Hall Sensor or Phase wire making my electrics burn? (With video)

Glyn

10 W
Joined
Oct 7, 2017
Messages
70
Hi all,

I just hooked up my Mac 536hf motor to the Kelly KBS48101X controller.

Initally I was over the moon to see the motor start spinning, however, I have a problem...

Video here: https://www.youtube.com/watch?v=k3fi_ggXLK8


The motor seems fine when spinning in one direction, but struggles to spin in the opposite direction.

Even worse, part way through this video, a warning light comes on the controller, and then although you don't see it, smoke starts to come out of the motor! Eek!

As I was recording, I quickly disconnected the battery and tried to film the smoke but the camera didn't pick it up. The motor housing is really hot to the touch, and the cables are quite warm too. I haven't opened it up or tried to power it on again to see if any permanent damage has been done.

Looking in the manual, the flashing red LED translates to error "Motor did not start - Motor did not reach 25 electrical rpm within 2 seconds start-up. Hall sensor or phase wiring problem".

However, I've checked the wiring diagrams for both the controller and the motor and it seems to be correct.

Any suggestions?
 
I've thought about this since, and as it's a geared motor with a 20T winding, surely it shouldn't have burnt out just in the time it took me to make that video?

It's a 48v controller and 48v motor, running off a 16s LiFePo4 battery pack, charge in the battery was 56v.

Rated current for the motor is 10a, max current is 15a - far less than the controller is capable of. My understanding is the motor shouldn't pull any more than it's max current, is this correct?

Wiring diagram for the controller:
DSC_1420.jpg


Wiring diagram for the motor:
mac-diagram.png


As you can see, in both cases:
Hall A - Yellow
Hall B - Green
Hall C - Blue

And the same is true of the phase wires. I've double checked my wiring and everything seems to be as it should.

What am I missing?
 
your images don't show for me but it sounds like the improper hall phase combo,

where did wiring diagram come from? usually not color to color that would be too easy,
 
(1)
Your photos cannot be seen.
Up load your photos to ES. If you link them from a 3rd party site they will go away in the future and your post will become as meaningless to future readers as it is today.

(2)
The error code on the lights is:
kelly manual said:
Motor did not reach 25 electrical RPM within 2 seconds of start-up. Hall sensor or phase wiring problem.

Whatever scheme they are using to start/reverse the motor is not working in reverse - perhaps due to the MAC hall layout or a self-learning mode issue with the controller that substitutes for sorting out the phase/hall color coding a priori.

(3)
Glyn said:
I've thought about this since, and as it's a geared motor with a 20T winding, surely it shouldn't have burnt out just in the time it took me to make that video?

It's a 48v controller and 48v motor, running off a 16s LiFePo4 battery pack, charge in the battery was 56v.

Rated current for the motor is 10a, max current is 15a - far more than the controller is capable of. My understanding is the motor shouldn't pull any more than it's max current, is this correct?
There is no such thing as a motor voltage - that's a backhanded means to help you get the rpm you want. I don't know where those recommended currents came from but they look like battery currents (e.g 48V * 15A ~= 720W). Those currents were supposedly what you could safely feed the motor before it overheated in 'normal use'. Feed it more - it cooks. But - a motor will use whatever you feed it and the real thing that counts is the phase amps which can be many times the battery amps - making the 'recommended' battery current pretty meaningless. The (unfortunate) magic part here is that a controller can develop far more phase amps than the rated battery amps under heavy load (don't ask). You need to read this article to get the dope on motor power ratings: http://www.ebikes.ca/learn/power-ratings.html.

In your case, the controller appears to be trying to start with the wrong phase/hall combination in reverse and the phases are fighting each other (think applying power with the wheel pressed up against a wall). The controller could be delivering huge phase amps as it tries to turn the motor against itself. In the video, you flicked the pot to full power in reverse after trouble begins - essentially causing the worst possible scenario. So the motor may have gotten kind of toasty as you ran it full throttle into a wall - even though it looked like little was happening.

I think you need to contact Kelly - this is beyond the scope of the manual. There's a chance that controller may not work with the MAC for the reversing application you have in mind. The Big gear motors (BMS, MAC, eZee) have been problematic for many controllers over the years, either because of the low inductance or because of rumored strange hall timing.
 
