MAC 8T

Kirill

1 W
Joined
Dec 10, 2019
Messages
58
Hi everyone!

I’m got a problem with my MAC 8T. After replacing the motor cable, the wheel motor spins very slowly about 7(seven) mph. I’ve checked many times everything, reassemble wheel motor, cooper winding looks good, no traces of burning, motor no twitching, controllet is fine, no burning smell, throttle is fine. I'm changed controller and throttle but nothing changes.
Kelly controler settings stay the same.

How fix it?
 
Kirill said:
Hi everyone!

I’m got a problem with my MAC 8T. After replacing the motor cable, the wheel motor spins very slowly about 7(seven) mph. I’ve checked many times everything, reassemble wheel motor, cooper winding looks good, no traces of burning, motor no twitching, controllet is fine, no burning smell, throttle is fine. I'm changed controller and throttle but nothing changes.
Kelly controler settings stay the same.

How fix it?
Depends on the cause.

First, did it ever work correctly?

If it did, then what happened to cause you to have to replace the motor cable? (what specific problem did you have before replacing it?)

When rewiring it, did the wiring pattern from controller phase and hall wires to the internal motor phase windings and hall sensors remain in the same order? If not, then you may have to use the Kelly setup program to relearn them, or manually reorder them to match the previous wiring order.

If it did not ever work correctly, then what did it do before replacing the cable, if that is different than it is now?
 
amberwolf said:
Depends on the cause.

First, did it ever work correctly?

If it did, then what happened to cause you to have to replace the motor cable? (what specific problem did you have before replacing it?)

When rewiring it, did the wiring pattern from controller phase and hall wires to the internal motor phase windings and hall sensors remain in the same order? If not, then you may have to use the Kelly setup program to relearn them, or manually reorder them to match the previous wiring order.

If it did not ever work correctly, then what did it do before replacing the cable, if that is different than it is now?

Everything works right. Never had a problems. My acquaintance with this motor began immediately with the replacement of the cable, since I decided to use a detachable cable. I had to learn a lot of information about this motor without having driven no one miles. But I replaced the cable safely and rode about one year. In any weather at winter and summer.

Yes, of course, if necessary, I did auto-identification of phases.

The reason for second replacing the cable was a break in the hall cables. Thin hall cables are trapped between very hard phase cables.

And I have a 19 Ah / 48 V battery and a simple Kelly controller.
 
Kirill said:
Everything works right. Never had a problems. My acquaintance with this motor began immediately with the replacement of the cable, since I decided to use a detachable cable. I had to learn a lot of information about this motor without having driven no one miles. But I replaced the cable safely and rode about one year. In any weather at winter and summer.

The reason for second replacing the cable was a break in the hall cables. Thin hall cables are trapped between very hard phase cables.

I'm a little confused. Let me state what I get out of what you said, and you tell me if I misunderstood anything:

"Never had a problems" means something different to me; it means there has never been any problem of any kind with the system at all, until this moment.

However, it sounds like the cable has been replaced twice already:
First time, when you first got it (before using it?), so you could have a "detachable cable" (not sure what this means).

Then it worked fine for a year.

Then a hall wire broke inside the motor cable (which to me *is* a problem), and so you have now replaced the cable a second time.

This time, the cable replacement didn't work as it did the first time, and you are now only able to ride up to 7mph, and that is as fast as it will go.

Controller itself has been replaced with a new one (same model as before?) with the same settings as before.

Phase/hall identification (self learn) has been successfully performed on the new controller.

Is that all correct?


If it is, it sounds like one of these possible problems exists now:

--New cable has a problem (probably another hall wire).
--Connection failure from new cable to something in the motor, or something in the connector to the controller end.
--Hall sensor failure inside the motor (causing controller to fall back to a sensorless mode, if it has one, that doesn't allow full speed operation).

However, I would expect all of those should cause the phase/hall identification to fail.
 
amberwolf said:
Is that all correct?


If it is, it sounds like one of these possible problems exists now:

--New cable has a problem (probably another hall wire).
--Connection failure from new cable to something in the motor, or something in the connector to the controller end.
--Hall sensor failure inside the motor (causing controller to fall back to a sensorless mode, if it has one, that doesn't allow full speed operation).

However, I would expect all of those should cause the phase/hall identification to fail.

You got it right. Just one correction the controller remains old. Settings as they were.
 
