Conference vmszoo::rc

     I'M not sure about the Javelin but on my RC10 I use velcro.
    
    
    
    
    
         REIS
    

    On the Javelin/Optima cars the battery is under the car, help up
    to the chassis by 2 straps.

    
    
    I have a Javelin and several battery packs (Im seriously into racing).
    I have replaced the reusable tie wraps by heavy rubber bands.
    
    Make sure bands adjust tight enough to prevent any battery slipping
    while taking a sharp curve. I have made some from an old bicycle
    tube.
    
    Once you have practice it should take you under 30 seconds to swap
    battery packs.
                                                                 
    Hope this helps,
    Orlando.

    Thanks, great idea! I even have a couple of old bicylce tubes hanging
    around. Thanks Orlando.

    I recently bought a car (Kangaroo), and its a lot of fun.  One question
    though, the battery only lasts about 4 minutes.  The car has a stock
    540 motor and I use Kyosho 7.2 Volt 1200 ma packs that I charge with my
    Aristocraft quick charger.  I occasionally trickle charge the batteries
    (14 hrs???) and I always make sure that they are all the way empty
    before I charge. 
    Am I doing something wrong, or is the cheap speed control the culprit? 
    Or is this the way it is, I always thought that the batery is supposed 
    to last 8-10 minutes. 
    Any ideas? 
    

    Regards
    J�rg

    Could be the battery itself -- I bought a couple of batteries in
    Hong Kong, thinking I was getting a good deal, but they were'nt
    so good when I got them home.  
    
    Could be the way you are driving the car.  Mine will go close to
    15 minutes if I'm just driving straight around a track at high speed.
     Driving slower uses the battery quicker, and fooling around with
    a lot of starts and stops and stunts will use battery  in a very
    short time.
    
    Most batteries get better after being charged for a while.  Mine
    take a better charge after 12 -15 times.  You can actually see that
    they are getting to a higher reading if you use a voltmeter to charge
    them.

    the machanical speed control reduces motor speed by dissapating
    energy that would normally be consumed by the motor as heat.  the
    special device that magically converts the unwanted battery energy
    to heat is called a *RESISTOR*.  an electronic throttle slows the
    motor by pulsing the electricty, producing a lower average voltage,
    and so a slower motor run.  The electronic speed control is more
    efficient because it is not just "throwing away" the unwanted speed.
    Try an electronic throttle.  Also, you might want to see if you
    could borrow a battery from someone, trickle it over night and see
    if it runs longer than yours.  If so, your battery is munged.

    I'm no expert but something I've found that helps (and by the way
    I also have the kyosho battery) is to fast charge the battery for
    the fifteen minutes and then feel it, if it's warm, it's charged
    if it's cold, give it another five to ten on the quick charge. If
    additional time is needed, keep an eye on it, you want it to get
    warm, not hot. I've heard it said also that after every 5 fast charges
    you should slow charge.
                                                           Jim

    Randy,
    	I beg to differ with you on electronic speed controls extending
    battery run time.  They do dissipate less power themselves, but the
    motor actually dissipates more power to make up for it.  If you
    run at the same average current draw (Motor torque is directly
    proportional to the current), then you will be taking exactly the
    same amount of amp-hours out of the battery!  The difference,
    if there is any can only be from the way you use it.  Smoother
    control action will limit current and give longer run times.
                    
    Charlie Watt (Volt-Ampere)

>< Note 528.4 by LEDS::WATT >
>                       -< Electonic speed control Folly >-
...
>    	I beg to differ with you on electronic speed controls extending
>    battery run time.  They do dissipate less power themselves, but the

I can't disagree with your logic but empirically speaking electronic speed
controls do significantly extend the run time on RC Cars.

Bye          --+--
Kay R. Fisher  |
---------------O---------------
================================================================================

    Kay,
    	I should have only argued with Randy about what the speed control
    does, not that it doesn't extend run time.  I can prove that there
    is no difference between using a resistor to limit current to the
    motor and using a pulsed control to do it in terms of battery drain.
    By the way, the motor gets much hotter if it is running on pulsed
    current than if it is running on DC current.  This is due to the
    fact that the power dissipated n the motor as heat is proportional
    to the RMS (root of the mean squared) current and the average motor
    torque is proportional to the average current.  For DC, these are
    the same, but for pulsed current, the RMS is larger than the AVG.
    Large motors have different power ratings depending on how they
    are controled.  Big motor controllers put a large inductor in the
    armature curcuit to smooth the current to reduce the motor heating
    caused by pulsed current controllers.
    
