs hot, the first thing you should check is the hall sensors. It is 90% of the motor problems. Someone posted a link to the tester, Lyen also has them locally, the labels on his are Chinese but he adds the english. Everyone with a sensored motor should have this in their tool kit. There are many posts on how to test sensors but I will give the run down and save you from searching.

How to test Sensors:
First, it is useful to think of the mysterious hall sensor simply as a NPN transistor. Instead of a base lead to turn it on, a magnetic field turns it on. The collector of an NPN floats when not connected so it is always pulled high with a resistor. It drives to ground and sinks current when on. So there are three leads +power, ground, and the collector output. Note that since the output floats, it will measure as an open circuit unless it is pulled up by a resistor. The controller has the required resistors internally, and supplies +5V to power the sensors. Therefore, it is easy to test the sensors with the controller connected and turned on. The only issue is probing the connections on the connector without shorting out the probes. This circuit is ground referenced so you always measure from ground (black wire on meter and connector) to the sensor output. To measure without the controller connected, you need to use a resistor to pullup the sensor lines, one at a time. The supply voltage, in this case +5V, is used. Simply connect the sensor lead to a 10K ohm resistor, then to the 5V lead on the connector. Any resistor from 2K to 100K will work fine (available at Radio Shack). You measure at the sensor lead to ground. While holding the probes, slowly turn the wheel, it should go from +5V to ground. A DD motor can rotate either direction, a geared motor has to be turned backwards. The sensor will operate up to 15V so you don't have to use 5V, you could use a little 9V battery or a 12V, whatever you have handy. If it is stuck high or low, or 2-3V, then you have a bad sensor. Obviously you have to check for broken wires which would measure 0V. If you can't determine if there is an open or a stuck low condition, simply measure from the sensor output to the +5V (or other supply voltage) which will read the supply voltage for a stuck low and zero for an open.

If you don't have a meter and are too cheap to buy one, you can use a 25 cent LED instead. You just need one more resistor, about 500-1000 ohms. Solder the resistor in series with the LED and use that instead of the meter. The difference is that you now go from +5V to the sensor output instead of ground. First, connect to ground and +5V (with resistor connected) to determine which is the plus and minus side of the LED. One way it will light up, the reverse will not. So now when the sensor turns on, it will supply ground to the LED, turning it on. Leave the pullup resistor mentioned above connected while testing.

Replacing sensors in MAC motors:
You don't have to replace the PCB, but you have to be handy at soldering. First, follow the colored wire for the bad sensor up to the PCB and locate the correct traces leading to the sensor. Carefully cut or chip away the epoxy above where the pins are soldered in. There maybe some of this glue under the PCB where the sensor is, if so this has to be cut loose so you can partially lift the PCB. Clip the leads close to the sensor so the legs are hanging down under the PCB. Unsolder the legs, pulling them out with the needle nose plyers. Use a solder sucker or solder braid to clear the pin holes. Next remove the old sensor. This can be a little tricky, so some patience is required. It's clued in so you have to force it out by tapping on it with a small screw driver or pointed tool. Don't worry about breaking it into pieces, just clear the side channels that the sensor case slides into. Be careful not to damage the layers of metal around where the sensor goes, or scrape the lamination off the coils. Put the new sensor in, it only fits correctly one way. While inserting it into the slot, first insert the three wires into the PCB. You have to bend the leads at a 45 deg angle and lift that section of the PCB while inserting the leads. Next glue the sensor in position with a drop of super glue. After that dries, solder the leads to the PCB and trim the leads. Test the sensor, then re-assemble the motor.

Any competent motor supplier can provide this service. Our charge is $60 plus shipping.

$60 + shipping? we used to buy the motors for less than that :lol: its easy to change hall sensors on these motors but I disagree its 90% of the problem, the gears and the one way bearing are more troublesome from my long experience with these motors.

I spend an hour posting a very detailed explanation on how to repair a sensor that will help a lot of people. And you reply with a stupid nitwit comment. You disagree all you want. You don't know what the hell you are talking about.

HighTekBikes said:

I spend an hour posting a very detailed explanation on how to repair a sensor that will help a lot of people. And you reply with a stupid nitwit comment. You disagree all you want. You don't know what the hell you are talking about.

Click to expand...

You have made 26 posts on here and I don't know what the hell I am talking about? You need to do your research, myself along with mark higgon at team hybrid were amongst the first to get hold of and test this type of motor BMC/MAC, 8 years ago, that's where the name Puma came from, go look it up, stripping gears and one way bearing failure has always been more of a problem, if you are seeing more motors with hall sensor failures then they are shipping motors to you with low grade sensors in and you are passing them on to your customers.

If you ask around these parts you will find I also spend a lot of time, years and years not just on here but on the old v is for voltage forum and way back on power assist on Yahoo helping loads and loads of folks with problems (not charging them a dime) if you check google video and youtube you will find a lot of my videos that I have been making for over 10 years promoting ebikes and their use, the electric BMX video had over 3/4 million hits which helped sell and promote 1000s of these motors.

You spent an hour detailing a hall replacement? here you go 10 seconds,buy ebike tester, remove side cover, locate bad hall, replace bad halls if not all halls with quality honeywell sensors, replace side cover there you go, dont make it sound so hard then go offer to replace them to folks for 60 bucks + shipping .

I have talked with Paul, and he will send me replacement hall and PCB for free. I'll post further updates as soon as the package arrives.

Knoxie, I don't give a damn who you are, you sound like a twit. The purpose of the forum is to help people with problems and share technical info. Your time must be worth nothing. Go stroke your ego somewhere else if you are not going to contribute something useful.

Efold, I would suggest trying to replace the bad sensor while you are waiting. It might be easier than replacing the whole board.

The first mac i tried to replace the hall sensor on that board was a pita. There was so much epoxy on the face of the board
it was almost impossible to chip it out without messing up the traces itself.

I am sure it has something to do with my own stupidity , but it was not until someone sent me a board w/no epoxy nor halls that
I could fix it.

The mac i am referencing was from cellman.

Heat the pcb up with a heat gun (maybe a hair drier on max), the epoxy gets soft and you can pretty much peel it off in one bit. Been there, done that. 8)

Figured there was a better way - I even have a heat gun should have used it. Thanks for the tip.

Efold, I would suggest trying to replace the bad sensor while you are waiting. It might be easier than replacing the whole board.

Click to expand...

I would, if I had the damn sensor. So it seems I have to wait anyway.