The Kelly controller is pretty smart thing. You cannot fool him. A green diode there means no problem. That's just the breakage of the hall wires I discovered with the help of the light of the red diode.
Hall cables are connected. The order is not important. And the phases wires are identified automatically. It is impossible to connect incorrectly. Even in a dark. Just connect thick cable with thick, and thin with thin. Moreover, it worked 5 thousand miles before.
Yes I thought about the breakdown of the hall sensors themselves, but when the hall sensor fails, the motor will twitch, spin with jerks.
But now motor works smoothly.
Just too slowly. Like on first gear. But I never had a speed switcher, I didn't set any restrictions in the settings. Checked many times.
 
FWIW, some of these smarter controllers actually run with one bad hall signal (even if they dont' report a problem). I dont' know if the Kelly is one of those. (I have a Grinfineon that will, and it's a pretty dumb controller). But some of them may not respond to just two halls the same way they do to three, and so might run at a different speed, or power level, etc. Just a thought.

Regarding the limits in the controller, I have had other systems (including a Cycle Analyst once) with user-changeable settings show me that certain settings were a certain way, but they definitely did not operate as if the settings were like that. Changing the setting to something else didnt' change the behavior, nor did changing it back. When I completely factory-reset all of the settings, *then* changed the settings to what I wanted, *then* it worked, in most cases. In a few devices, it simply never worked like it should again, for no detectable reason. :(



Kirill said:
You got it right. Just one correction the controller remains old. Settings as they were.
Ok. This was why I thought you had replaced the controller:
Kirill said:
I'm changed controller and throttle but nothing changes.

The only problems I have definitely seen cause a motor to be limited to a certain speed are:
--a speed limiter in the controller
--a problem with hall signals either
----at the halls
----in the wiring
----in how the controller reads the incoming signals (or is set to respond to them, or what it expects them to be vs what they actually are).

A couple of times it was suspected that a wrong phase/hall combo and/or a faulty "self learn" / phase/hall angle detection caused it, but the people I was helping never responded so I don't know if they fixed it or what did.

The only Kelly controller I've ever dealt with personally has a definite problem where it doesn't correctly spin a motor, so I don't know enough about which settings affect motor operation in the way you see to suggest specific ones to change.

If you havent' already, you might check with Fany at Kellycontrols, to see if they have a suggested complete reset method to establish a baseline working system, then you can re setup your settings to recustomize it to the way you actually use it.
 
This is of course all true. But I was connecting another controller. It’s works same.
The new cable is the old cable. This is because the motor cable is too thick. The factory-braided cable doesn't fit into the motor axle. I've to removed a few centimeters of the outer braid, then put on a heat shrink tube and insert a slightly thinner structure into the motor axle. I had to shorten the cable for the new connection.
Yes I tried to do a factory reset or something like that, but didn’t find the information about this.
But in this situation, I will forced to change the hall sensors too.
 
By detachable cable, do you mean one with an additional connector near the motor, to make maintenance easier?
If the cable is two parts, did you replace both parts (and how did you identify the original problem)?
Did you purchase, or make/assemble the cable yourself (plug n play or custom assembly)?

EDIT: I see you response above. Did you do a continuity check on all of the wires before/during/after assembly? Does braiding mean shielding?
 
E-HP said:
By detachable cable, do you mean one with an additional connector near the motor, to make maintenance easier?
If the cable is two parts, did you replace both parts (and how did you identify the original problem)?
Did you purchase, or make/assemble the cable yourself (plug n play or custom assembly)?

EDIT: I see you response above. Did you do a continuity check on all of the wires before/during/after assembly? Does braiding mean shielding?

Yes, a two-piece cable. Both parts are old and have already been used successfully.
I removed only the outer plastic braid.
Shielding have only two thin wires. I decided it's for sensitive signal like temperature or speed.
 
Sorry for randomly barging in.
Regarding the limits in the controller, I have had other systems (including a Cycle Analyst once) with user-changeable settings show me that certain settings were a certain way, but they definitely did not operate as if the settings were like that. Changing the setting to something else didnt' change the behavior, nor did changing it back. When I completely factory-reset all of the settings, *then* changed the settings to what I wanted, *then* it worked, i
I had exactly same issue myself with CA3. Induced by an uninitiated user who returned the bike to me holding both brakes because "the motor doesn't stop". She apparently set cruise control without realizing it. CA quit working and didn't respond to settings changes. Total reset and re-entering all settings fixed it.

Not likely it's Kirill's problem but who knows.
 