    Charlie
          

>    	I beg to differ with you on electronic speed controls extending
>    battery run time.  They do dissipate less power themselves, but the
>    motor actually dissipates more power to make up for it.  If you
>    run at the same average current draw (Motor torque is directly
>    proportional to the current), then you will be taking exactly the
>    same amount of amp-hours out of the battery!  The difference,
>    if there is any can only be from the way you use it.  Smoother
>    control action will limit current and give longer run times.

   I also beg to differ. An electronic speed control WILL extend the run
   time. What makes an electronic speed controller different is that it
   acts as a constatnt power DC to DC voltage level converter. The power
   drawn out of the battery will be the same as the power drawn by the
   motor (plus the small dissipation in the controller). Note that I say
   POWER and this is important. If for instance a particular motor were to
   need, say, 4V @ 1A to run at some particular speed then the power that
   it's dissipating will be 4V * 1A = 4W. If your battery voltage is, say,
   7.2V then the current draw from the battery will be 4W / 7.2V = 0.555A
   (neglecting controller dissipation) so you are still drawing only 4W
   from the battery. With a resistive controller the battery must supply
   the full 1A to the motor so the battery will be supplying 7.2v * 1A =
   7.2W. This is similar to the effect that you see in an AC power
   transformer where you have large currents at low voltage on a secondary
   winding being supplied by low currents on a high voltage primary
   winding. To delve any deeper would require going into switching
   regulator circuit theory so I won't go any further. For anybody REALLY
   interested in the whys and wherefores then just about and good text on
   switching power supply design will explain what is happening. 

G. Schrader

RE:.6  -  I saw this one after I entered .7

   OK, for pulsed current and voltage, RMS is the same as average. The
   place where they become different is for SINUSOIDAL and other similar
   waveforms. Any waveforms which have flat levels bounded by sharp
   transitions between levels will have equal RMS and average values. Any
   waveform which has smooth changes over a significant perecentage of the
   waveform's period will have RMS not equal to average. I can derive the
   equations from first principles if you need proof. The reason for the
   motor overheating is probably that the output of the controller is not
   properly filtered (either a failed component or bad design). What is
   happening is that the magnetic material used in many armatures is a low
   frequency material. If you apply a high frequency AC signal (and the 10
   - 30 KHz pulsed output from a switching regulator would do nicely) the
   magnetic flux changes generate heat in the armature core. This of course
   uses battery power just like driving the vehicle does and causes a net
   efficiency loss. The trick is to get a controller which has been well
   designed in the first place.

G. Schrader

    re .7
    
    It is true that you can limit power using a switching controller,
    but you can do this with a resistive controller too.  The difference,
    (Assuming no inductance in the circuit) is that instead of dissipating
    power in a resistor, you dissipate it in the motor instead.  In
    a switching supply, you smooth out the pulsed voltage in a capacitor.
    This allows efficient power control.  If you assume that the motor
    needs some average current, you need to give it some average voltage
    across the armature.  With a fixed voltage supply (battery), you
    either drop the voltage across a resistor, or you pulse the voltage
    to make the average current what you want.  What happens when you
    do this is that the motor dissipates more wasted power as heat.
    
    Charlie
    

    re .8
    Bullshit!!!! The rms and average are only the same for DC.
    
    Are you going to tell me that the RMS of a square wave is the same
    as the average.  Let's take an equal duty cycle square wave that
    is +1volt for half cycle and -1volt for the other.
    
    The average is zero and the RMS is 1.0  In my math, zero is not
    equal to 1
    
    Charlie
    

    I found some interesting things in experimenting with various pinion
    gear sizes on my Optima Mid...
    
    I had bought a Twister 702 motor that I ran for a while with the
    stock 20 tooth pinion. The car had a very good top speed, but the
    running time was only three minutes or so.  After a run, the motor
    was red hot, as was the battery.
    
    I installed an 18 tooth pinion, and the difference was astounding.
    The run time went up to 5 or 6 minutes, with little or no loss in
    top speed.
    
    The point here is that you have to be carefull to gear the car in
    such a way that the motor doesn't act like a short circuit or a
    locked rotor.  If motor manufacturers would provide torque vs RPM
    curves, and torque constants, you could estimate how many amps your motor
    would draw.
                                                              
    Ed.

    
    The equation for the RMS value of a wave form is...
     
    
    	 __                __  ^.5
      	|          /         |
     	|	   |         |
        |  (1/tau)*|f(t)^2 dt|
      	|          |         |
      	|         /          |
      	|_                  _|
                               
    where tau is the period and f(t) is the waveform to be analyzed.
    
    As Charlie says, this evaluates to .707 for a 1 unit peak to peak
    sinusoid.
    
    Ed.

    ref .-1
    Ed,
    	You mechanical guys are not supposed to be spouting equations.
    (Just kidding).  You are correct.  .707 * peak is the RMS of a 
    sinewave.  The average is considerably less than this. (.636 I think)
    This of course assumes you full wave rectify it. Otherwise, the
    avg is zero.  If anyone is interested, I can put in some theory
    on how dc motors work and how to control them most efficiently.
    
