3rd QS 273 8KW burnt

kebekua

100 mW
Joined
Jan 18, 2016
Messages
49
Fellows, please help me!

This is the third QS273 8KW I cook in no time and quite unfortunately I'm neither a chef nor a millionaire.

Using same bike, same controller (ND841800 not maxed out), same settings, same battery. First setup peaked at 42KW, 2nd at 35KW and 3rd failed immediately at 30kw.
The first motor was 16'', I changed to 17'' with a much higher tire so I took a different winding to obtain a similar resulting speed and torque.

My first QS273 8KW (323T) lasted over 5 years, 30 000KM and failed for known reasons (corrosion caused the core to expand details in thread below).


The second motor (434T) had a set of hall sensor fail after a month, and burnt after just 5 months despite having power limited to 80% of what the previous motor handled. QS passed the motor under warranty. The new motor (434T) seemed to be working normally until I pushed it a bit. I popped one wheelie and that was it. The temperature sensor was always connected and controller limit set to 120C. Readings never exceeded 128C. Not RPM related, never went much over 100KMH (800RPM).

I'm really puzzled at what could be the problem. Could it be the controller? Is there a way to test it without risking to cook another motor?

The different windings could explain part of the problem, the 434T has 48 strains instead of 58 for the 323T and the coils are 33% longer. That means roughly 60% more resistance, and more heat, for the 434T. But that doesn't explain why the second 434T burnt almost instantly with less power than the previous one.

mmexport1718902394218.jpgmmexport1718902401918.jpg
 
Last edited:
Don´t you have a temperature sensor installed?
Or do you use the calculated motor temperature which could be selected in the app of the fardriver.
Are you using weakening field or have you turned it of in the app.

I had a Fardriver ND72680 and a 5kW QS260 motor in my 190kg heavy scooter, without FW I could draw up to 25kW without killing the motor, but with activated FW the motor got really fast very hot near Topspeed. Without FW I sometimes had the Fardriver to reduce power on longer uphill runs or on very hot days. The Fardriver reduces power when the motor gets over 140°C to nearly zero at 158°C and I think to "no" Power at 160°C.

Now I have a installed ND721800 and a QS72V10000W motor , but my bigger fardriver does not suport the temperature sensor which is installed in the motor. I hadt to install a parallel resistor to make the resistance fit the 140-160C range. I´m running this motor with 400bA and 1400pA with 21S (77V nominal). Under normal conditions the motor stays cold.
 
Last edited:
I assume those two wires next to the first hall sensor or the temp sensor mounted directly to the windings so that should be working, in the controller are there settings to set the type of thermistor that could be set incorrectly so it's reading the temp wrong or maybe the controller is set correctly but just reading the temps wrong. Either would be easy to test if you connect the temp sensor to the controller and heat up the sensor to a known value to check if it's correct. If not setting the temp limit lower and adding some cooling mods to the motor may be a good option.
 
If the sensor is connected to the controller the app should show ambient temperature before you start driving. If it shows strange values it could be set wrong in the app.
My 72680 had a lot of available sensors, my 721800 has fewer options available.
 
The temperature sensor works normally and the correct option is selected in the controller. As you can see on the pictures, the motor didn't heat up evenly, this is why the actual maximum temperature wasn't captured.
 
So, just to sum up:

-1st motor: 16 inches, 323T, 58 strands according to QS*, 42KW recorded peak -> worked for 5 years (rust)
-2nd motor: 17 inches, 434T, 48 stands, 35KW recorded peak -> worked for 5 months (phase burned)
-3rd motor: 17 inches, 434T, 48 strands, 30KW recorded peak -> worked for 5 minutes (phase burned)

All with correctly connected and configured temperature sensors.
Also, all with the same controller (ND841800)

I feel like this is a mystery for the real motor experts of the forum, hopefully they are still around? @Doctorbass @liveforphysics @John in CR @Arlo1 @APL ?

*I counted only 54 strands when I took the motor apart
 
Looking at the burn pattern, there is remarkable symmetry with about 8 windings or stator coils separating them, which is kind of reminiscent of a harmonic standing wave. Since each winding alternates phases as you go around the motor, if I count from each burned coil in sets of three, it seems that all the burned windings belong to the same phase. I'm leaning towards the controller putting out unequal phase voltages, causing the failure. Voltage imbalance among the phases leads to current imbalance, overheating and damage.

3 phase 2.PNG
 
How do I put this?

Ur doin it rong.

Is that motor 30kW to 42kW rated? Guessing not.
Rating is continuous not peak. If you read the post, it lasted 5 years so the setup itself ain't wrong. I have owned several builds with less powerful motors handling that range of peak power.
 
Last edited:
Looking at the burn pattern, there is remarkable symmetry with about 8 windings or stator coils separating them, which is kind of reminiscent of a harmonic standing wave. Since each winding alternates phases as you go around the motor, if I count from each burned coil in sets of three, it seems that all the burned windings belong to the same phase. I'm leaning towards the controller putting out unequal phase voltages, causing the failure. Voltage imbalance among the phases leads to current imbalance, overheating and damage.

View attachment 355482
Thanks, I'm also leaning more and more towards a controller issue
 
Can you measure the current of the three phases at the same time?
That would be the only way you could be shure not investing money in the wrong part.
Probably you know someone which has a Fluke 1736 power logger or something similar, or three identically current clamps.
 