Glyn said:
The motor seems fine when spinning in one direction, but struggles to spin in the opposite direction.
Educated guesses without being able to see any pics (you should always attach them to the post itself, so anyone that can see the post can see your pics):

--The phase/hall combo is wrong, causing excessive current draw, and overheating. You can go to the thread Determining The Wiring For A Brushless Motor, and use the info there to find the right one.

--The motor has hall "timing" that is "advanced", rather than neutral, so it requires different settings in the controller for forward vs reverse.

It oculd be both, either, or neither of those things.


Side note: the model of motor you give comes up as a regular geared front hubmotor, which typically has a freewheeling clutch inside. The manufacturer site doesn't say if yours does or not. But if it does, then under load it doesn't rotate the wheel in the reverse direction. It can appear to be rotating with no load, but it will be much slower htan in the forward direction, and if you place your hand on it it won't rotate at all (but you'll still hear/feel the motor spinning insid teh casing).


ALso, according to Kelly's page forthat controller, it'sa 40A controller, not a 10A. Unless you've changed it's currnet limit via programming it, it can put out a few times the 10-15A motor limit.
 
I have MACs and the highest wind I know is 12t. If that is a 20t it is spinning way too fast. Even a 12t is slower than that. As others have said it must have a phase combination wrong. That is called a wrong positive combo. I run my 12t on 66.6v and it's slower than that one. As also said, color to color is normally incorrect. And even from MAC to MAC it can be different. I have six and one is different. You should read the info Amberwolf mentioned and learn how to get it right. I doubt the motor is burn out.

Dan
 
Thanks for everybody's input, this is really helpful.

Have updated the images in my original posts.

The motors are 48v 500w Mac 536hf 1:5 geared motors with a 20T winding (not advertised on their website) 32 magnet 16 pair poles. According to Mac, their rated current is 10a and max current is 15a. I'd always thought previously that the motors would only ever "pull" the current they need, in increasing increments until they reached their "max current". If I've understood your post correctly though teklektik, this is completely wrong and the motor will consume as much current as you care to give it. I'll give that "motor power ratings" article a read, thanks for sharing.

The motor came from Mac with a woodruff key on the shaft so that it could operate in both forwards and reverse directions:
motor-2.jpg


The controller is a Kelly KBS48101X. It's a 48v programmable controller which is good for up to 100a.

One of the controller's programmable parameters is the amount of speed reduction in reverse. Originally (whilst I was recording the video) I thought this was why it was turning slowly in reverse.

As I say, the wiring diagrams from both manufacturers suggest the same coloured cables are used for phase and hall A, B & C, however, the controller did report a phase/hall error. I couldn't find the Determining The Wiring For A Brushless Motor thread which you mention Amberwolf, however, Kelly (Fany) did reply to my email and said to try other phase/hall combinations, and whilst searching for the thread, I was able to find the spreadsheet which shows the 36 possible phase/hall combinations. I'll hook up another motor and have a play (whilst keeping a closer eye on temperatures and not applying power if it doesn't seem right).

Whilst I'm here, I do have another question which you might be able to help with: The motor has 16 pair poles. In the software for programming the controller there is a drop-down-box which I presume lets you select how many pair-poles your motor has. However, this drop-down-box is disabled so the value it contains is stuck at 8. What effect might this have?

As always, thanks for all your support folks. This is a great forum and I would never have been able to get this far without you.
 
Woohoo!!! I found a phase/hall combination which works! :mrgreen:

Speed in both directions looks the same to the naked eye.

DAND214, as you suspected, now that I have a working phase/hall combination, it has a much slower rpm.

I ran the motors in both directions for 60 seconds whilst applying as much resistance as I was able to with my hand and the motor showed no signs of overheating. I'm very impressed with the torque; even if apply as much pressure as I am physically able to with both hands I'm still not able to prevent it turning even at a very low speed. And this was with the "burnt" motor too.

It was a very sad day yesterday but now I'm over the moon! I can't wait to see what six of these motors can do! Thanks again for your advice everybody :mrgreen:


Questions arising:
  • Although the motor seems to be working brilliantly, when looking at the photo above which shows the "burnt" coils, should I be concerned?