    Charlie
    

OK, my initial statement was a little too broad but only a little. 

What we're talking about is unipolar pulsed waveforms, not bipolar. For unipolar
waveforms my statement is perfectly true. For instance, a 50% duty cycle
waveform with levels of 1v and 0v has both a average and an RMS of 0.5v. 

G. Schrader

RE: .12,.13

Your equation is correct but we are not dealing with sinusoids, we are dealing
with square pulses, which is a different thing entirely. If you evaluate the
equation using square pulses what you get is average = RMS (for unipolar
pulses of course).

G. Schrader

OK, I concede on the RMS issue. Sorry about that. That's what you get
when you work from memory... sigh. I hereby retract (at least some of)
my previous statements.

I kind of figured that i'd start a good shouting match when I
replied to this note 8^).

G. Schrader

    So it sounds as though it's safe to say that a pulsed speed controller
    (note that not all electronic speed controllers are pulsed!  Some
    have a pass transistor and merely give you better control over your
    speed but no better power conservation.  However, that's just another
    rathole) will give you better efficiency than a resistive speed
    controller, though it may be a little less efficient that ideal,
    it's sure not wasting (3.2 volts at one amp gives 3.2 watts for
    the previous example) as much power as a resistive one.
    
    Willie

    re .16
    Thanks for admitting your error.  Just to finally clarify things,
    The formula for RMS that ED A. presented is correct for any waveform.
    The formula for avg is different.
    
    
    AVG = 1/T { integral of(f(t) dt)} evaluated from 0 to T
    
    ONLY FOR DC is the RMS equal to the AVG.  This is a common
    misconception!  The narrower the pulses, the bigger the difference
    between RMS and AVG.  
    
    In an electric car application (assuming negligable armature
    inductance) the only reason to pulse the speed control is to reduce
    the need for a heatsink on the power devices in the speed control.
    It does put more heating load on the motor than a linear regulating
    speed control.  All of this means nothing if you spend most of the
    run time either full throttle or zero throttle.  If you never use
    full throttle, then the efficiency could be improved by reducing
    the available voltage or rewinding the motor so that you do use
    full throttle.  Also, Ed's matching of gearing to the job makes
    much sense, because it reduces the torque load on the motor.
    
    Here is the thing to remember:
    
    Heating losses in a motor go with I squared where as the Torque
    goes with I.  Work done by the motor goes with I * RPM (Horsepower)
    
    So much for speed control theory!  I'm not sure whether many of
    them are pulsed anyway.
    
    Charlie
    

    The Futaba models MC(***) are advertised as being of the voltage
    pulsing design.  You can actually feel the motor cogging (not torque
    ripple) at very low RPMs.
    
    The pulse rate is set by a small trim pot, and is a snap to adjust
    to set the range and peak values.  All adjustments are made with
    no load on the motor.
    
    Novak also makes pulse type controllers.  All the winning racers
    use them.  I use a Futaba, which I guess makes me a loser. :^)
    
    Ed.

    I have the Futaba MC108, which I believe has a series pass transistor,
    as it gets quite a bit warmer and the MC112 I used to have (the
    112 had FETs...).
    
    Willie

    
    re: .0
    
    J�rg,
    
    Don't go out and buy an expensive speed control (at least not for
    this problem). You should be looking for binding in your drivetrain.
    If your battery and charger check out or you find that after four
    minutes your motor is hot, pull your motor out and make sure everything
    turns very easily in the transmission. And make sure the pinion
    gear (on the motor) mates properly with the transmission.
    
    Running time will get better as things break in, but at four minutes
    something is wrong.
    
    Chris
    
    
    
    
    

    I was begining to think we'd have to change the title of 
    the base note ;^)...
     
    J�rg,
    
    Just a thought, but have you checked to see that the charger
    is really fully charging the battery pack...... It might be
    helpful to charge your packs on another charger and see it
    that makes any difference. In my experience, the aristo-craft
    chargers have a high failure rate as compared to some of the
    others available, though the usual symptom is they are dead
    as a door nail. Like the one sitting on my kitchen counter.
    
    :^(
    
    DougT

    When I charge my batteries I use a DVM (Digital Volt Meter) to obtain
    a peak charge. What you do is hook up the meter and watch the voltage
    rise. When it starts to drop the battery is at peak charge.
    
    I sometimes can get 10 minutes on one battery. 
    
    It usually takes me 22 minutes to get a peak charge.
    