Identifying a possible malfunction of the hall sensors, I checked all possible combinations of hall wires.
Here's the picture. Rotation of the wheel with two options is 11 km / h, and for 4 other is 20 km / h.
Is it mean all hall sensors for replacement?

ws4isb.jpeg
 
Kirill said:
Identifying a possible malfunction of the hall sensors, I checked all possible combinations of hall wires.
Here's the picture. Rotation of the wheel with two options is 11 km / h, and for 4 other is 20 km / h.
Is it mean all hall sensors for replacement?

ws4isb.jpeg

Checking combinations of halls doesn't test for if there is a connection problem to them or test if they work. You have to actually check the signals the controller itself is receiving from them.

Some of the Kelly setup software shows the actual hall signals recieved on screen, as you manually rotate the wheel with your hand on the tire. (For the Mac you have to rotate it backwards, so the gearing will force the motor to spin inside; otherwise the internal freewheel/clutch will keep the motor from being spun by the wheel).

I recommend performing this test:
https://ebikes.ca/documents/HallSensorTestingFinal.pdf
from here
https://ebikes.ca/learn/troubleshooting.html
 
amberwolf said:
I recommend performing this test:
https://ebikes.ca/documents/HallSensorTestingFinal.pdf
from here
https://ebikes.ca/learn/troubleshooting.html

Thank you so much. I've spent huge time seeking how make check by multimeter.
 
With Kelly controller all three wires work the same way: when rotating backward, the voltage is constantly about 5.56V
After stopping about 0V and sometimes changing after rotation to 11.15V
 
Kirill said:
All three wires work the same way: when rotating backward, the voltage is constantly about 5.56V
After stopping about 0V and sometimes changing after rotation to 11.15V

Are you rotating the wheel backwards by using your hand on the tire, very very slowly? If you move it too quickly then they change too fast for your meter to keep up. Move it slowly enough to watch them change.

If they don't change back and forth from nearly zero to nearly full voltage at a constant rate as you turn the wheel at a constant very low speed, then there may be a problem.

If they do, then they are working ok.
 
Installed another noname controller.
Yes, rotate by hand very slowly. Doesn't change, just about 2.65V

After stopping 5V or 0V
 
Kirill said:
Installed another noname controller.

Does the system operate normally with the noname controller, vs the kelly?

If not, it's more likely still something at the motor or wiring from motor to controller.

If it does, it's something about the Kelly, it's wiring, or it's settings (even though you've checked all that).

The only other thing common between the two would be your throttle, like if a magnet got displaced so it doesn't output full throttle voltage, or your battery, if it's not high enough voltage to spin any faster.


Yes, rotate by hand very slowly. Doesn't change, just about 2.65V

After stopping 5V or 0V
If they are all changing, and never being stuck, then the sensors themselves, and the wiring from them to the controller, is all working basically correctly.

There could still be noise in teh signal at higher speeds that causes problems with the ocntroller reading them, but you can't check for that without an oscilloscope, or by shielding the hall sensor cables away from the phase wires all the way from motor to controller. (can't do much inside the axle, though, since you don't have room).


The 2.65v you are seeing is the sensor output changing too fast for the meter to keep up with. You'd have to turn the wheel slowly enough watch it change. The MAC has a gear reduction, so the motor is spinning several times faster than the tire. So by very slowly, I mean *really really really* slowly.

It's not necessary for the test (which seems to have passed), but if you wish to see what I mean about the sensor changing during spinning, you can turn the motor just barely, like not even a quarter the distance between two spoke nipples, and probably see the value change.
 
Worse. There were jerks and even stopping of rotation. But very rarely. It is more difficult for the noname controller to choose the right combination of phase and hall wires. Basically everything is fine. Noname controller and costs a lot less.
I tried riding with noname controller too. Speed was the same.

Yes I understand. Still, it's hard to reproduce the slow backward rotation. But as you said moving the wheel a quarter or one-fifth back, the value changes. 0V or 5V. But sometimes not immediately, that is, two or three movements back and then changes.
 
Kirill said:
Worse. There were jerks and even stopping of rotation. But very rarely. It is more difficult for the noname controller to choose the right combination of phase and hall wires. Basically everything is fine. Noname controller and costs a lot less.
I tried riding with noname controller too. Speed was the same.

Well, the jerks and stops are unlikely to be caused by a throttle problem. An intermittently broken ground wire to the throttle could cause sudden speedups. An intermittently broken power wire to the throttle could cause jerks and stops, as it connects and disconnects.

A battery could do it but normally that would be from a BMS shutdown, and that would stay shutdown, usually.


If there is an electrical noise problem in the hall sensors/wiring, and the nonname controller has less noise filtering on the hall signals than the kelly, then it would make sense for it to have more problems running the motor.

