Testing Hall Sensors

14 posts / 0 new
Last post
mvadventure
Offline
Last seen: 14 years 2 months ago
Joined: Tuesday, July 22, 2008 - 06:07
Points: 3
Testing Hall Sensors

A recent ride through the rain has fried my Crystalyte Phoenix Cruiser (5304) motor and the controller. The controller stayed dry but the motor got very wet and apparently that is what fried the controller.

Have already got the motor apart, cleaned everything and tested the wiring on the little plastic board where the hall sensor wires actually connect to the hall sensors and have continuity. Following directions on an Endless Sphere thread I tried testing the Hall sensors but to no avail, my results simply don't match with either good or bad examples. Odds are the Hall sensors are fried, I've ordered a new motor and controller but truly don't want to put the new motor on if this one has a snowballs chance of being good. I've heard/read that a fried motor will burn out a new controller instantly so testing it by installing the new controller and testing it is a good way to ruin a new expensive controller.

Any help on how to effectively test the hall sensors? Step by step?
Thanks,
Mike

ronin149
Offline
Last seen: 12 years 3 months ago
Joined: Tuesday, August 26, 2008 - 12:01
Points: 17
Re: Testing Hall Sensors

I can help if you are still interested. I need the pinout of the connector for my own use, we should be able to help each other.

ronin149
Offline
Last seen: 12 years 3 months ago
Joined: Tuesday, August 26, 2008 - 12:01
Points: 17
Re: Testing Hall Sensors

To test hall sensors, one needs to do a little wiring.
The hall sensors have what is called an "open-collector" or "open-drain" output. They are like a switch connected to ground on one end. They are open until they detect the required magnetic field, then they close, connecting their "output" pin to ground.
So what does this mean? It means that one cannot measure a voltage at the output pin; there is nothing to measure.
If all one wants to do is check that the sensors are functioning, then the "diode check" function on a multimeter will do. Connect the black lead from the multimeter to ground, connect the red lead to the output of the hall sensor. Rotate the motor. When the hall sensor is active, it should show anything from .700 volts all the way down to zero volts (for an open-drain.) When it is not active, the multimeter will show "open," usually a 1 followed by blanked digits. It will be the same reading as when the multimeter is not connected to anything.
If one needs a better signal, say to feed an oscilloscope, then a pullup resistor is needed. These may actually be installed in the motor, but I have never seen it. They are usually on the controller board. Connecting the motor to the controller board should not only give power, but the requisite pullups are also active. If not, one will have to temporarily install some. The resistor need only be 0.25 watt. 10K ohms is ideal, but anything from 2K up to 1M should do the trick. It needs to be connected between the positive supply (for the hall sensors) & the output to be tested. Then, the output will show the positive supply when not active, and ground (or close to ground) when active.
Each hall sensor should change states smoothly. An individual sensor should be active roughly one third of the time as the wheel is rotated, in evenly spaced increments. All three sensors should cascade their outputs in order. I know of two sequences. The only motor I have ever seen that does not use one or the other is a custom unit I built (Yes, I had my reasons! But I digress...)

30 Degree spacing:
U-V-W
0 0 0
0 0 1
0 1 1
1 1 1
1 1 0
1 0 0

60 Degree spacing:
U-V-W
1 0 1
1 0 0
1 1 0
0 1 0
0 1 1
0 0 1

The spacing is the number of degrees between the sensors. U, V, &W are the sensor outputs. An active sensor is indicated by a "0." An inactive sensor is indicated by a "1." Notice that only one sensor changes at a time. This eliminates ambiguity in position detection; the motor would still run well if one or more sensors were not exactly where they should be. Of the two, I prefer the 60 degree spacing, as it has no valid combination for "all sensors off." This means that if you forget to connect the hall sensors, the motor will not try to run. The 30 degree motor will assume it is in position four and try to start.

Krilson
Offline
Last seen: 5 years 1 month ago
Joined: Tuesday, September 9, 2008 - 15:17
Points: 15
Re: Testing Hall Sensors

Question;
How many poles is this motor?

Thanks

Bingo Sun Noon

ronin149
Offline
Last seen: 12 years 3 months ago
Joined: Tuesday, August 26, 2008 - 12:01
Points: 17
Re: Testing Hall Sensors

Krilson,

I don't know. One does not need to know to test the sensors, if that is your concern.

mvadventure
Offline
Last seen: 14 years 2 months ago
Joined: Tuesday, July 22, 2008 - 06:07
Points: 3
Re: Testing Hall Sensors

Thank you for the input. Now that some time has passed and I'm back on the road I'm not sure the motor fried the controller, it could be that the controller got wet (even though it was well sealed) and caused the problems all by itself. After I disassembled the motor though the wires from the hall sensors where they join in the center looked a little burnt but still showed continuity. Not knowing what to do I ordered a new motor and controller from Electric Rider and a new watts up meter from another vendor.

While waiting on the the new motor the old one was disassembled and in my air conditioned shop. I had posted this same problem on Endless Sphere and received several ideas and suggestions to testing the Hall Sensors so when the new stuff finally got here I installed the new controller and put the old motor back together and on the bike. I tested the sensores using a volt meter prior to hooking it up to the new controller and the tests were fine. With a little trepidition I hooked it up to the controller and punched the throttle and it ran fine and is still running. I was able to return the "extra" motor to Electric Rider with no penalty.