    Coooop
    
    

Thanks for all the hints and the truckload of theory about speed
controllers. I might have no been clear, but I have two identical
battery packs and the car has an el-cheapo electronic speed control
(huge heat sink on it!). I took the motor out and made sure that
everything turns easy, I also made sure that the wheel bushings and
dog-bones have plenty of graphite in and on them.  I furthermore
changed the gearing from a 16 to a 20 teeth, but it does not prolong
the life of the battery-charge and has only minimum impact on
speed and acceleration. I will check the trick with the DVM tonight. 
Can I top the batteries of at a later point or should I do it
rightaway, when I do the initial charge? 

    Thanks again 
	J�rg 
   

    a charged nicad can be "topped" at any time before use.  Most
    heavy-duty racing types top their cells off with the pack in an
    ice chest just minutes before the race.  These types of extra effort
    charging techniques are good only for an additional 10 or 15 seconds
    of high current at the start of the race, they do not really prolong
    the run of a peak charged battery significantly.  It works because
    NiCads can be "overcharged" to a small extent without damage (the
    cell actuall puts out slightly more than its rated voltage) but
    they can only hold this charge for a short time (it dissapates fairly
    quickly) and so the top-off technique mustbe used just before a
    race.  Definately try and peak charge your batteries, this will
    certainly tell you if your charger is screwy.  It is the only way
    I fly my electric planes (since the few times I have not bothered
    with a peak charge I had to panic land the plane, or just suffer
    with a very unsatisfactory flight).  Dont bother topping them off
    just before you run the car, the extra effort won't make an appreciable
    difference.

    A pulsed throttle controller WILL increase your run time if you're
    not running at full speed. The theory about RMS is nice, BUT REMEMBER,
    the RMS delivered to the motor is the same value RMS being supplied
    by the battery. Heck, you don't even have to go to RMS. RMS is used
    to figure the effective voltage for resistive loads. It makes a
    fake calculation of the power (hence squared) and converts it back
    to voltage. You can figure the current directly. When switched on,
    it supplies battery voltage at torque current. When switched off,
    it supplies no power.
    
    The primary reason you don't want to use RMS anyways is because
    the motor is not a pure resistive load. It changes with torque.
    Add to this the fact that you are breaking the circuit, and you
    have anything but a resistive load.
    
    In short, a pulsed throttle control will improve your runtime.
    
    							John.

    re -1:
    Believe me, you don't know what you're talking about.  THe main
    source of power loss in the motor is heating.  With pulsed current,
    the heating loss is related to the RMS power where as the Torque
    generated is related to the AVG power.  RMS is a true, not a FAKE?
    power calculation when measuring power in a resistive load based
    on load current.  THe facts as I presented them previously are
    accurate.
    
    Charlie
    

>    A pulsed throttle controller WILL increase your run time if you're
>    not running at full speed.
...
>    In short, a pulsed throttle control will improve your runtime.

That part is true.  I assume you disagree with the technical explanation
 - right?

Bye          --+--
Kay R. Fisher  |
---------------O---------------
================================================================================

    RMS measurements were originally intended as a relative reading
    of effective power expressed in volts. The reason why it is squared
    is because in a truly resistive load, current is proportional to
    voltage. In the speed control application, you don't need to use
    RMS, in fact, you shouldn't use RMS because the motor isn't a true
    resistive load. You can calculate power directly.
    
    Switch on - full voltage and current proportional to torque.
    Switch off - full voltage and no current ( the motor will have some
    back EMF )
    
    Now, there are two types of loss that makes the controller less
    than 100% efficient. The first is the voltage drop across the
    transistor, which is .7 for bipolar silicon, much less for VMOS
    FETs. This is the DC load. The AC load is much more for VMOS, as
    it is recommended that you have 4.7 uF on the brush terminals to
    prevent temporary power outage ( you should also have some .1 ceramic
    or equivelent caps going from each brush to chassis to reduce EMI)
    So that when initially switched on, the VMOS FETs are going to drop
    some voltage. This is AC switching load.
    
    The reason why you have extra ceramic caps on the brushes is because
    the non polarized elecrolytics in the 4.7 uF range do not have a
    very good frequency response, and tend to roll off before offering
    any reduction in EMI.
    
    In short, RMS is a force fit to this application, and doesn't work
    because the controlled motor isn't a resistor. Lamp dimmers work
    the same way, and even though the lamp itself is a resistor, when
    you put in the switch, it ceases to be a resistive load. This is
    why the dimmer control on your wall doesn't heat up. A motor is
    really more of an inductive load. Instead of generating magnetic
    field, however, it generates angular momentum.
    
    							John.
    						( Senior Engineer )

	RMS is only applicable to sine waves, not squarewave,
	swatooth waves or pulses.....

	bob

RE: < Note 528.30 by KERNEL::DAY >

> RMS is only applicable to sine waves, not squarewave,
> swatooth waves or pulses.

