Another Possible Hall Problem?

[synthrider]

Hi guys,

Here's the history: motor stopped responding, first intermittently, then permanently. I thought the most likely culprit was the stock controller, so I tried putting in a programmable Lyen: 6 FET 3077 Mark II Mini Monster LYEN Edition Controller. Problem is, still couldn't get it to work, even though it was receiving power.

Lyen was very helpful and responsive in troubleshooting... he asked me to do a Hall Sensor Test (like this one: http://www.ebikes.ca/troubleshooting/HallSensorTestingFinal.pdf). When spinning the wheel backwards and testing on the black hall wire and the other hall wires, the voltage does change but sort of erratically:

(nothing happens when I spin it slow, I have to spin pretty hard. and it only seems to go to zero if I turn it really hard)

b/w black and...

Blue, goes from 3.7V to between 2.9 and 4.2, sometimes flashing to 0

Green, goes from 3.7 to 5.0, sometimes flashing to 0

Yellow, goes from 3.7 to 2.0, flashing to 0

When it goes to zero, sometimes I'm not sure if it's because I'm losing a good connection with the voltmeter or what... The yellow does go to zero pretty consistently. The other ones I'm less certain about.

Does this make any sense to you guys?

(I haven't heard from Lyen for a while... I think I got lost in his inbox or something). If it's a hall sensor issue, is it conceivably possible to program the Lyen controller into "pedal first" mode, where it wouldn't need the hall sensors at all?

Or could I get a different "pedal first" controller, and bypass the problem that way?

Thanks!

PS Bike Stats: bike is X-treme Scooters XB310-Li, 36V10ah Li-Ion battery, 300W Bafang motor (could actually be 250W despite manufacturer claims)

 

[motomech]

I use an analog meter for Hall testing and the readings are pretty definitive. The needle deflects or it doesn't.
Sound like it could be the wires in your case.
I have a Lynn 6 FET Mini-Monster and it will work with senor or sensorless without any re-programing.
I usually have to get the wheel to turn about a half a turn before the motor{MXUS geared) and the controller get in "sync".

 

[gwhy!1]

the readings that you are getting from the hall sensors are not what they should be. they really should only go from 0v to 5v and not really anything inbetween. find it very unlikly that all 3 sensors to go faulty all in the same way , so just to recap the black probe on the meter fixes to the black wire on the hall plug, connect red probe to R ( red should be 5v all the time) Y,G or B should be 0-5v switched, power up the controller and spin wheel to test. If you get the same results then try connecting the black meter prob onto the black battery wire ( instead of into the hall plug ) and re-test.

 

[dogman dan]

Those results would tend to make me think, wiring problem, perhaps just the difficulty of doing the test with the probes. But more likely a problem with a wire, which could be the black or red one.

You are really really sure that your main power wires and connections are good? no chance of a nicked power wire at the hub entry?

 

[dnmun]

you do not spin the wheel. i think you are just doing the hall sensor test wrong. how do you make the connections to test? you should see the voltage toggle between almost 0V and almost 5V when you move the wheel only one of two spokes at most.

 

[synthrider]

I use an analog meter for Hall testing and the readings are pretty definitive. The needle deflects or it doesn't.
I have a Lynn 6 FET Mini-Monster and it will work with senor or sensorless without any re-programing.
I usually have to get the wheel to turn about a half a turn before the motor{MXUS geared) and the controller get in "sync".

Here's what I just tried: I plugged the throttle in. Unplugged the hall connection entirerly. Then I tried doing the following with all 6 possible ways of connecting the three phase wires (yellow to yellow, blue to blue, green to green; Y to B, G to G, B to Y, etc...):

1) I turned the power on,
2) I gave the pedals a good crank to get the wheel moving
3) I turned the throttle

The motor never engaged, although in most cases it did grumble for about 1 second when I depressed the throttle. In most cases, I noticed that the hall wires got warm, sometimes hot.

- - -

You are really really sure that your main power wires and connections are good?

Yah if I hook up a cycle analyst to the controller, all the readings look good...

no chance of a nicked power wire at the hub entry?

