Open Source Electric Supercharger

[jamesm]

Originally Posted by FT_Monk

How about using two big motors and gear it up to get the desired RPM. Big motor with lower speed has the torque to spin gears and draw less current.

For a given voltage, you're going to draw the same current for any given amount of total power, as power is current * voltage.

If I were building it i'd start with 3 geared 700-class heli motors and lipo power. That would ne simple and work for the purpose of getting all the hard parts (compressor, mounting, etc) figured out. Delay decisions on the electrical system until after everything is figure out and you know how much power you really need to drive the motors.

[neutron256]

Originally Posted by Bigmaxy

Bearing quality of the motor you choose should be a factor that you consider also. I have used Turnigy motors and Neu motors and the Neu are definitely a quality unit that are efficient and can spin happily at very high revs.
I've not used or researched the water cooled unit you have chosen though and Turnigy are getting better and better all the time so I'm definitely not saying anything bad about them. It could be an awesome choice.
As you know magnets don't like excessive heat either so a motor that is working in it's comfort zone will obviously prolong life. That water cooled jacket could be a good thing even though it adds a bit of complexity.
Good idea to give the charging system a miss for now also IMHO.
Great work!!

I fully expect to burn up at least one motor in testing, and when I do I think I'll be getting a water cooled one to replace it. The fact that it would be sealed also addresses any issues with air leaking past the motor is another big benifit.

Originally Posted by FT_Monk

How about using two big motors and gear it up to get the desired RPM. Big motor with lower speed has the torque to spin gears and draw less current.

Originally Posted by jamesm

For a given voltage, you're going to draw the same current for any given amount of total power, as power is current * voltage.

If I were building it i'd start with 3 geared 700-class heli motors and lipo power. That would ne simple and work for the purpose of getting all the hard parts (compressor, mounting, etc) figured out. Delay decisions on the electrical system until after everything is figure out and you know how much power you really need to drive the motors.

Gearing/multiple motors seem like a great idea on the surface, but it adds a huge amount of complexity to the system, and at these speeds you'd also have to consider lubricating any sort of transmission.

If you're going to add gearing it would probably be more practical to just get a centrifugal supercharger that all ready has gearing in it and attach an appropriatly sized electric motor to that. I considered this but the even used a working centrifugal supercharger starts at over $1000 and that's before you start adding things like a motor.

I actually ran into one of the guys from Grimmspeed last night at our local Rice Killers meet and we got talking about this. He remined me of something I had seen years back where a guy named Thomas Knight had taken three starter motors built a transmission for them and mated it to a twinscrew super charger. It was a very cool idea but didn't make it I'm sure because the cost was just too high for the benifit it provided. This is the same problem I fear Phantom may run into which would be really unfourtanate because it it's such a cool idea.

http://www.turbomagazine.com/tech/04..._supercharger/

[mid_life_crisis]

Starter motors are specially wound to be able to create huge amounts of torque for very short bursts. They are not designed for continuous running. This is why if you crank your car too long trying to start it you get that smell that warns you that you're about to let the smoke out of the wires.

Edit: Just looked at the article. He used the starters as the core for his own windings. They were no longer starter motors.

[Wild Weasel]

Interesting that the Thomas Knight stuff came up. For years, on my website talking about electric supercharger scams, I've referred to those as the only thing I was aware of that may have actually worked (until now, obviously). It bothered me a bit though. I mean... I don't doubt that they worked, but I don't actually know whether anyone EVER bought one! I've never heard of a single person buying and installing one and then posting about it on the Internet. Ever.

[neutron256]

Originally Posted by Wild Weasel

Interesting that the Thomas Knight stuff came up. For years, on my website talking about electric supercharger scams, I've referred to those as the only thing I was aware of that may have actually worked (until now, obviously). It bothered me a bit though. I mean... I don't doubt that they worked, but I don't actually know whether anyone EVER bought one! I've never heard of a single person buying and installing one and then posting about it on the Internet. Ever.

I had the same feeling. The first time I came across the Knight setup I thought "Cool how much does it cost?". Then I did some more searching and realized that there hadn't been any new news about it in several years.

I'm sure that it was just something he did, and even though it may have worked it probably wasn't viable as a commerical product. Most people don't know how hard it is to make a profit sellling something like this even when it is a really great idea.