  • There are 36 possible phase/hall combinations. The one which worked for me was only the 5th in the list. Is there any point trying the other combinations if this combination appears to be working? I guess what I'm asking is, is there a way to be absolutely certain that this is the best combination?

  • Please also see the question in my last post about pole numbers.


Working Combination
And for completeness, here's the working phase/hall combination for a Kelly KBS48101X controller and a MAC 536hf motor:
motor phase.PNG
 
Glyn said:
Have updated the images in my original posts.
I can see some but not all.

If you attach the images to the post (see "attachments" tab below text entry box) then everyone that can see your post can see your images. ;)

Plus then they will always be there as long as the forum is there, and not go away whenever your hosting site is gone or changes it's permissions, etc. (which happens frequently, so that information others could have used is now permanently lost).


I couldn't find the Determining The Wiring For A Brushless Motor thread which you mention Amberwolf,

Apparently I got the name slightly wrong, but a simple google of the phrase still finds it and the wiki article too:
https://www.google.com/search?q=Determining+The+Wiring+For+A+Brushless+Motor&ie=utf-8&oe=utf-8



Whilst I'm here, I do have another question which you might be able to help with: The motor has 16 pair poles. In the software for programming the controller there is a drop-down-box which I presume lets you select how many pair-poles your motor has. However, this drop-down-box is disabled so the value it contains is stuck at 8. What effect might this have?
It migh tbe disabled if the contorller automatically determines htis info, or if the software is in an auto-learn mode, and might let you change it once out of that mode.

If it doesn't automatically determine the actual number of poles, then it probably isn't going to switch with correct timing. What that will cause I don't know, but I'd guess it could cause overheating or other issues.

FWIW, since you have a geared motor, it effectively has a number of poles equal to the motor poles * the gear ratio, and that's usually what the controller needs to know.

If the controller has the dropdown disabled because it can't handle more than 8 poles (which seems odd) then it's possible it won't work with your motors.
 
Glyn said:
Speed in both directions looks the same to the naked eye.
I'd expect the control software to be able to read the RPM, if it talks to the controller "live", and display that on a basic info screen (along with stuff like how much current the controller is drawing, batteyr voltage, etc).



Questions arising:
  • Although the motor seems to be working brilliantly, when looking at the photo above which shows the "burnt" coils, should I be concerned?
I can't see that image, just a blank spot. Try attaching it to the post.


  • There are 36 possible phase/hall combinations. The one which worked for me was only the 5th in the list. Is there any point trying the other combinations if this combination appears to be working? I guess what I'm asking is, is there a way to be absolutely certain that this is the best combination?
By testing all the combinations (though you don't really have 36, since most of them are the same ones shifted forward or backward in phase around a circle--see the previously mentioned thread and it should have info about that too) and finding the one that satisfies the requirements of correct direction, lowest no-load current, smoothest operation, etc. ;)

But most likely it's the right one based on the info you've provided so far.
 
I've put a clamp meter over the positive lead between the controller and battery and the largest reading I was able to get was 6 amps.

Photo of burnt coils attached:
burnt coils.jpg

I've just sent an email to MAC to request another motor. Ultimately, I don't want to risk using one that might fail halfway up a mountain.
 
Glyn said:
I've put a clamp meter over the positive lead between the controller and battery and the largest reading I was able to get was 6 amps.

I'm surprised the kelly software can't tell you what's happening. :/


Photo of burnt coils attached:
burnt coils.jpg

I've just sent an email to MAC to request another motor. Ultimately, I don't want to risk using one that might fail halfway up a mountain.
Yeah, it might be ok, but if it isn't, it could fail once it gets hot again under load. I don't know how that motor is wound but it could just be a single phase that's fried (not sure how that would happen, since current always flows thru at least two phases at a time--even a locked rotor stall should burn at least two sets of windings).

FWIW, though, I've seen worse burns that poeple kept using perfectly safely.