The hub entry wire looks good. The insulation is intact... there is a spring protector thing surrounding it, also in good shape. As far as I know that part is fine.

- - -

Any thoughts? I will check my hall pins and redo the test... but it seems like it should be working without the hall sensor connections, if I spin the wheel manually to start...

 

[NeilP]

Here's what I just tried: I plugged the throttle in. Unplugged the hall connection entirerly. Then I tried doing the following with all 6 possible ways of connecting the three phase wires (yellow to yellow, blue to blue, green to green; Y to B, G to G, B to Y, etc...):

1) I turned the power on,
2) I gave the pedals a good crank to get the wheel moving
3) I turned the throttle

The motor never engaged, although in most cases it did grumble for about 1 second when I depressed the throttle. In most cases, I noticed that the hall wires got warm, sometimes hot.

Doing this and leaving the throttle pressed even a little way, for more than a second or less with halls and or phase wires misconnected is a good way to blow the controller

Well this is not testing halls or anything else..you only try to get the wiring combination sorted IF you know you have a working controller and motor.

Forget the phases and spinning the wheel and concentrate at one thing at a time..like testing the halls and the windings on the motor.
- - -

You are really really sure that your main power wires and connections are good?

Yah if I hook up a cycle analyst to the controller, all the readings look good...

Hooking up the CA tells you very little. it may register a voltage, but that can happen even with crap connections. As soon as big current is drawn and things start moving then the problems will show,

but it seems like it should be working without the hall sensor connections, if I spin the wheel manually to start...

WRONG..unless it is a Sensor-less controller that does not have hall connection any way. or you have a fancy new type of controller that only uses halls for start up and then ignores them after that.



So test the halls again. Really you should buy an e-bike tester from Ed Lyen or e-bay but if you do not have one, get yourself 3 AA batteries to create a 4.5 volt supply.

Can you find a secure way to hold 3 batteries together with two wires connected to them? if so do it so yo have a stable battery supply of 4.5 volts.
Turn bike upside down so you can spin the rear wheel freely.
Connect the three batteries up to the red and the black of the halls at the motor. No controller connected AT ALL.

Now get volt meter and in low volt range..10 v or so..connect black to the black hall wire and the battery negative.

Now connect the red of the volt meter to one of the hall out wires from the motor.
It should read 0 volt or 4.5 to 5 volts..whatever the battery voltage is.
Now turn the wheel very slowly while watching the meter.
As you turn the wheel, the voltage should go from 0 to approx 4.5 volts, then zero again.
it should just be Voltage on/off/on/off/on/off. not progressive..just on and off.

Try each individual hall output and see if they all do this or not.

If they are all wrong..then there is various explanations.
Still bad testing technique.
All halls bad.
bad wiring or a short internally

 

[synthrider]

I redid the test just as above (thanks NeilP!).

I was giving it about 4.5V with three AA batteries.

None of the colored hall wires performed as expected when I turned the wheel.

They were all at 0 or just a little over when stationary, and at most went up to maybe .2V.

So it's either the halls or an internal short/wiring problem. Any way to determine which? Without replacing the halls?

Thx.

Can you find a secure way to hold 3 batteries together with two wires connected to them? if so do it so yo have a stable battery supply of 4.5 volts.
Turn bike upside down so you can spin the rear wheel freely.
Connect the three batteries up to the red and the black of the halls at the motor. No controller connected AT ALL.

Now get volt meter and in low volt range..10 v or so..connect black to the black hall wire and the battery negative.

Now connect the red of the volt meter to one of the hall out wires from the motor.
It should read 0 volt or 4.5 to 5 volts..whatever the battery voltage is.
Now turn the wheel very slowly while watching the meter.
As you turn the wheel, the voltage should go from 0 to approx 4.5 volts, then zero again.
it should just be Voltage on/off/on/off/on/off. not progressive..just on and off.

Try each individual hall output and see if they all do this or not.