[Wild Weasel]

The problem was that, at the time, it really wasn't a very good idea at all. It was an interesting experiment, but since it cost as much as a proper turbo and had the weight of a Hummer in batteries, there just wasn't any practical reason to actually want it. I only mentioned it because it did appear to work as designed... showing that the concept was viable and it acted as a great contrast to all the crap out there that isn't worth the bandwidth it takes to read about.

Technology has come a LONG way since then.

[SloS14]

Got a MAP sensor. I'm able to blow in it and get a reading so.....yeah. I need to brush up on my OOP.

I figure the motor needs 3 modes "Off","Cruise", and "TargetBoost", targetBoost (reads from array of RPM and or TPS)

[neutron256]

Originally Posted by SloS14

Got a MAP sensor. I'm able to blow in it and get a reading so.....yeah. I need to brush up on my OOP.

I figure the motor needs 3 modes "Off","Cruise", and "TargetBoost", targetBoost (reads from array of RPM and or TPS)

How were you thinking of getting RPM?

[SloS14]

Originally Posted by neutron256

How were you thinking of getting RPM?

right now it's just a POT. From the actual car, I'd have to do a tap I guess - If I decide to go that far. The code as-is is messy as hell. Really needs to be done with real OOP methods, etc. which I'm a bit rusty with. After coding for 8 hours straight, it's difficult to come home and do more. Right now it looks like this.

#include <Servo.h> //for ESC control
#include "pitches.h" //for speaker output




// This is our motor.
Servo myMotor;

String incomingString; //for COM READ

//INITIALIZE ESC SETTINGS
int motorPin = 9; //pin ESC is attached to
int minThrottle = 0; //in PWM, 
int maxThrottle = 180;
int currentThrottlePercent = throttlePWMtoPercent(minThrottle);
int currentThrottlePWM = 0;

//INITIALIZE ACCELERATOR
int accelPin = 2;
int accelValue = 20;
int minAccelValue = 0;
int maxAccelValue = 1023;

//INITIALIZE ENGINE SETTINGS
int engineCurrentRPM = 0;
float engineCurrentRPMsignal = 0;

//INITALIZE MAP SENSOR
int mapPin = 3;
int lastMapRead = 0;
int minMapMinRead = 650;
int maxMapMinRead = 810;


//GENERAL INITIALIZATION VARS
boolean ESCinitializedBool = false;


//SETTINGS
int targetMapRead = 750; //target boost


int speakerPin = 8;

void setup()
{
  myMotor.attach(motorPin);

  // Required for I/O from Serial monitor
  Serial.begin(9600);
  // Print a startup message
  Serial.println("initializing Program");
  if (ESCinitializedBool == false) {initializeESC();}
}

boolean initializeESC() {
       Serial.println("Initializing Motor. Please Wait...");
       Serial.print("Sending value of "); Serial.print(minThrottle); Serial.println(" To ESC");
       myMotor.write(minThrottle);
       delay(2000);
       Serial.println("Motor Initialized");
       ESCinitializedBool = true;
}

int throttlePWMtoPercent(int pwm) {
 int x = map(pwm,minThrottle,maxThrottle,1,100);
 return x;
}
int throttlePercentToPWM(int percent) {
 int x = map(percent,1,100,minThrottle,maxThrottle);
 return x;
}

int accelToPercent (int accelVal) {
  int x = map(accelVal,minAccelValue,maxAccelValue,0,100);
 return x;
}

void loop()
{
  accelValue = analogRead(accelPin); 
  
  // If there is incoming value
  if(Serial.available() > 0)
  {
    // read the value
    char ch = Serial.read();
  
    /*
    *  If ch isn't a newline
    *  (linefeed) character,
    *  we will add the character
    *  to the incomingString
    */
  
    
    if (ch != 10){
           
      // Print out the value received
      // so that we can see what is
      // happening
      Serial.print("I have received: ");
      Serial.print(ch);
      Serial.print('\n');
    
      // Add the character to
      // the incomingString
      incomingString += ch;
    }
    // received a newline (linefeed) character
    // this means we are done making a string
    else
    {
      // print the incoming string
      Serial.println("I am printing the entire string");
      Serial.println(incomingString);
    
      // Convert the string to an integer
      int val = incomingString.toInt();
    
      // print the integer
      Serial.println("Printing the value: ");
      Serial.println(val);
    