Also, if it is electrical noise in the hall sensors/wiring, the fact that the noname uses only 5v instead of 12v (which the kelly uses) for a pullup on the signal line makes the difference between on and off much less, so noise will affect it more than the kelly.

So...at this point my best guess is there is electrical noise on the hall lines, or a poor ground between halls and controller, or both.

If the ground is good, then you can try a small 0.01uf ceramic capacitor (usually marked 103) on each of the hall signals, to ground. Easiest place to do this is at the connector from the motor to the controller. Best place is right at the controller itself, but that's not usually practical. A very small capacitance will help absorb higher frequency noise that is riding on the signal, that may be affecting the controller's ability to read the actual hall signal.

If this helps but does not eliminate the problem entirely, then if you can at least temporarily physically separate the hall wires from the phase wires for the whole length of the cable from motor to controller, by as far as possible, at least a couple of inches or more, then that will help reduce induced electrical noise from the phase wires to the hall wires.

If that helps even more but still doesn't totally make it go away, shielding the hall wires with a metal braided tube, etc., may work. But that's pretty extreme, and indicates some other problem to be solved (grounding issue, bad sensors, etc).


The thing that is still wierd is that the motor runs that slow (same speed?) for both controllers. That might not be part of the same problem, and might be the throttle itself simply never reaching full voltage. The kelly software should show you what the kelly sees from the throttle as you move it thru it's range.

I just don't see a reason that both controllers would run the motor at the same slow speed with a hall signal interference issue. Run it badly, incorrectly, etc., and even slow, but not at the same speed...
 
I have a question why nothing like this has happened before?
The same cable was used. The cable connection has been soldered and insulated the same.

My motor cable has 3 thick phase wires and 9 thin wires. Of which 2 thin wires are wrapped in a plastic foil and is shielded. I decided that two this shielded wires for the temperature and speed sensor. Since there are very sensitive signals.
 
I make measurements on the throttle handles.
Both throttle handle used by myself before.
Replaced because of the convenience.

So, both handles have the maximum value 3.58V on the signal wire. Use with noname controller. And speed 11 km/h.

Use Kelly controller both handles have the maximum value 4.2V. And speed as wrote before 20 km/h after choosing the best combination of hall wires.

Is it mean what at least throttle handles fully functional.
 
Kirill said:
I have a question why nothing like this has happened before?

Don't know. If we knew that, we'd probably know what was wrong and be able to fix it. ;)

My motor cable has 3 thick phase wires and 9 thin wires. Of which 2 thin wires are wrapped in a plastic foil and is shielded. I decided that two this shielded wires for the temperature and speed sensor. Since there are very sensitive signals.
Could be. But Motor Hall signals are sensitive signals, too (same type as the speed sensor, but actually are higher frequency and more noise-susceptible, because the sensors are actually in the windings/stator teeth, in the middle of the actively-changing high current magnetic fields). I'd expect all those thin wires to be shielded, if any of them were.
 
Kirill said:
I make measurements on the throttle handles.
Both throttle handle used by myself before.
Replaced because of the convenience.

So, both handles have the maximum value 3.58V on the signal wire. Use with noname controller. And speed 11 km/h.

Use Kelly controller both handles have the maximum value 4.2V. And speed as wrote before 20 km/h after choosing the best combination of hall wires.

Is it mean what at least throttle handles fully functional.
Sounds like the throttles work normally.

And speed as wrote before 20 km/h after choosing the best combination of hall wires.
I just need to clarify:

Your first post said you only get 7mph (11km/h). Another post shows that for some hall combinations you get that, and for others you get 12mph (20km/h). But I don't see anything that lists the normal speed you get.

Is 20km/h (12mph) the normal maximum speed you get out of the system when it is all working?

I assume that it is not, but if it is, then why not just use that combination?

If it is not, then there is still something wrong that is common to the parts on the bike that have *not* been changed. Meaning, not the controller, not the throttle.

The most likely problem is induced electrical noise into the hall signals, and/or a bad ground to the halls. (one hall, or all of them).

Additionally:
If the controllers both have an "auto learn" or "auto identify" feature for figuring out the phase/hall combination themselves, but you get different results with them by manually changing the combinations, this also points to an issue with the hall signals, either in the wiring, or the connection (not necessarily detectable with just a multimeter ohm test), or the halls themselves.

It might feel like there's a ghost in there somewhere, but whatever the problem is should be findable, by eliminating things one at a time. :)
 
Back
Top