Rating is continuous not peak. If you read the post, it lasted 5 years so the setup itself ain't wrong.
But the setup changed, according to your description.

You said your first setup lasted 5 years, which sounds great, even with 42kw peaks. Then:
The first motor was 16'', I changed to 17'' with a much higher tire
Bigger tire means the motor is going to require more torque (higher amperage) to achieve similar performance. And yet:
The different windings could explain part of the problem, the 434T has 48 strains instead of 58 for the 323T and the coils are 33% longer. That means roughly 60% more resistance, and more heat, for the 434T.
Therefore, by your description, you swapped to a motor with a 60% higher resistance, while also switching to a bigger wheel requiring more torque. I agree that QS motors have a reputation for being underrated, and I know the difference between continuous and peak ratings. However, it's still not surprising why the motor burnt. You're even admitting to pushing it:
The new motor (434T) seemed to be working normally until I pushed it a bit. I popped one wheelie and that was it.
Wheelies take a ton of startup current. So much so that the heat can rapidly rise in the coils, past their rating, before residual heat can make its way to temp sensors and the controller can engage a temp-related shutdown. That high temp will eventually make it's way to the thermistor, and it sounds like that is what happened:
The temperature sensor was always connected and controller limit set to 120C. Readings never exceeded 128C
So even though you set the temp limit to 120, it got to 128. Which is rather indicative of a rapid current draw and subsequent temp rise. It's entirely possible that this could be controller related, or faulty thermistor, sure. But given all the evidence you've presented, it sounds a lot more like you gave motors much more current than what they're rated for, even for short-term peak currents.
 
So, just to sum up:

-1st motor: 16 inches, 323T, 58 strands according to QS*, 42KW recorded peak -> worked for 5 years (rust)
-2nd motor: 17 inches, 434T, 48 stands, 35KW recorded peak -> worked for 5 months (phase burned)
-3rd motor: 17 inches, 434T, 48 strands, 30KW recorded peak -> worked for 5 minutes (phase burned)
,
The different windings could explain part of the problem, the 434T has 48 strains instead of 58 for the 323T and the coils are 33% longer. That means roughly 60% more resistance, and more heat, for the 434T. But that doesn't explain why the second 434T burnt almost instantly with less power than the previous one.

Hello,

Seeing the info provided on this thread I think is just too much amps on each phase. The difference between motor 1 ,2 and 3 is not that small at all.

For example if the wires on the winding are the same on all motor just the change from 58 from 48 strains is big, imagining each strain at 40 amps:
48x40= 1920amps
58x40= 2320amps
400 amps difference just in this. Please note this numbers are just examples, I’m not saying this is the limit of your motor.

Also you state the windings on the motor 2 and 3 are 33% longer and that would mean 60% more resistance. This would mean the difference between the 2 motors is even bigger in terms of how many amps they can handle.

Also I see you use a ND841800, but this controller can go to about 1800 amps or 1500 amps depending on the version, I don’t see any info on the post on how many phase amps you have been using.

About the temperature, well just don’t trust the temperature sensor if you see your motor can quickly heat up. The temperature sensor probably will not be right on the spot that overheats, if you have a quick temperature raise and the temperature sensor is not of the top of this hot spot, by the time heat gets to the temperature sensor is already too late.
 
,


Hello,

Seeing the info provided on this thread I think is just too much amps on each phase. The difference between motor 1 ,2 and 3 is not that small at all.

For example if the wires on the winding are the same on all motor just the change from 58 from 48 strains is big, imagining each strain at 40 amps:
48x40= 1920amps
58x40= 2320amps
400 amps difference just in this. Please note this numbers are just examples, I’m not saying this is the limit of your motor.

Also you state the windings on the motor 2 and 3 are 33% longer and that would mean 60% more resistance. This would mean the difference between the 2 motors is even bigger in terms of how many amps they can handle.

Also I see you use a ND841800, but this controller can go to about 1800 amps or 1500 amps depending on the version, I don’t see any info on the post on how many phase amps you have been using.

About the temperature, well just don’t trust the temperature sensor if you see your motor can quickly heat up. The temperature sensor probably will not be right on the spot that overheats, if you have a quick temperature raise and the temperature sensor is not of the top of this hot spot, by the time heat gets to the temperature sensor is already too late.
Looks like @Portugaline and I were both typing the same conclusions in different way, at the same time.
 
How can the temperature limit be set? I was not able to change it neither on my 72680 nor on my 721800.
I heated my 5kw QS Motor a few times beyond 150C and very often between 120-140C without any problem at the windings. A few months before I even melted and shorted the phase wires, but the winding still looks like new.

The fardriver only has two current sensors, the third phase current is calculated. If one sensor has an offset the controller wil not recognize this and the overloaded phase will probably heat its winding much faster than the other two.
 
How can the temperature limit be set? I was not able to change it neither on my 72680 nor on my 721800.
I think you can set it only in the computer programming interface, for some reason.
 
I should have chosen a 260 instead of a 273. 68x333 strains for the kv I want
 

Attachments

  • 260 winding.PNG
    260 winding.PNG
    10.4 KB · Views: 12
Back
Top