BL motor acts as worn brushed motor...

[DCLXVI]

I'm running a Sphere ESC with a 3-star motor that recently stopped running as it should - it not acts as a brushed motor with worn brushes IE, it needs a push to get going and then it runs more or less as normal...
I've resoldered the sensor-cable and measured it so I know all the sensor wires has a connection (could I have reversed the connectors?)
What else could be amiss?

[pcnutz]

prolly poped as few power fets on esc

[Grizzbob]

There's still a few possibilities, first off, I'm not familiar with your particular motor, so I wonder, does it have a sintered rotor, or an older style bonded one(a sintered rotor will basically be nickel-plated, while a bonded one will look a bit more like the magnets of a brushed motor, same kind of material)? I ask this because in my own experience with a bonded rotor(using one in my Novak 13.5, when my local track would only allow us to use the motor with that instead of the sintered rotor), after it heated up several times, i.e. hard racing, its amp draw rose significantly, & it could draw enough that it could make my batteries' voltage drop too far on starts & cause a kind of glitch where the battery couldn't supply enough juice to run the ESC, receiver, servo, etc. for a split second, & would cause that hesitation, which behaved just like a hung brush in a brushed motor. And the other thing I'd question is the battery you're using, is it a good, race-quality one, or some unmatched stick pack you got for $20 at your local shop? Or beyond that, is your battery very old? If it's old, or just really cheap, then it might not be able to handle the amp draw of that motor any more, & that would cause the same problem as I described it. IF any of that might be the issue, then the first thing you could try(if you haven't already), is to try a capacitor plugged into the receiver(Spektrum & Novak sell some, & maybe a few others too). What they do is draw a little energy from the system's BEC(just like the receiver & servo does), & when the BEC's output voltage drops too far to maintain everything, the capacitor will discharge, keeping the servo & receiver supplied with enough energy to maintain control for a moment(not long at all, but usually just enough for the amp draw of the motor to go down so your car's battery can handle the rest)....:cool:

[DCLXVI]

Hmmm....

The thing that bothers me the most is that the *only* thing that I can notice that there's something amiss with my motor is that the car won't start from a standstill, I just have to give it a little nudge and I can accelrate away and it still have the same performance...to me that shouldn't be if the FET's are damaged...
But perhaps I should look into that amp draw thingy...

Btw, the motor is a LRP/Associated Reedy 3-start (their first that came out in 2005-2006)

[Demon-TC3]

Sounds to me like the sensor isnt kicking in. make sure your esc is set to sensored brushless, if it is, either the sensor port is gone or the motor's sensors are gone. its a really strange and difficult thing to check unfotunately but its acting exactly as a sensored motor would do if it was on a less-tuned sensorless esc.


take care,
Demon

[chilledoutuk]

it might be worth taking apart the motor cleaning it and reassembling it. I would remove any shims if you have added any as when i shimmed my ss5800 a while ago it seemed to run worse that when the rotor had lots of slop so i took them out.

if it still is problematic you could try a new sensor cable as far as i know the lrp motors have a detachable sensor cable (oh make sure the contacts are clean as well).

Your problems are a good example of the main negative point of running a sensored system ie less reliable and harder to diagnose when things do not go well.

[GSMnow]

I agree, it sounds like a sensor issue. Carefully check all the sensor cable wires for any damage. Do you (or do you know anyone who might) have a scope? You can use a scope to measure the signals from the sensors. The A, B, and C sensor lines should all have a very clean 0.0 or 5.0 volt level on them, and as you turn the motor, they will all toggle high/low at sifferent times as the rotor passes each one. Even with the trigger at neutral, you should see the sognals when you turn the motor by hand. The one true benefit of a sensored system is that it KNOWS the rotor position, even when stopped or coasting. But if one sensor is not working, it does not know and can't start, but if at least 2 are working, it will continue to run once started, but it will be down a bit on torque as it will basically be coasting for 16 to 1/3 of a turn. Assuming the motor wonding are all good and the wiring from the FET's to the motor is all intact, it could also be a dead bank of FET's. The symptom would be the same. To better test the FET's themselves, you can set it for high power brushed mode with a mild brush motor. Instead of connecting the three output wires together though, try one at a time to the MOTOR- with the MOTOR+ going to BATT+. All three should give you drive and brake. This will show if all three banks of FET's are able to pull to BATT- (drive) and to BATT+ (braking). If they all work, then all six banks have at least one good FET.

[DCLXVI]

I agree, it sounds like a sensor issue. Carefully check all the sensor cable wires for any damage. Do you (or do you know anyone who might) have a scope? You can use a scope to measure the signals from the sensors. The A, B, and C sensor lines should all have a very clean 0.0 or 5.0 volt level on them, and as you turn the motor, they will all toggle high/low at sifferent times as the rotor passes each one. Even with the trigger at neutral, you should see the sognals when you turn the motor by hand.

Would it be possible to see this on a multimeter?

Thanks for all the replys, I just have four things to test right now: Battery, motor, ESC and sensor-wire...though, now I know what to look at on each item...