 

[NeilP]

Since none of the halls are firing it sounds more like a power supply problem to the halls, so may be that you have a broken red or black wire within the axle or hub. Still possible that all the halls are fried, but now the only way to find out is to open the motor up and have a look

 

[itchynackers]

The way I test hall wires for cuts/grounds is to do a continuity test with a simple $2 multimeter from Harbor Freight. Put it in "diode" mode and place one probe near your hall leg, and the other end outside the motor on each hall wire (at the plug). This way you can see if your hall wire is cut (no continuity), and if it is grounding with other wires (continuity). Do this for all combinations for each hall wire.

 

[synthrider]

Hi itchynackers thx for the advice,

I do have a multimeter with continuity mode... Can you please rephrase where to place the probes?

I think I follow that the basic idea is to test between the plug-end (easily accessible) and the motor-end...

There is only one thick wire coming out of the motor... do you mean to cut into the insulation to get access to the hall wires near the motor?

Thx

The way I test hall wires for cuts/grounds is to do a continuity test with a simple $2 multimeter from Harbor Freight. Put it in "diode" mode and place one probe near your hall leg, and the other end outside the motor on each hall wire (at the plug). This way you can see if your hall wire is cut (no continuity), and if it is grounding with other wires (continuity). Do this for all combinations for each hall wire.

 

[NeilP]

Since you have not got the motor open yet, then don't cut the wire , but test from the plug end and just see if you have a connection from ANY of the wires to the motor or axle. Any very low or continuity from ANY of the wires to the motor axle or case is bad, and will indicate a problem.

Sam goes for continuity between any two wires, any zero ohm readings indicate a break in the wires and they are shorting together

Itchynackers is talking about a test once you have the motor apart and can get to the actual legs of the halls

 

[synthrider]

Thx for bearing with me... can you please read what I did, and if I screwed up, rephrase how to do the continuity test?

I set it in continuity mode and tested between the different hall wires (not hooked up to anything except the motor, no power source etc.)...

Everything read "0.F" except when I tested between the black and red. (black/red tested 0.016 V in "resistance" mode and 33.3 ℧ in "diode" mode)

Also, when I changed the mode to check for shorted or open circuits... it said "Open" for all the wires, except between the black and red which tested "Shorted."

Since you have not got the motor open yet, then don't cut the wire , but test from the plug end and just see if you have a connection from ANY of the wires to the motor or axle. Any very low or continuity from ANY of the wires to the motor axle or case is bad, and will indicate a problem.

Sam goes for continuity between any two wires, any zero ohm readings indicate a break in the wires and they are shorting together

Itchynackers is talking about a test once you have the motor apart and can get to the actual legs of the halls

 

[NeilP]

(not hooked up to anything except the motor, no power source etc.)...

Assuming you are meaning just the thick phase wires are connected to the controller and not the hall connector and you are pushing the test probes in to the wire end of the connector.
If you are then that is wrong. Disconnect motor completely ..all wires, halls and phase wires.



┣━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┫

Everything read "0.F"

Don't know what "0.F" actually means. Never seen a meter with that as a display reading



┣━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┫

(black/red tested 0.016 V in "resistance" mode and 33.3 ℧ in "diode" mode)

I guess you got that the wrong way around
Resistance mode will read Ω
Diode mode will read volts. What diode mode is reading is the voltage drop across a junction, like the junction between internal parts of the hall sensor



┣━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┫

Also, when I changed the mode to check for shorted or open circuits... it said "Open" for all the wires,

What are you checking between? From one wire to another ? or as I said? between the wire and the case and axle of the motor

but test from the plug end and just see if you have a connection (SHORT) from ANY of the wires to the motor or axle.

Any SHORT or very low resistance or continuity from ANY of the wires to the motor axle or case is bad, and will indicate a problem.

┣━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┫

except between the black and red which tested "Shorted."

I do not have a hall or motor her at home to check, though wil be down the workshop later, so will check between rd and black. Would expect some connection between them..possibly ...since that is the power in and out for the halls.but no idea if it is going to read a short or some resistance value on your meter...or even what I would expect to see on my meter.