    As I remember it from college, RMS may be applied to ANY waveform. 
    If it is a sine wave, the formula may be simplified from the general
    case which involves integration to get the "Mean" portion of "Root
    Mean Square" (RMS).

    Charlie, although I can't come up with any formal proof, I tend to
    think that what was written in .29 is true.  A motor is not simply a
    resistive load but is mostly inductive.  (I'm not trying to throw
    fuel on the fire - I'm still not certain what the REAL answer is and
    would like to figure it out.)

                                - Dan Miner

    Oh, as an aside, I have a speed controller between a 1700 mAH battery
    and a 240 motor, which I rarely run at full speed. I get run times
    of a little less then 15 minutes. That turbo ultima has got a fast
    motor, I have hit 30+ with it, but I spend most of my time practicing
    corners.
    
    Speaking of corners, the best technique is to do nice, smooth, wide
    corners, coming near the curb only at the apex of the turn. The
    act of turning the car will slow it down, so that for the most part,
    you only want to apply medium throttle, and only enough to get it
    at speed. Speed controller is kind of a misnomer, it's really a
    throttle, and adjusts the acceleration. You want to apply enough
    throttle to make up for the loss in the turn. It's really important
    to keep traction throughout the whole turn. Not only does this
    technique give you better control, but it translates some of the
    forward momentum into the turn. The only exception would be on hairpin
    turns, where you would want to apply full brake, then full throttle.
    A controlled spin would give you faster braking. ( It'll also wear
    out your tires pretty quick )
    
    Just for comparison, watch some of the full scale racers. You'll
    see nice smooth lines traced on the track ( as a matter of fact,
    the lines are usually etched in rubber ). The only real difference
    is that the 1/10th scale racers have a better weight/traction ratio.
    ( a 1/10th racer going 30 mph is going at a relative 300 mph ).
    Choice of tires is absolutely critical. The suspension should be
    adjusted so that the suspension is as soft as it can be without
    bottoming out, and height determined mostly be sway ( which is why
    antisway bars are desirable ). You want to have some castor so that
    the tires are flat in the turns.
    
    Back to the original topic:
    
    Try measuring the charge in the battery. For a 1200 mAH battery
    on a fifteen minute charge, you should see an average of 4-5 amps.
    A speed controller is definitely a good thing to have, from all
    respects, I put mine in right from the beginning, but if you're
    getting short runtime, chances are it's the battery.
    
    							John.

    I should have said regarding the type of load the motor is, that
    it acts electrically very much like a capacitor. It opposes a change
    in voltage. This, too, is inaccurate.
    
    The way motors are generally rated is no load RPM and stopped torque.
    If you have a cartesian coordinate system with torque on one axis
    and RPM on the other, the graph traces a nearly straight line through
    both points. Given a certain voltage, the motor will generate enought
    torque to produce an equal back EMF voltage. During acceleration,
    the motor will draw current that will trace an negative exponential
    curve, approaching, but not quite completing, final RPM.
    
    If you were to model the car as a velocity capacitor, you could
    figure out the winding resistance for the RC network. The value
    of C would vary with the weight of the car. A load resistor across
    the capacitor would vary depending on whether you were going uphill,
    downhill, or turning. Load resistance also varies with velocity due
    to friction and drag.
    
    So, electrically, an RC car has a constant capacitance, series
    resistance, and a variable load resistance.
    
    The series resistance is dependent on the motor, if you have a throttle
    control, I recommend a reasonably fast motor. The 240 I have has
    significantly more than the stated 4 minutes when used with a light
    throttle. The capacitance is dependent on the weight of the car,
    so make it light wherever practical. The load resistance is dependent
    on a lot of factors; bearings, traction, drag, and course. A good
    suspension will increase your runtime, as it gives you better traction.
    Ball bearings increase your runtime, for obvious reasons. A good
    aerodynamic body will increase your runtime ( a wing helps for twisty
    courses, but increases drag on straights ).
    
    The suspension and the body are tuned for the course you intend
    to run. One trick I haven't seen is a wing that is attached to the
    steering to adjust the angle of attack with the turn.
    
    As a general rule - anything that makes your car go faster also
    makes it more efficient.
    
    							John.

    This is the last comment I am going to make on this subject.  What
    we are talking about when comparing pulsed control vs resistive
    control is efficiency.  This is the ratio of power dissipated as
    heat in the controller or motor vs work done by the motor to drive
    the car.  The losses in the motor (neglecting friction) are due
    to the resistive nature of the armature ONLY.  The inductance only
    limits rate of change of the current.  (There aren't many turns
    on these motors.)  If you go to a good textbook on DC motor control,
    you will find that the pulsed controller will cause the heating
    losses in the motor to increase and the heating losses in the
    controller to decrease.  Guess what!  You heat up the motor more
    instead of heating up the resistor in the controller.  This can
    all be easily explained using true RMS of the current waveform as
    was explained earlier.  Any waveform has a valid RMS value - even
    DC.  The RMS of a sinewave is .707 of the peak, and the RMS of DC
    is equal to the peak.  
    
