SHENZHEN Controller Bafang Geared Motor Testing

[Knuckles]

So I just bought a permanent magnet gear reduction (PMGR) planetary-gear brushless Bafang Motor. This motor is in Shenzhen, China in Keywin’s lab. It will stay there until we have thoroughly tested it on the ecrazyman 72V controller. We are trying to reproduce the ‘jitter’ and ‘shift happens’ phenomena as reported by fechter’s most excellent testing.
See viewtopic.php?f=2&t=4282

This is the test motor ...

Bafang_Geared_Motor.JPG

(153.3 KiB) Downloaded 1901 times

We will post our results (good or bad) here.
Maybe we can get to the bottom of the PUMA mystery once and for all.

cheers

[Knuckles]

Well Keywin tested it at 72V (exactly 72V) and it ran super fast "like jet engine" says Keywin. Draw 0.8 amps steady at high speed no-load test. We figure 400 rpm wheel speed and 3000 rpm rotor speed. Keywin reported that the motor constant (K) = 50 rpm/V. so 50 x 72 = 3600 rpm max rotational rotor speed at 72V. Testing continues of course.

-cheers

[The7]

How many pole-pairs are there in this motor?
What is output frequency from the controller?

The frequency for 1 pole-pair (or 2-pole) motor at 3600 rpm is only 60Hz.
The critical frequency as found by Fechter is about 325 Hz. So this motor needs to be 5 pole-pairs or more to reach the critical frequency at 3600 rpm.

[Knuckles]

I'll find out.

[The7]

It looks as if this is the motor.

This motor has 10 pole-pair. The gear ratio when estimated from picture is about 5:1

If the wheel speed is 400 rpm,
the rotor speed = 400 x 5 =e 2000 rpm.

Then the power frequency = 2000 x10 /60 = 333 Hz.

Now it will reach the critical frequency as per Fechter.

So it is important to know the wheel rpm and gear ratio OR the power frequecncy.

[rkosiorek]

Was this motor previously marketed using the names PARN or TARN? it looks very much like a motor that was tested by various Crystalyte dealers 2 years ago.

rick

[Knuckles]

How many pole-pairs are there in this motor?
What is output frequency from the controller?

The frequency for 1 pole-pair (or 2-pole) motor at 3600 rpm is only 60Hz.
The critical frequency as found by Fechter is about 325 Hz. So this motor needs to be 5 pole-pairs or more to reach the critical frequency at 3600 rpm.

OK. Keywin says ...
... There are 10 pole pairs
... The gear ratio is 1:4.3
... and "i am sorry, 2400 rpm at 48v is not bafang. 1720 rpm at 72v is bafang motor"
(hopefully this data is much closer to correct than before ... gotta LOVE China!)

I guess that means the motor (rotor) constant K = 1720rpm/72V = 23.9 rpm/V

Heck I own the darn test motor. You want I tell Keywin to try a no-load test at 100+V? Just to see the jitter?

btw ... this statement by you (The7) "Apparently the board of this controller was designed on the understanding that the power frequency will never exceed the critical frequency (about 325 Hz) which is valid for gearless hub motors."

Well you may be on to something. Keywin writes the code and he got quite 'perplexed' by this concept.

HMMM. The plot thickens!

sherlock.jpg (7.81 KiB) Viewed 3032 times

Elementary My Dear Watson! Quite so Mr. Holmes! Ay, that's remarkable! (But WTF are you talking about?)

[The7]

OK. Keywin says ...
... There are 10 pole pairs
... The gear ratio is 1:4.3
... and "i am sorry, 2400 rpm at 48v is not bafang. 1720 rpm at 72v is bafang motor"
(hopefully this data is much closer to correct than before ... gotta LOVE China!)

I guess that means the motor (rotor) constant K = 1720rpm/72V = 23.9 rpm/V

1720 rpm for 10 pole-pairs => 1720 x 10 / 60 = 286 Hz which is lower than the critical frequency by 13%.

Perhaps 82 V could run this motor into the critical frequency.

Heck I own the darn test motor. You want I tell Keywin to try a no-load test at 100+V? Just to see the jitter?

Why not? Would like to hear the results at 82 V as well.

btw ... this statement by you (The7) "Apparently the board of this controller was designed on the understanding that the power frequency will never exceed the critical frequency (about 325 Hz) which is valid for gearless hub motors."

Well you may be on to something. Keywin writes the code and he got quite 'perplexed' by this concept.

Glad if Keywin could tell us the following since he wrote the code: ( Otherwise I would be more perplexed)
a) Is there any overspeed limit program written in the code?
b) What is the expected limit of the operating frequency ?

[docnjoj]

So after all the math (my favorate subject) would it be fair to say that these gear motors would be fine with either Keywin or Crystalyte controllers at "normal" voltages, like 36-48??????
otherDoc

[fechter]

With the Puma motor (and I assume the BMC would be similar), I can hit the critical frequency at around 30v. Below this voltage, you won't have any problem.

[Knuckles]

So after all the math (my favorate subject) would it be fair to say that these gear motors would be fine with either Keywin or Crystalyte controllers at "normal" voltages, like 36-48??????
otherDoc

I think this is a fair statement for the Bafang PMGR motor. If you operate it below the critical frequency then all should be fine.

[Tiberius]

Hi Knuckles,

Is this a rear wheel or front wheel motor? Can you tell us the part number?
I may be in a position to run similar tests?

Nick

[Knuckles]

It's a front wheel motor. I don't have the part number (motor is in China now, not with me in NY) but you can see some of the numbers in the photo at top of thread.

I just had a thought. Would a 'pedal first' controller be immune from the 'critical frequency' limitation when running a PUMA or BMC motor?
Or would it suffer the same 'jitter' at a critical rpm?

Heck if a 50-amp 72V controller was made without the need for Hall Sensors (pedal first) and it did not suffer from 'the jitters' then this could be a nice alternative.
I posted this question here as well ... viewtopic.php?f=2&t=3896&start=135

[docnjoj]

The problem with "pedal first" is that U have to. If U R stuck on a hill in the wrong gear, U have to go back down just to turn on the motor! No thanks!
otherDoc

[tostino]

Yeah, i'd not like a pedal-first controller as a solution, because I am going to be making detachable pedals, and not running with them all the time.

That would be more of a hack-job rather than a real fix.