But any way what you should be checking for is SHORTS between each wire and the case. ( and even the phase wires)




┣━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┫
What we are trying to achieve is see if ANY wires is broken and touching anything that it should not.
The Only place you should see any continuity ( or a short on your meter) or (nearly) zero Ω is between any one of the thick wires and and another thick phase wire.
You may see low readings between black and red and between hall outputs but they should not be a short..( I stand to be corrected on that)

┣━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┫

First we will test to shorts from ANY of the wires to the axle

So, since you have a mode that reads SHORT or OPEN..run with that for now. leave Diode and resistance mode ...for now..one stage at a time..we can get to those later if need be.


1) PUT TESTER TO THE MODE THAT READS SHORT OR OPEN
2) Disconnect ALL motor wires from controller..phase, halls, any temp sensors..EVERYTHING
3) Connect the red wire from your tester to the axle of the motor...Use a pair of vice grips, ,OR take the wheel off the bike and remove the axle nut from one side and put it on the other axle side, so you now have two nuts on one axle. Clamp up the Red test probe from the meter between the two nuts. make it a good solid connection that you can then ignore for now.
4)..Test that connection by touching the other probe (black) of the meter to the axle...It reads a SHORT right? If not then your connection to the axle is poor. Sort that first, the proceed to step 5

5) Now touch the black probe to all of the wires coming out from the motor, ONE at a time.

ANY SHORT INDICATION ON THE METER MEANS A PROBLEM. NONE OF THOSE WIRES SHOULD BE SHORTED TO THE AXLE, SO THE MOTOR WILL NEED TO BE STRIPPED

6) NOW JUST TO BE SURE, SWAP THE PROBES AROUND. So Black to the axle and use RED as the testing probe and go back and test all the wires again. ANy short indicates a problem. Any wires that indicates as SHORT to the axle IS BROKEN and needs replacing... you have found the problem.



┣━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┫

Second we will test to shorts from ANY of the wires to the rotating part of the case. A loose stray wire that was not tied/glued down /tucked in properly could have started rubbing against the rotating part of the motor
1) So go back up to the first test, and instead of connecting one probe to the axle, connect it to the case. Use your imagination..scrape some paint off, connect to one of the cover securing screws..remove a screw, put a short length of wire under it and tighten down again and fit probe to that. The test the connection again by touching the other probe of the meter to another part of the case and test for short again..indicating that your connection is secure.

2) Go around all the 5 hall wires and the 3 phase wires with your other probe.

3) ANY SHORT INDICATES A PROBLEM AND THE MOTOR WILL NEED TO BE STRIPPED

4) Now swap the probe that is connected to the case and run these Second tests tests again..probably overkill but best to check twice




┣━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┫





Third we will test to shorts from ANY HALL wire to the thick phase wires

Hall wires often run across the phase windings.
Windings get hot. This can melt the hall wires ..AND possibly ( unlikely but possible) short them to a bit of damaged phase winding.

So now connect one probe to a thick hall wire and test to all the Hall wires. ONLY THE HALL WIRES>>>NOT TO THE OTHER PHASE WIRES. there will be a short indicated to the other phase wires, as they are all connected together internally.

ANY SHORT FROM PHASE WIRES TO ANY HALL WIRE INDICATES A PROBLEM AND THE MOTOR WILL NEED TO BE STRIPPED




┣━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┫


That should be enough for you to get on with for now. Shoudl only take 10 minutes...not the half hour I have just wasted typing this out when I have a battery pack that should be getting my attention for now.

I'll leave someone else to hand hold you through the testing of the halls...assuming that you do not get any short results from the tests above. Any shorts of the tests above are going to mean stripping the motor apart.

 

[dnmun]

i din't read all this, but you need to have a source of current going to the hall sensor leg as well as having the 5V on the red or power lead.

when the hall sensor is triggered ON: current flows out of the controller into the hall sensor lead and that current is delivered to ground inside the hall sensor IC. this current flows out of the controller through a resistor that causes the voltage on the outer end of the resistor to drop to almost zero at the 5 pin plug on the hall sensor leg.

when the hall sensor is OFF: no current flows, and the voltage at the 5 pin plug remains at the 5V from inside the controller.

to test: connect the controller to the battery and turn it on. no need for the phase wires to be connected, but it doesn't damage the hall sensors if it is still connected. if it is a geared motor, you have to turn the wheel backwards, just one or two spokes is enuff to see. as the hall sensors are triggered ON and then OFF, the voltage on the hall sensor wire where it goes through the 5 pin plug will jump from almost 5V to almost 0V then back to 5V. this is what we call toggling on and off.

direct drive motors can be rotated either direction but it will not toggle as fast since the gear reduction is not present. but the voltage at the 5 pin plug must toggle on and off.

if you spin the wheel like this guy does, it is a useless test. i don't think he knows even now if the hall sensors work. or not.