    Amen,
    
    CHarlie
    

    
    I don't know why people keep saying that RMS only applies to sine
    waves.  Some meters can only give RMS of a sine wave because they
    approximate it instead of integrating, but that doesn't mean you
    can't calculate the RMS voltage, current or power for any waveform!
    Just perform the integration.
    
    Now back to the other question, because I got lost somewhere.  The
    claim is that a pulsed motor controller gives longer battery life,
    but I didn't see any proof of why it would do any better than a resistor.
    Is it in these arguments somewhere?
    
    Bill
    
    P.S. I'm a principal engineer and wouldn't think about arguing with
    	 Charlie about motors, considering the fact that he's been
         designing servo systems and motor controllers for more years
         than he'd probably want to admit.

    Thanks, Bill - I wouldn't want to admitt to how many years I've
    been fooling with motor controls.  I designed my first one in 1974
    for a full scale electric car that ran on 120 volts worth of lead-
    acid batteries.  Runtime and performance were just as critical as
    they are with our toys.  Unfortunately, lead-acid batteries are
    not as good as nicads but we couldn't afford a car full of nicads.
    The controller (an SCR chopper) was rated over 500 amps and had
    current feedback control.  The wire that connected the batteries
    together was the size of a garden hose.  I like the little bitty
    motors we get to play with now.  You're right about meters.  Most
    meters measure average voltage or current and some are calibrated
    to read in RMS for a sinewave input only when on the AC scale.
    Special RMS power meters are available for special applications.
    They measure true RMS by integration or measuring IR heating in
    a resistor.

    Well, I thought long and carefully about the problem, and I have
    to say it's not intuitively obvious. You're right, the motor will
    dissipate the power that would otherwise be dissipated by the resistor.
    I set the seeds of my own destruction when I constructed the resistor
    capacitor model. Taking the average current ( which is allowable
    as long as the applied voltage is steady state ), the resistor has
    a lower voltage drop across the motor.
    
    From the standpoint of pure energy ( and I like to construct two
    models whenever I'm in doubt ), the battery supplies its voltage
    times current power, and the motor applies the back EMF (speed)
    times current (torque) power. This means that the voltage difference
    between the battery supply and the motor back EMF must be dropped
    somewhere.
    
    Sorry for the confusion, maybe my coming to understand this problem
    and subsequent description will help others understand it as well.
    The resistor capacitor model works pretty well, in fact, here's
    a picture:
    
             R winding       C back emf
    *---------/\/\/\/-----+------][------+---------*
                          |              |
                          +----/\/\/\/---+
                               R load
    
    The only power that's being applied to the car is the voltage on
    the capacitor ( back emf ) times the current. The power in the winding
    resistor is dissipated.
    
    Since the power loss is in the winding resistor, you can use RMS
    current as an indication of power loss, although I can at this
    conclusion after the fact, and used power calculations.
    
    There is a solution to the general problem of how to avoid power
    loss, the one I'm thinking of uses a flying core transformer. The
    schematic would look something like this:
    
    -----0 0--------)||(---->|-----+-----------+
    Vb   -+-        )||(          ___          |
    Vb   FET        )||(          ___        Motor
    Vb              )||(           |           |
    ----------------)||(-----------+-----------+
    
    The idea here is similar to what they do in power efficient switching
    supplies. Power the transformer by turning on the FET. Because the
    diode is reverse biased, no current is induced in the secondary
    winding. However, leave the transformer turned on so that it induces
    current, so that the FET thinks it's driving an inductor. Next,
    turn off the FET switch, so that the magnetic flux in the transformer
    induces a positive voltage at the secondary winding, supplying current
    when it forward biases the rectifier.
    
    The transformer itself would be rather special. It would have a
    small number of windings and a closed core made probably of ferrite
    to reduce flux loss due to hysteresis.
    
    Performance wise, it would remove the speed limit on the motor,
    allowing a fairly large voltage to accumulate, limited only by motor
    resistance. You could have an extremely large torque at low speed,
    and an extremely high speed at low torque. There would be some loss
    in the transformer and rectifier, so these would have to be on the
    larger side to handle the power.
    
    Well, there's a project for ya.
    
    							John.

    
    .37 has me confused now... 
    