[chilledoutuk]

you would be able to test with a multimeter as you should see the 5v or 0v signal alternate as gsm mentioned.

[kufman]

definitely sounds like a sensor problem. I had the same problem with my ss5800 when one of the wires was broken right down by the esc connector. the wire insulation was still holding, but the conductor was broken.

[Tmaxxinman]

It isn't a sensor issue i don't think.
If even one sensor is flawed or not connected the Novak or LRP ESC's will not give the motor ANY power.

It seems like it's a loose magnet in the motor. If a motor looses a magnet the ESC will not notice, and it will act like this. But it will still work, because 7 of teh 8 magnets will still be running.
It's like a V8 not firing on a cylinder, it still works, but power just isn't there and it does not run smooth.

So, crappy power, sometimes need push-starts... definetly sounds like a magnet that broke loose on the inside of the motor can.

-Johnno-MX-

[DCLXVI]

I just did some checking with a multimeter and all the sensor connector pins on the motor has a connection with each other (resistance is about 500 omhs) exept for one of the pins with is not connected at all...
I'm starting to suspect that this might be the fault...

Btw, would it be possible to run the motor on a sensorless ESC? Could help with my error diagnosing...

[Demon-TC3]

Yes, get the motor running on a Mamba max with 1.17 Firmware and see what happens. should run just fine if it has a sintered rotor.


good luck,
Demon

[kufman]

It isn't a sensor issue i don't think.
If even one sensor is flawed or not connected the Novak or LRP ESC's will not give the motor ANY power.

It seems like it's a loose magnet in the motor. If a motor looses a magnet the ESC will not notice, and it will act like this. But it will still work, because 7 of teh 8 magnets will still be running.
It's like a V8 not firing on a cylinder, it still works, but power just isn't there and it does not run smooth.

So, crappy power, sometimes need push-starts... definetly sounds like a magnet that broke loose on the inside of the motor can.

-Johnno-MX-

This is not correct. There are only 2-poles of one magnet to make up the rotor on this type of motor. There are 3 sensors that are aligned 120* from each other. If one sensor is broken, the controller will never get a signal from that sensor. If the rotor comes to a stop under that sensor, the controller can't figure out how to start the motor because it cannot determine where the rotor is. Sensored controllers are not like sensorless. If they can't figure out the rotor position, they don't make a guess, they just don't do anything.

[SS Pede]

I've heard of similar problems being described as a "burnt phase" or some such thing, like one phase isn't firing properly. Although you would think a dead phase would severely hurt performance even after the rotor is turning. Would this be the same as the sensor problem you are all describing or is it different?

[GSMnow]

In a Delta wound motor, a single burt open phase would probably still run with reduced torque but it might have a small dead spot where it won't self start. A "Y" wound motor with a burnt open phase would have a much larger dead spot where it won't start and will lose much more torque even if push started. The ROAR rules actaully mandated "Y" wound, not sure if that has changed in the latest update though. The symptom will be very similar for any failure that kills power to a phase of the motor. It does not really matter if it is an open wind in the motor, a fried FET bank, a bad connector to the motor, or (on a sensored system) a failed sensor signal which could be due to a bad hall sensor, out of position rotor, a break in the sensor cable, or a damaged sensor decoder in the ESC. Any of these conditions will very likely cause a brushless motor to fail to start on it's own and will reduce torque when running due to not having any torque for part of the rotation. Imagine if a plate of the comutator was missing in a brush motor. This would do the same thing as it would have no power for 1/3 of the rotation.

I think a company needs to come up with a few diagnostic toold for brushless. I would do this myself, but I certainly don't have the means to go into production.
1) motor sensor checker.
A small box, maybe self powered, or even better, powered by the ESC sensor cable with LED's. When you turn the motor by hand, you should see the 3 LED's all step in order so you know the sensors are working.

2) ESC output tester.
This one needs 6 LED's. As each phase gets energized to Batt+ or Batt- it will light a different LED. This would be very easy for a sensored system as you just pull a little throttle and slowly turn the motor and it should step through the phases as the sensors switch. Sensorless will be trickier as you need a way to make the ESC step aong, but slow enough to watch.

3) brushless motor winding tester.
Due to the very low internal resistances, this will be tricky to make accurate. A Delta wound motor will show continuity between each pair of wires, even with a winding totally open, and a shorted winding is almost impossible to detect. To get around these issues, I think it would have to test at a fairly high current and compare the voltage drop between opposing pairs. This will most likely require a microcontroller to power each of the 6 steps and store the results to compare them. The wide range of motor internal resistances makes this a little tricky to make it give a reliable good/bad reading on even just 540 sized motors, let alone if people want to use it on much smaller or bigger ones.

I know how to do these tests with my test bench having a variable power supply, an array of powr resistors, and a good scope and meter. With these devices, anyone could test their system right at the track.

[kufman]

Device #3 could get easier if you use an AC voltage to do the test. The impedance will be higher at a higher frequency and the device would not need to push as much current. Z=jwL You could probably build this device using a multimeter, and LM555 and an OPAMP.