 

[NeilP]

but you need to have a source of current going to the hall sensor leg as well as having the 5V on the red or power lead.

Just so synthrider is clear on this, that is not the case for checking for shorts. What dnmun is saying applies to the hall tests, not testing for broken wires, as I outlined in my previous post

 

[dnmun]

ok, i did not read through that, i just wanted to summarize what the test is looking for when done normally. his spinning the wheel part is what i have tried to clarify. plus needing the resistor inside the controller to drop the voltage at the 5 pin plug.

 

[synthrider]

if you spin the wheel like this guy does, it is a useless test. i don't think he knows even now if the hall sensors work. or not.

I don't know if the hall sensors work or not. It's that or wiring. I was doing the test wrong at first, but the halls failed to toggle as expected even when I did it right.

I am going to move on to the continuity tests outlined above (thx NeilP)

- - -

Does anyone have any thoughts about using a pedal first controller to bypass the halls altogether? If you think that might work, anyone have thoughts about where to get one and which one to get? (though it sounds like if its a wiring issue, could be the phase wires might be fused also...)

 

[CellBalance]

1) PUT TESTER TO THE MODE THAT READS SHORT OR OPEN
2) Disconnect ALL motor wires from controller..phase, halls, any temp sensors..EVERYTHING
3) Connect the red wire from your tester to the axle of the motor...Use a pair of vice grips, , take the wheel off the bike and remove axle nut from one side and but it on the other axle side..so yo have two nuts on one axle..clamp up the test probe from the meter between the two nuts. make it a good solid connection that you can then ignore for now.
4)..Test that connection by touching the other probe (black) of the meter to the axle...Ir reads a SHORT right? If not then your connection to the axle is poor.

5) Now touch the black probe to all of the wires coming out from the motor, ONE at a time.

ANY SHORT INDICATION ON THE METER MEANS A PROBLEM. NONE OF THOSE WIRES SHOULD BE SHORTED TO THE AXLE, SO THE MOTOR WILL NEED TO BE STRIPPED
...
Any shorts of the tests above are going to mean stripping the motor apart.

Got that. My Crystalite 408 hub motor runs fine till it cuts out at about 75% throttle. Examination shows there is continuity between the 3 phase wires and the axle, even with the stator removed from the case and the phase wires not touching the axle.

Please advise the next step in diagnosis and repair. Please define "stripping the motor apart". It doesn't sound pretty.
Thanks.

 

[NeilP]

UUUgggg...

Ummm.. if it runs fine up to a certain speed AND you say you have a short between phase wires and axle.,
I am guessing that the 'short ' is either normal or a fault in your testing method. if there is a short between phase wires and axle, any standard controller / motor combo I know of will not run at all with the short.

SO this then points to another problem, I would say at the controller end...or possibly throttle .

What happens above the 75% throttle?

 

[NeilP]

I just noticed...first post...Welcome

First. before woking on the problem...go the the Forum "User Control Panel" link at the top left of this page, just below the Endless Sphere logo,
Click the profile tab, and go down and enter your location...if you don't, others will soon ask you to do it.
If we know where you live, geographically, it often helps...there maybe another ES'er living 'next door' that can help you in person.

 

[dnmun]

i just never saw where he had tested the halls properly.

from what i could determine he was using 3 alkaline cells to supply voltage to the hall sensors and not power from the controller and the hall sensor wires were not connected to the controller through the 5 pin plug when he was rotating the tire.

never saw this but maybe it was in here somewhere.

 

[NeilP]

Eh?

back up a bit dnmun, I just replied to the new poster..not the OP...
OP was 'synthrider' I was replying to "CellBalance" and his new post regarding the motor cutting out on his first post. His post history says Posts:1