    No matter how much voltage I apply to any DC motor, the limiting
    speed will be voltage/torque constant.  If I apply a step voltage,
    the velocity will look like:
    
    	RPM = C*Volts/Kt*(1-(e^-t/T)) 
    
    C is a conversion constant to get all the units right;
    Kt is torque constant (or back emf constant in MKS usnits)
    t is instantaneous time
    T is mechanical time constant, resistance*inertia/(Kt^2)
    
    At t = forever, the limiting speed is still V/Kt now matter what
    fancy scheme I use.
    
    If I could apply a constant current to a DC motor, it would accelerate
    until the parasitic losses equalled Kt*current.
                                                    
    Ed.
    

    re .38:
    
    The equation you use assumes voltage drop from only the losses.
    A motor is a transducer, it operates in both directions, both as
    a motor and as a generator. There is an EMF produced by the motor
    while spinning, proportional with the RPM's. The standard graph
    for depicting a motor has a straight line that crosses two points:
    torque at 0 RPM, and RPM at 0 torque.
    
    In the capacitor resistor network I drew in .37, the voltage on
    the virtual capacitor is proportional to the RPM's on the motor.
    Instead of the energy being stored as a charge on the plates of
    a real capacitor, the energy is stored as kinetic energy in the
    system that is being driven by the motor.
    
    Study the resistor capacitor model for a while, keeping in mind
    the torque-RPM graph, and see of it makes sense.
    
    							John.

    
    My only contention with your theory is that a voltage driven motor
    will have no RPM limits.  I'm afraid that I'm not intelligent enough
    to comment on your driver scheme.  :^) I'm just a dumb mechanical
    engineer, but I have designed a few motors.
    
    The equation that I posted in fact does include the back EMF effect.
    The model used to derive the velocity equation is a classical voltage
    driven motor with inductance neglected.  The block diagram looks
    like...
    
               |---------|  I  |-------| Torque |------| Accel |----| Veloc   
    V-->(X)--->|  1/R    |---->|  Kt   |------->| 1/J  |------>| 1/s|-------->
         ^     |_________|     |_______|        |______|       |____|   |
         |     (resistance)                    (inertia)    (integrator)|
         |                     |--------|                               |
         |---------------------| K emf  |<------------------------------|
                               |________|
    
    Most MEs, myself included, are uncomfortable with LaPlace transforms
    and block diagram algebra, but you can get the exact same results
    by solving the differential equation...
    
    	Theta'' + (Kt*Ke/J*R)*Theta' = 0
    
    and using some common sense boundary conditions, such as no velocity
    at t=0.
    
    Anyway, this is pretty standard stuff.  I hope I'm missing something
    basic here.  If a new speed controller could be invented that would
    extend run time, I'd be delighted.  My personally experience, having
    used both mechanical and pulsed ESCs is that the pulsed variety
    does seem to give a bit longer run time.  However, I could easily
    be convinced that the extra run time is due to the infinite throttle
    resolution, which makes the car less likely to be wide open at any
    given time.
    
    Ed.
                                                            
    

Well, I got lost in some of the theory about 15 notes back (after all,
thats why I got into this digital stuff, everything is 1 or 0, no 
integrals :-) but I would like to understand this as best I can.  I 
am thinking about getting a pulsed power controller for my almost
complete Goldberg Electra (electric powered sailplane, .05 motor,
1700 mah 7.2 v battery) on the assumption it would lengthen run time
at partial power.  In fact, I have seen that stated several places.

I used to build stereo amplifiers as a hobby and I remember a bit of what
I read about various power supplies.  I accept (I think) the statement
that a pulsed waveform will waste as much extra power in the motor windings
as you save in the controller.  But, can't you effectively turn the pulse
train in to (almost) straight DC with an inductor/capacitor filter?
Borrowing from an earlier example, don't you get something like this:

                                      .-- (Say, how DO you draw an inductor
                                     |       on an ASCII terminal ?? )
                                     V
           
          .---------.     .------~~~~~~~--.
        + |         |     | +             |
   .--------.     .---------.             +------------. 
   |        |     |         |             |            | +
   | battery|     | pulsed  |           -----      .-------.
   |        |     | power   |           .---.     (  motor  )
   |        |     | contrlr |             |        `-------'
   |        |     |         |             |            | -
   `--------'     `---------'             +------------'
        - |         |     | -             |
          `---------'     `---------------'

       |
 7.2v -|-----------        _   _   _   _ 
       |                  | | | | | | | |
       |                  | | | | | | | |
       |                  | | | | | | | |        _   _   _   _ 
   4v -|                  | | | | | | | |       ' `-' `-' `-' `-'
       |                  | | | | | | | |
       |                  | | | | | | | |
       |                  | | | | | | | |
       |                  | | | | | | | |
   0v -+------------      ----------------      ----------------- 

Doesn't this make it so the motor sees almost pure DC at a reduced voltage?
Assuming the switcher is efficient, doesn't this save most of the 3.2 volts
times X amps that would be wasted by a resistive controller?  If the
switching rate is 10s of Khz neither the inductor nor the cap would have
to be very big.  If this isn't right, what am I missing?

