Data Meter
Jun 24, 2010
Data Meter
My electric truck motor started stuttering.
I'm afraid that the motor is about to fry again.
I spoke with a few people who are extremely knowledgeable about this setup.
There is no consensus about the issue.
But, everyone asked a question that I could not answer.
That is, how much power is going to the motor?
I decided to see if I could answer this question.
Please see the other pages about this
project for more detail including the
Schematic and PCB design.
|
|
|
Summary
The basic plan is to rapidly measure the current and voltage using a PIC Microcontroller
and save the results. There were many challenges to this plan.
There were many technical challenges in this project:
- SPI to interface with SD Memory Card.
- USB interface in the Microcontroller.
- USB Linux kernel module to interface with the device.
- Isolated power supply to provide a floating ground.
- Communication protocol between multiple devices.
|
click to enlarge/reduce
|
Results
|
I designed the current sensor to operator on a shunt
that would yield up to a 50 mVolt drop.
To test this, I hooked up a resistor and a potentiometer
so that I could create a 0 to 50 mVolt drop across the
potentiometer. I then hooked this to my device.
Here is a graph of one of these tests. This graph consists of
71004 data samples recorded in 15.5 seconds.
The average sample frequency is nearly 4.6 kHz.
|
click to enlarge/reduce
|
I hooked the device to my truck and took it for a spin. I have a shunt in between the battery
and the power controller. This will only measure the current going to the controller. It does
accurately reflect the power going to the motor. Here are the graphs from two separate tests.
In both of these tests, I felt the motor stutter. I don't see anything surprising in the graph.
click to enlarge/reduce
|
click to enlarge/reduce
|
I ordered another shunt, some wire and some connectors. I assembled the second data acquisition
unit and got the two devices working together. The schematic shows how these devices
connect to the high current system. The current sensor unit measures the voltage across the shunt.
The voltage sensor unit measures the voltage across the motor. I mounted the devices, the shunt
and a few fuses to a board. The photo shows how these sit in the truck.
click to enlarge/reduce
|
click to enlarge/reduce
|
Data Collection
drive 4
I hooked the system up to the truck and went for a calm drive around the neighborhood.
Just before I pulled into the driveway, I heard a loud pop.
So, this graph should record an anomaly.
I decided the best way to view the two graphs is a scatter plot with an overlay.
You can select to show or hide the voltage or current graph by clicking the appropriate check-box.
Each graph shows about 4,400 points per second.
Narration:
| 20-40 | I backed out of my driveway. |
| 44-65 | Drove up a slight hill and down a slight hill. Came to a stop. |
| 75-90 | Went down a hill and around a curve. |
| 105-125 | Up a steep hill. |
| 128-140 | Took off from the hill into my driveway. |
At the end of my drive, I was stopped at the top of a hill.
When I accelerated from the stop, I heard the pop.
A 500 Amp current spike is visible toward the end the graph from Time=127.9 to Time=128.6 seconds.
The following shows that part of the graph in wider detail.
drive 5
The next day, I took another drive around the block the same path as before. When I took off at the top of the hill into my drive way, I heard a grind this time... not a pop, but a grind.
The sample rate on this graph is about 5kHz.
Narration:
| 20-40 | I backed out of my driveway. |
| 44-65 | Drove up a slight hill and down a slight hill. Came to a stop. |
| 75-90 | Went down a hill and around a curve. |
| 105-125 | Up a steep hill. |
| 128-140 | Took off from the hill into my driveway. Heard a grinding sound! |
drive 6
A few minutes later, I took another drive around the block the same path as before.
When I took off at the top of the hill into my drive way,
again I heard a popping, something louder than before. This spike is near 600 Amps!
600 Amps @ 155 Volts = 93 kWatts
The sample rate on this graph is about 5kHz.
Narration:
| 15-25 | I backed out of my driveway. |
| 30-55 | Drove up a slight hill and down a slight hill. Came to a stop. |
| 60-80 | Went down a hill and around a curve. |
| 90-110 | Up a steep hill. |
| 110-120 | Took off from the hill into my driveway. Heard a loud pop! |
drive 7
July 4, 2010, I spoke with someone who suspected it was commutator flash-over.
That is, I arcing occurs between the brushes around the commutator.
I bypassed three of the batteries and reduced the source voltage from 165 to around 140.
When I was on the hill, I also tried first gear instead of second.
I heard two quick pops and saw the gear shift jump violently each time.
This corresponds to two spikes in the graph.
The sample rate on this graph is about 5kHz.
Narration:
| 25-40 | I backed out of my driveway. |
| 45-62 | I took off in first gear. Drove up a slight hill. |
| 62-80 |
I shifted into second.
I pressed the accelerator for just a second since I was already going fast enough.
I then came to a stop.
|
| 82-115 |
Took off slowly in second. Drove down a slight hill and around a curve.
|
| 115-140 |
I did not come to a complete stop as I turned up the hill.
I arrived at the top of the hill, came to a stop, and shifted back into first.
For a moment, I forgot that the motor spins a lot longer than in a gas vehicle.
I let off the clutch and the vehicle jerked.
This jerk was all my fault, but it does not appear in the graph.
|
| 140-160 |
Took off from the hill in first gear.
I heard two loud pops and saw the shifter jerk each time.
I coasted into my driveway.
|
| 160-170 |
Yea, I don't know what that last spike is. Maybe I pressed the pedal one last time. |
Here is the last few seconds in wider detail.
|
