Self Study Controller doesn't work and behaves weird

[Kollege_122]

Hello!
I got this Controller (https://www.ebay.de/itm/363427692253) and connected it to my motor harvested from a self-balancing unicycle. It works with the original controller.
I connected the Hall wires and Phase wires to the controller. The so called "electric door lock"-wire is connected to positive Terminal coming from the battery. I also connected a Potentiometer as a throttle.
The only way by which I can get the motor to move only just a little is by connecting the two Self-study wires into each other and then disconnecting the battery. After disconnecting I can see the motor moving a tiny bit.
Maybe someone has an idea about my problem and how I can possibly solve it.
Thanks!
Kollege

 

[amberwolf]

A potentiometer (pot) has a wider voltage output signal range than a hall throttle (whcih the controller is designed for); it's possible the controller detects the throttle input is too low and is erroring out because of this (even when you're not using it) and causing undesired behavior.

You an add resistors to the two outside pins of the pot (from one pot pin to 5v, and the other pot pin to ground) to cause it to have the narrow range required for normal operation on most of these controllers. What resistance each of those needs to be can be calculated, once you know:
--the potentiometer end-to-end resistance (this is usually stamped into the metal of the pot somewhere, or printed on it)
--the voltage range the controller expects on the throttle input (this is usually on the sticker on the controller; I was going to check the ebay page for this but the pictures there are of at least two different controllers, so none of the label information in them may be applicable; you'll need to look at your actual controller to see what it says. It's typically around 1v-1.4v up min to 3v-4v max.)

The total voltage across the pot now is the ~5v the controller supplies. The goal of the resistors is to make the "top" pin of the pot throttle only go up to the highest throttle input voltage the controller expects, and the "bottom" pin start at the lowest throttle input voltage the controller expects. The rest of the voltage will be across those resistors.

If you have three identical pots you can put two of them in place of the resistors, and adjust them to experimentally determine the right resistance. Wire each of those two "outside" pots with their center pin connected to one of their outer pins (doens't matter which one). Then set your multimeter to 20 DCVolts, and measure from ground (black meter lead) to the "bottom" outer pin of the center pot (the throttle) with the red meter lead, and adjust the "bottom" pot until that point reads whatever the min voltage the controller expects. Do the same for the "top" pot, adjusting it till the "top" outer pin of hte center pot reads whatever the max voltage the controller expects from the throttle.

 

[Kollege_122]

Hello!
Thanks for your response.
Sadly I don't find any informations about the input range. My (really short) manual says the self study progress should also work without a Throttle connected.
It also doesn't make any difference if the potentiometer is connected or not, the controller still behaves the same.
Do you have any other idea why it isn't working?

 

[TommyCat]

Testing with correct voltage and amperage supply? (battery)
Leave the brake connections disconnected, as well as any other peripherals not required for start.
To verify that the controller is "energized", is 5vdc input present at the throttle connector. (or motor's hall connector...)
As it looks like a dual mode controller, try it with the motor hall sensors disconnected.


Regards,
T.C.

 

[Kollege_122]

I use a 16S Battery Pack, I wouldn't expect any issues from that side, especially while having the motor running without any load.
I can measure 5V on the throttle and hall connection.
Disconnecting the Hall-cables doesn't make any difference. I thought the dual mode means it can handle both 60 and 120 degrees phase angle.

 

[fechter]

Most controllers will have several things that prevent the motor from running as safety features. If the self-study feature works, then the hall sensor arrangement should be OK. Other things are like battery under or over voltage, brake input, and throttle voltage out of range. As ambewolf pointed out, the throttle signal is typically from 1v - 4v and anything outside this range will trigger an error and cause shutdown. You can use your voltmeter to measure the throttle signal and get it around 1v, then cycle power on the controller and see if it runs.

I didn't see a diagram on the web page link. If you can post the diagram, it might be helpful.

 

[Kollege_122]

I have tried it and measured 1V as throttle input. Still nothing has changed. I also don't have a complete diagram, but I try to attach the bit of documentation I have.

 

[amberwolf]

I have tried it and measured 1V as throttle input.