      /*
      *  We only want to write an integer between
      *  0 and 180 to the motor. 
      */
      if (val > -1 && val < 101)
     {
       // Print confirmation that the
       // value is between 0 and 180
       Serial.println("Value is between 0 and 100");
       // Write to Servo
       currentThrottlePWM = throttlePercentToPWM(val);
       Serial.print("Writing PWM to motor: ");
       Serial.println(currentThrottlePWM);
       myMotor.write(currentThrottlePWM);
       
       Serial.print("TPS %");
       Serial.println(accelToPercent(accelValue));
       
       lastMapRead = analogRead(mapPin);
       Serial.print("MAP read:");
       Serial.println(lastMapRead);
       
     }
     // The value is not between 0 and 180.
     // We do not want write this value to
     // the motor.
     else
     {
       Serial.println("Value is NOT between 0 and 100");
      
       // IT'S a TRAP!
       Serial.println("Error with the input");
     }
    
      // Reset the value of the incomingString
      incomingString = "";
    }
  }
  
  //START MAKE MUSIC
  tone(speakerPin,accelToPercent(accelValue)+80 ,64); delay(10);
  
  //END MAKE MUSIC
  
  //make motor follow acceelerator for now
  currentThrottlePWM = throttlePercentToPWM(accelToPercent(accelValue));
  myMotor.write(currentThrottlePWM);
  //Serial.println(currentThrottlePWM);
  //end make more follow accel
  
  delay(1); //lets keep things reasonable...
}

[neutron256]

Okay I'm almost as bad at reading code as I am writing my own, but what is that bit about IT'S a TRAP!" going to the display? LOL. I like it I'm just wondering what triggers that to show up.

[SloS14]

Originally Posted by neutron256

Okay I'm almost as bad at reading code as I am writing my own, but what is that bit about IT'S a TRAP!" going to the display? LOL. I like it I'm just wondering what triggers that to show up.

I copied several bits of code from other sources. Basically the way it works now is it initializes the motor, then awaits a serial input (legacy code) that once entered, will display the TPS, MAP inputs. Like I said, it needs a total revamp with proper OOP coding.

I just installed "Processing" which will let me view some things visually in real time. Ideally I'd like to write an engine emulator for testing, but that kind of sounds like a PITA.

The "it's a trap" would show if an out-of-range integer for the motor speed was input. Currently I have the "throttle" mapped to the motor.

moving the throttle lever and sucking/blowing into the MAP sensor I could get the rhythm of but I would really like the motor RPM to offset the MAP signal. probably a bit too cumbersome for a bench prototype. I'd rather be spinning a compressor and making actual pressure.


Edit: Also, is your wheel actually balanced? In my experience usually the retaining nut is shaved for that.

[neutron256]

Originally Posted by SloS14

I copied several bits of code from other sources. Basically the way it works now is it initializes the motor, then awaits a serial input (legacy code) that once entered, will display the TPS, MAP inputs. Like I said, it needs a total revamp with proper OOP coding.

I just installed "Processing" which will let me view some things visually in real time. Ideally I'd like to write an engine emulator for testing, but that kind of sounds like a PITA.

The "it's a trap" would show if an out-of-range integer for the motor speed was input. Currently I have the "throttle" mapped to the motor.

moving the throttle lever and sucking/blowing into the MAP sensor I could get the rhythm of but I would really like the motor RPM to offset the MAP signal. probably a bit too cumbersome for a bench prototype. I'd rather be spinning a compressor and making actual pressure.


Edit: Also, is your wheel actually balanced? In my experience usually the retaining nut is shaved for that.

Yep the wheel is balanced. They shaved a small spot on the underside, and on the top.

[neutron256]

More progress...

My second test print for the compressor backplate. The motor mounts nicely. Next I need to work on the recess for the compressor wheel.

I had an idea for mounting the wheel to the motor shaft need to think it through some more though...

[mid_life_crisis]

Originally Posted by neutron256

Okay I'm almost as bad at reading code as I am writing my own, but what is that bit about IT'S a TRAP!" going to the display? LOL. I like it I'm just wondering what triggers that to show up.

It doesn't actually display (from what is shown there) "It's a trap". That's a comment letting anyone look at the code know that it is what is called an "error trap". If the value returned is out of range, an "error", it will trap it there and post a warning message and not send the value on.