The Crystalyte Phoenix system is absolutly fantastic as long as it doesn't get wet and while it's often difficult enough not to go into the rain when you have to be somewhere it's impossible if you're five miles from nowhere and the rain decides it's time to come down. I've gotten away with light rains several times but when everything got fried it was a real downpour and sustained at that. I try to be more careful now.
Mike

Krilson
Offline
Last seen: 5 years 1 month ago
Joined: Tuesday, September 9, 2008 - 15:17
Points: 15
Re: Testing Hall Sensors

Hi
Good advice all.
Be sure to keep in mind that the aforementioned truth table is electrical degrees.
You need to know the pole count to convert to mechanical degrees and to calibrate the system.
A 10 pole motor will repeat the pattern every 36 mechanical degrees or ten times per revolution of the wheel.
This is more than you need to know to simply test the halls. If you test them off the bike, (isolated) be sure to add a pull up resistor between Vcc and the output pin. 1k is good.
Also, it is possible that the timing can be adjusted to suit your riding, that is, advanced for flat high speed work and retarded for hills. You really need an oscilloscope to do much with this, though. Hard to kill a hall effect sensor.

sapo
Offline
Last seen: 8 years 8 months ago
Joined: Friday, April 23, 2010 - 22:30
Points: 10
Re: Testing Hall Sensors

When I was wheeling my bike up a hill using the throttle to assist the climb, the motor failed. Diode testing indicates that all three sensors have blown. Can you explain why?

www.beyond-oil.com for electric bikes and scooters.

ronin149
Offline
Last seen: 12 years 3 months ago
Joined: Tuesday, August 26, 2008 - 12:01
Points: 17
Re: Testing Hall Sensors

If I had to guess, I'd say it was inductive kickback from the motor. If this caused a surge in the power supply for the hall sensors, they could be damaged. The speed control is supposed to take care of such surges. But the operating conditions you describe would cause the worst kickback.

Check the power supply for the hall sensors to make sure it is working properly before replacing them. It should not be more than 15 volts (for most sensors.)

sapo
Offline
Last seen: 8 years 8 months ago
Joined: Friday, April 23, 2010 - 22:30
Points: 10
Re: Testing Hall Sensors

Thanks for that.

What action would minimise the risk of inductive kick back?
For example,
keep a steady throttle position,
no blipping the throttle,
don't twist the throttle until the bike is moving,
do not use a lot of throttle when the bike is moving slowly,
never apply throttle if the bike is rolling backwards?

I can't get my head around the dynamics creating a large inductive kick back?

Any help appreciated.

www.beyond-oil.com for electric bikes and scooters.

ronin149
Offline
Last seen: 12 years 3 months ago
Joined: Tuesday, August 26, 2008 - 12:01
Points: 17
Re: Testing Hall Sensors

You seem to already have a good grasp of what will help; everything you mentioned could help a lot.

However, it boils down to speed control design.

If it was mine, I'd replace the hall sensors (after checking the power supply.) If it didn't fail again, I'd forget about it. If it did, I'd get a new speed control.

Inductive kickback is caused when current through a coil is interrupted. This happens many times a second in a brushless motor. It is mitigated by the speed control (ideally.) The higher the current through the coil, the higher the kickback.

Keep in mind that I am guessing, hard to say what's wrong without it in front of me. It could be that your whole controller failed due to heat or some other problem. You will need to replace the hall sensors or try another motor to be sure.

sapo
Offline
Last seen: 8 years 8 months ago
Joined: Friday, April 23, 2010 - 22:30
Points: 10
Re: Testing Hall Sensors

A basic hall effect sensor is not like a transistor ... It does not have a diode junction. It cannot be tested via the diode check function on a multimeter.
A hall effect sensor has a voltage at it's output that is produce by a current flowing in a magnetic field. The output voltage is usually amplified before it is output.
I wasted a few hours trying to figure out why I wasnt getting a reading on my diode setting. I hope this saves others that pain!

www.beyond-oil.com for electric bikes and scooters.

Mick46
Offline
Last seen: 8 years 1 month ago
Joined: Friday, June 26, 2009 - 01:57
Points: 55
Re: Testing Hall Sensors

A good rule of thumb, mvadventure, is to invest in some waterproofing materials
if in fact you plan on riding in the rain. This will save you much $$$ and
time knowing your system is safe. I do not recommend leaving the controller with
no air, so when you get to your first destination let the controller "breathe"
inside while the motor has a chance to cool as well.

Dauntless
Offline
Last seen: 11 years 5 months ago
Joined: Thursday, May 27, 2010 - 16:20
Points: 220
Re: Testing Hall Sensors

Ummm, if it was me, I wouldn't want to leave the old motor unreplaced if there was a snowballs chance in hell that it was going to fry the new controller. You run a test and you don't get a successful reading, but you don't get the failure reading you expect. That means it's broken differently than what that person who told you to expect that other reading was thinking about.

Now, if a bad controller was hooked up to a new motor, could it fry the motor? If not, maybe you can test the old controller that way, just in case it's fine and the motor caused all the trouble. If that too is a hazard, forget the whole thing. Unless you really want to gamble that new conrtroller on that old motor. (How many bricks do you need for this house you're building?)

Meanwhile, if anyone knows how to fix that motor (If it was bad as he thought) and controller, I'd find that answer interesting.

WHo dares, WINS!!!!

Log in or register to post comments


Who's online

There are currently 0 users online.

Who's new

  • Nikson420
  • mhufnagl
  • Jamesato
  • VectrixPTLx
  • Schnabelwesen

Support V is for Voltage