Could you be more specific about what "it" is, and under what specific conditions you measured 1v as a throttle input?

These are important for us to help you troubleshoot the problem.

For instance, if you measure 1v as a static voltage on the throttle input, and the motor is not moving, that would typically be normal.

But if you measure 1v on the throttle input at powerup, and then you turn the throttle slowly to increase the voltage to say, 2v, and you get no motor movement or other system reaction, and no current flow out of the battery, then this means the controller is not responding to the throttle input for some reason (of which there could be several).

If conditions and actions are the same as above but you do get current flow increasing with increasing throttle voltage but still no motor movement, it is likely the hall/phase combination is not yet correct, and the "self-study" either isn't working or hasn't found the right one on it's own. If this is the case you can find the combination manually, if necessary. Or you can leave the hall wires disconnected completely, as this controller should work sensorless according to the first paragraph in the documentation.



Something else to note:
You said originally:

The so called "electric door lock"-wire is connected to positive Terminal coming from the battery

however, the documentation shows several possiblities for this wiring.

One shows there are two wires for that whcih should be connected together, and neither one connects to anything else at all (not the positive of the battery, either).

Another shows the more common option for controllers, which is a single wire that connects to the batt+ instead.

Then there are others that are a bit less clear.

Just to be sure, are you certain that yours has just the single wire for door lock? Or a pair?

It depends on what that wire actually does in the ocntroller for whether this will work, or if it will damage the controller, etc. Controllers that connect it to B+ use the wire to power the low voltage power supply from the battery, and then the LVPS powers the brain, hall sensors, throttle, etc. But it would be just as easy for a controller to have a logic-level (5v or 3v) "enable" line that is called the same name, and that could cause controller damage or failure if connected to B+.

 

[Kollege_122]

I set the potentiometer to multiple settings. I moved it to multiple positions between 1 and 4 Volts and still got no response.

I did not check for current flow, however if the motor is moving slightly after disconnecting the battery I expect it to be connected in a correct way.

Yes, the controller has one wire labeled "electric door lock". I connected it to B+ since I have no other accesories which could deliver a signal for it.
The controller also has some connections labeled something with alarm system, but since those were not mentioned in the documentation and the plugs are not fitting together I left them without connection.

 

[amberwolf]

I set the potentiometer to multiple settings. I moved it to multiple positions between 1 and 4 Volts and still got no response.

But what was it *at system power on*? That is very important, because some controllers will disable themselves for safety reasons if the throttle is above a certain voltage at power on, and they may also do it if the throttle is at zero at that point (though I haven't run into one yet).

The common hall throttles are usually around 0.8v minimum actual output, so setting the pot to soemthing between 0.8-1.0v should be safe. Then disconnect battery from system, and then

I did not check for current flow, however if the motor is moving slightly after disconnecting the battery I expect it to be connected in a correct way.

That's not necessarily the case.

There are typically 36 possible ways to connect the hall and phase wires. Different combinations will produce different results, including some where it will "lock" the wheel in position rather than spinning it, and then the motor will "jerk" a little when power is disconnected as the "lock" is released. Generally all of the non-correct combinations use more current (some of them a LOT more) than the correct ones, and that's where monitoring current flow out of the battery during these tests is helpful.

(technically there are only 6 combinations (3 forward and 3 reverse), with 3 offsets each of those, as each is "rotated" by one position to get the others, but for whatever reason there are certain motor/controller combinations that only work in one single combination...requiring all 36 combinations to be tested to verify that's the only one.



Controller self-learn or self-study usually tests each of these by internally changing timing on the phases and reading different orders of the hall signals, then choosing the best result and saving that in memory to use for actual riding.

Sometimes it chooses a "backwards" combination, so a direct drive hub will then spin backwards (a geared hub will not spin at all in this case, though you will hear the motor running inside it, because the freewheeling clutch in it is not engaged when motor is in reverse).

If there is a problem wiht one of the hall signals (wire, connector, bad sensor, etc) then the controller won't be able to find a correct combination. In that case, the results aren't predictable without knowing how the controller chooses combinations and what it does in the case of test failure.

