Testing NEOs?

Being a relative novice when it comes to motor technicalities, and after a bit of searching here to no avail, let me ask my question. I’m sure you’ll tell me if it’s a dumb question or not, regardless I’ll learn something. :slight_smile:

We used NEOs on our swerve this year and they got some pretty amazing workouts. Is there a way to test them to see if they’re still putting out the torque they’re supposed to? Or do they just not die? I see lots of ways to measure the torque of a motor on the web, is there a way some of you test them?

Thanks, Jim, mentor for 4153

Resistance across the coils is easy enough to check with a multimeter and compare to a known good motor, that’s usually about all you’ll need to confirm a motor is dead or on its way out.
I guess you can test stall torque easily enough with your choice of measuring device, but there’s some annoying variables to control for there, mainly surrounding the input voltage.

In FRC timescales, BLDCs don’t die unless you start to smoke them (there are other failure modes, but they are usually binary), and a DMM can tell you that.

3 Likes

Coil resistance is a good health check as Troy said.

Another is free speed. Multimeter probes across the correct two terminals at the back of the sensor connector (one Hall signal, and GND) with the meter set to read frequency will give you the electrical speed in Hertz; that can be converted to shaft speed in RPM, multiplying frequency by 60/7 because Neos are 14 pole motors. Compare this free speed to the nominal data sheet value, or better yet to that of a known-good motor you have tested.

You can also just listen to them. Their whine at free speed will be recognizable to an ear that is familiar.

4 Likes

Back when we ran 775 pros / redlines, we used to test motors by running them as generators. Spin then at a constant speed (using a drill and tachometer, or a second motor in a gearbox) and measure the output voltage. It gave a clear signal of a degraded motor.

I’m not sure if the same thing would work well with brushless motors, maybe using a scope? (I have not noticed any defecation degredation in Neo performance yet)

Not the OP but I have not noticed that coming from Neos. Can you tell by smell? :thinking:

7 Likes

They smell like hell when you let out the magic smoke :wink:

Yes, if you spin a NEO it will make AC voltage. The voltage output will be related to the speed you spin it at and the frequency will match the speed you spin at times the number of poles.

OK, test data (cross posted to my brushless testing thread too):
Makita drill, 1,300 RPM nominal speed
Motor #6 2 Volts AC Good motor
Motor #3 0.92 Volts AC This motor shows 0 inductance and is clearly shot
VERY heavy drag
Motor #4 0.85 Volts AC This motor also shows 0 inductance and is shot.
VERY heavy drag
Motor #5 1.96 Volts AC
And when I say “VERY heavy drag” I mean somewhat less than taking a three headed Anderson shorting bar and shorting the motor. Drag similar to ONLY shorting two leads, but without the pulsating torque you get from only doing two.
Note that motors 3 and 4 both run on a test SparkMax, just really slow due to excessive current draw preventing them from spooling up.

So, spinning a NEO and checking Voltage is a VERY useful test!

More test data and how to test work on NEO motors: Brushless Motor Controller Tester - #23 by Weldingrod1

4 Likes

I just chucked up a couple of NEOs in my Makita and looked at the “output” on my scope. One NEO had gotten really hot, the other was brand new. I looked at the red/black and white/black pairs and got a nice sinusoidal signal. Sinusoidal? I thought BLDC motors had trapezoidal back EMFs? Or are NEOs PMSMs? Regardless, the new motor put out 1.5V peak to peak, the “heated” one 1.2V peak to peak for the exact same drill RPM. Off to check all the other motors now.

4 Likes

The applied EMFs of many BLDC controllers is trapezoidal, to simplify the firmware. The back EMF is sinusoidal, provided by rotating geometry.

6 Likes

Thanks for getting an actual scope trace! If you have enough channels it would be interesting to compare all three phases for the smelly motor.

Oh, and I tried my NEO tester hooked into a motor and spinning the motor with a drill. I get a little light on the phase Voltage LEDs with a good motor and -no- light with a fried motor.

Scope has 4 channels, I just need to find more BNC cables to wire them up (I’m at home, not at the Bot shop). Dumb question, there are red, white, and black connectors/wires, I did measurements between red and black and white and black to compare. How exactly are the insides of one of these wired up? I didn’t see that in your other thread.

Three phase delta connected. One coil between each wire pair.

1 Like

I used to have a slide when I worked at TI with all of different names that people give to ‘motors that have three electrically commutated phases and no brushes’. I agree the common one is what you mention as BLDC = trapazoidal bemf, PMSM = sinusoidal bemf, but the words all mean the same thing (except maybe the ‘DC’ part of BLDC).

Nomenclature aside, I suspect if you zoom in more it’ll start looking a bit more ‘trapezoidal’. I suspect its as sinusoidal as it is due to the high number of poles, and the fact that these are outrunners. If you take the motor apart, the magnets are certainly not rounded like you find in a PMSM, and there is no special pattern to the windings. Either way these motors are good candidates for SVM and FOC as a replacement for simple trapezoidal commutation.

1 Like

12 slot 14 pole outrunners have relatively sinusoidal BEMF when measured line-to-line. And that’s usually the only way it can be easily measured, since there’s really no reason for the motor manufacturer to provide a neutral (star connection point) terminal. If you did measure line-to-neutral BEMF, it would appear more trapezoidal, and its wave form would line up when superposed with loaded current wave shapes, if both measurements were made at a reference speed and synched to the Hall sensor edges.

All of that is of mostly academic interest of course. The main point, as Gus and Will both pointed out, is that BEMF wave shape follows the geometry – specifically the geometry of stator teeth and rotor magnets.

2 Likes

I found that a brand-new NEO has approximately 0.6 ohms of resistance between the coils, and a brand-new NEO550 has approximately 0.5 ohms.
If anyone else also has data for this, comparison could be interesting.

Hmmm. I don’t have access to a decent milliohmmeter at the moment (fires in NM have us ready to evacuate!) but a quick check of our used ones with a DMM shows several closer to 0.2 ohms. Hmmmm. Yes, if anyone else has data for coil resistance on brand new NEOs, let us know.

1 Like

Y’all definitely need a milliohm meter to take that measurement. IIRC a NEO is closer to ~36 milliohms, and the NEO550 closer to ~62, milliohms.

5 Likes

Adding onto this, if you have a current limited bench power supply using constant current mode, you can run a few amps through it and measure the millivolts with a multimeter. Then use Ohm’s Law to find the resistance.

2 Likes

In that case, do you have any idea what would cause the high values I got? Just me using a regular multimeter on resistance mode instead of a milliohm meter?

You’re probably within the range of error for the meter itself. A milliohm meter will have 4 leads to allow four terminal sensing which becomes very important when the leads, connectors, and internal electronics have similar resistance to what you are trying to measure.

1 Like

Ok, that makes a lot of sense. Thanks.
Is there a way to measure this resistance without a specialized tool? Would code running on the RIO be helpful in any way?