Interesting to note they have a spot for an NTC 10k thermistor (not installed) on the PCB, as well as what is probably an LM35 (populated) temperature sensor. So you can monitor temperature in there if you want to--I don't know if the kelly can do it or not (and autorollback power if it gets too hot), but there's threads about using those sensors with external voltmeters/etc to manually monitor temperatures if you want to.
 
amberwolf said:
Interesting to note they have a spot for an NTC 10k thermistor (not installed) on the PCB, as well as what is probably an LM35 (populated) temperature sensor. So you can monitor temperature in there if you want to--I don't know if the kelly can do it or not (and autorollback power if it gets too hot).
Yes, the Kelly controller does have this functionality but I'll need to change the thermistor as the controller will only work with KTY84-130 or KTY83-122 thermistors.


amberwolf said:
Glyn said:
I've put a clamp meter over the positive lead between the controller and battery and the largest reading I was able to get was 6 amps.
I'm surprised the kelly software can't tell you what's happening. :/
It has the ability to connect to an analogue ammeter (the ones with the needle). However, it uses the same pin as the thermistor so you can only use one or the other.

The current limit is programmable though so I guess setting that to a reasonable value and using the pin for a thermistor makes the most sense.
 
Glyn said:
It has the ability to connect to an analogue ammeter (the ones with the needle). However, it uses the same pin as the thermistor so you can only use one or the other.

The current limit is programmable though so I guess setting that to a reasonable value and using the pin for a thermistor makes the most sense.

What I meant was that the programming software ought to be able to read the values as they change, realtime, so you can use the computer you'r eusing to program it to read things during testing. I guess theirs simply may not have these features, but it would certainly be handy, since the controller already knows all this stuff and there is communication already in progress between the comptuer and controller.


The other method would be useful while on-road (where keeping a computer connected wouldn't be practical), but there are better ways for that, like a cheap RC wattmeter that actually monitors volts/amps/watts/wathours/amphours and remembers this info for teh whole trip (as long as power is not disconnected), so you can see how it performs both as you go and at the end. You'd only need one if you're monitoring battery usage as a whole, though if you want to know how each controller is doing you have to have six. ;)
 
Thanks Amberwolf. Unfortunately this controller doesn't provide any kind of live data via it's USB connection. Neither does it perform any kind of data logging. For interested parties, I've started writing a Definitive Guide to Programming the Kelly KBS48101X Controller.

An RC wattmeter sounds perfect for my needs. Eventually I plan to use a Raspberry Pi or similar for functionality such as home-brew traction control and data logging etc, but I've got a long, long way to go before I reach that point.
 
I just came across this little device which looked so good I thought it was worth sharing: The Eagle Tree eLogger V4

elogger-overview-med.gif


Although it's been designed for RC airoplanes, some of the features which I thought would be really useful for my application:
It's good for up to 80v and 180a and has Live data logging via USB. Importantly, you can hook up a GPS, inclinometer and thermistor, and record throttle position and motor RPM.

It can then plot graphs to show your current usage, voltage drop, throttle, rpm, inclination and temperature (max 4 hours duration), and display the graphs along side a map of your route. This sounds like a great way to understand what might be causing current spikes.

Some of the other cool features
  • Adjustable logging rate (from 50 samples per second to one sample every 5 minutes)
  • All data logged to permanent memory, which retains the data even when the power is removed
  • Accepts optional inexpensive sensors for three Temperatures, RPM (brushless, magnetic or optical), Airspeed, Altitude, GPS, Servo Current, and more
  • Fully Compatible with Windows 7 (32 and 64 bit), Vista, 98SE, ME, Win2K and XP™
  • Weighs about 0.8 oz (22 grams). Versions with Wire Leads may weigh slightly more.
  • Comes with our state of the art Windows virtual playback display and graphing software
  • Powerful Graphing software has advanced charting features, such as amps vs volts
  • Internet Updatable firmware
  • Full support for English and Metric units.

If you're in the US, you can find them on ebay for $30.
If like me you're in the UK, you can get them from Hobby King, new, for £42.04 (description says 100a version but photos are of the 150a version?).
 
Glyn said:
Although it's been designed for RC airoplanes, some of the features which I thought would be really useful for my application:
It's good for up to 80v and 180a and has Live data logging via USB. Importantly, you can hook up a GPS, inclinometer and thermistor, and record throttle position and motor RPM.
Keep in mind it's little LCD is *just* a glass LCD wafer, and fairly fragile. You'd want to make a good housing for it to keep it safe in use on the crawler.
 
Back
Top