However, since that controller should support sensorless operation according to the documentation, simply disconnecting the halls completely, then powering the system off, and back on, should let it operate correctly in sensorless mode.

When using this mode, self-study no longer applies and is not used, because that is only to determine which hall order matches which phase order.

It's an easy quick check to see if the controller/motor system is operating at all, and if it works then you can use it as-is or you can begin troubleshooting the hall sensor wiring/connectors, etc.

 

[jimmyhackers]

does it work without the hall effect sensors pluged in? (motor position sense wires)

try it without the motor sense wires first (unplug them). then try self learning. if it works afterwards.....there is a problem with your hall sensors or their wiring or detection

the other possibility is that the wrong colour wires were soldered into the wrong ports in the factory.

if it still doesnt work.... keep the hall sense plug out still. but go through swapping every combination of the 3 main phase motor wires.
hopefully one will make the motor spin smoothly.

if one does make it go smoothly....make sure its spinning in the right direction. if not, disconnect the self learnig wires with battery still attached and reconnect to change direction.

if this does work/fix it.. you will have to replug the hall sense wires in again. and potentially go through swapping their 3 signal orders over/about if you dont get a smooth rotation.

 

[amberwolf]

if it still doesnt work.... keep the hall sense plug out still. but go through swapping every combination of the 3 main phase motor wires.
hopefully one will make the motor spin smoothly.

There is no purpose to swapping around just phase wires. The only thing you can do with just phase wires to change anything about the behavior of the motor is to swap two, any two, and only two, of them to reverse the direction of a motor running on only phase wires (no hall sensors or other position sensors).

 

[jimmyhackers]

there is a purpose if man in the factory soldered the wrong colour wires in the wrong ports on the esc board.

its also possible the same error could of been done only to the sense wires.
or both.

i just think that in order of operation for this particular fault find, youd want to try and make the motor spin in the right direction sensorless first.

then see if the hall sensor parts messes it up if you plug them in.

testing 3 phase wires and 3 sense wires in all various combinations.....is a lot more of a headache than just swapping the 3 phase wires round by themselves.

 

[amberwolf]

there is a purpose if man in the factory soldered the wrong colour wires in the wrong ports on the esc board.

No, because the only thing you are doing is changing the motor direction, if you are *only* using the phase wires.

If a motor spins sensorless at all, the phase wires are as good as they can possibly get.

If it's spinning in the wrong direction, you swap ANY two, but ONLY two, of those phase wires to change it's direction.


If swapping around just phase wires in a sensorless setup makes any difference other than that, you have connector installation problems, damage, or failures, or wiring installation problems, damage or failures, that are being shown to you by the physical movement you're doing with the connectors and wires.

 

[jimmyhackers]

i guess your logic process is different to mine. i was under the impression if all three were swapped/wrong, the motor may not turn at all.

i think if im trying to run the motor sensorless and it still doesnt turn, or jerks etc....swap the phase wires about.
(and tbh if it was still messing up id put it on my esc and hub motor tester, but i guess the OP doesnt have one).

also. i guess even if they have got the hall sense wires wrong....and the main phases are ok...it might be easier to swap a main phase cable than a little annoying hall effect cable.

i guess we will have to wait to see what was wrong and what fixes it.

 

[amberwolf]

i guess your logic process is different to mine. i was under the impression if all three were swapped/wrong, the motor may not turn at all.

That's only true if you are running with hall sensors.

When running sensorless, there are only three phase wires, and there can only be an order that goes forwards, or an order that goes backwards. You can draw it out on paper, in a circle, and see this, if you like.

Let's say you have BGY. If you write it as YBG, it's still the same order, just shifted by one. Same for GYB. Remember that this is a circle of letters, not a line, so it's continuous, rotating.

If you put it YGB, then it's the reverse order, and will spin backwards. Same for BYG, and GBY.

When running with sensors, then there are different results depending on which phase matches with which hall, so it does matter for that case.

 

[Kollege_122]

Thank you all for your help!
I could not get the controller to run, I think it is broken. I ordered a new one and with this one the motor works perfectly fine.