Thanks,

  Kevin C.
          

    re .40:
    
    Yes, that block diagram looks about right. Now, study it and see how
    velocity reaches a limit as t goes to infinity. We have another
    convert.
    
    re .41:
    
    That circuit would work for about one cycle, where the high impedance
    on the switch side of the inductor would cause a voltage spike
    guaranteed to blow your switching. In order to dampen the spike,
    you would need a capacitor, low and behold, you have a voltage drop
    across your switches. I honestly don't believe there is a DC coupled
    solution. As long as the current going to the motor is equal to
    the current being supplied by the battery, you will always have
    a power loss at lower speeds. The flying core transformer works
    because you effectively change the load impedance by changing the
    operating frequency. Another way of doing it might be:
    
    *------------0 0----+----()()()()()----+----0 0----------*
                 S1     |        L1        |     S2
                        0                  0
                          S2                 S1
                        0                  0
                        |                  |
    *-------------------+------------------+-----------------*
    
    When S1 is open, S2 is closed, and vice versa. We therefore have
    on DPDT switch with an inductor between the commons. With a duty
    cycle of 50%, the lower the frequency, the more power that gets
    transfered. You would need some caps to compensate for the break
    before make on the switches, or make the switches make before break
    and lose some power. Let's call this the flying inductor scheme.
    
    						John.

re: 42

Alright, put a diode in front of the inductor.  Doesn't this force the 
collapsing magnetic field of the inductor to deliver power to the
capacitor/motor setup rather than blasting the VFETs (or whatever)?


          .---------.     .-->|--~~~~~~~--.
        + |         |     | +             |
   .--------.     .---------.             +------------. 
   |        |     |         |             |            | +
   | battery|     | pulsed  |           -----      .-------.
   |        |     | power   |           .---.     (  motor  )
   |        |     | contrlr |             |        `-------'
   |        |     |         |             |            | -
   `--------'     `---------'             +------------'
        - |         |     | -             |
          `---------'     `---------------'

       |
 7.2v -|-----------        _   _   _   _ 
       |                  | | | | | | | |
       |                  | | | | | | | |
       |                  | | | | | | | |        _   _   _   _ 
   4v -|                  | | | | | | | |       ' `-' `-' `-' `-'
       |                  | | | | | | | |
       |                  | | | | | | | |
       |                  | | | | | | | |
       |                  | | | | | | | |
   0v -+------------      ----------------      ----------------- 

    This one hit me like a bolt out of the blue. Take a look at it and
    see what you think. 
    
    from switch *------+------()()()--------* to motor
                       _        L1
                       ^ D1
                       |
                *------+--------------------*

    This circuit was inspired by .41. It doesn't have the nifty features
    of the flying core or flying inductor scheme, but it will do the
    trick. This should work with a standard controller, and consists
    of only two parts ( the capacitor already on the motor should be
    sufficient ), a switching diode, and a low winding ferrite torroid.
    Both components can probably be salvaged off a broken power supply.
    
    Maybe I'll stick my neck out further and actually try it.
    
    						John.

    Well, what's the lesson here? Basically that electronic speed controls
    do not automatically give you longer run times. The circuit in .44
    is but one example of a circuit that is more power efficient, however,
    what the manufacturers impliment is anyone's guess. My guess is
    that any manufacturer that says its electronic throttle is power
    efficient is likely using a special scheme to do so, and it will
    be a bit more expensive. Any old switcher won't necessarily be
    efficient, but it is possible, and I'm sure someone will do it,
    if it hasn't been done already. All throttles are not alike. Keep
    a watchful eye before you make that purchase.
    
    							John.

    re .37   If you want to solve the problem, you need an energy storage
    device in series with the motor in the form on an inductor.  You
    need a big one, so it is not practical for cars or planes, but this
    is how it's done in electric cars (Full Size) and in big machine
    drives.  The inductor smooths the current pulses just like the
    big capacitor in a switching power supply.
    
    CHarlie
    

>    Well, what's the lesson here? Basically that electronic speed controls
>    do not automatically give you longer run times. The circuit in .44

I won't pretend to understand the theories that you guys are talking
about BUT...

The things DO increase run time.  I challenge you to ask anyone who has
switched form a resistor speed controller to an electronic speed controller.
I'll bet you can't find anyone who did not also see an increase in run time
when not at full throttle.

I'm still confused when I read your explanations as to weather or not
you believe that they work.  Could we avoid the theories for a couple of
notes just to hear empirical evidence and opinions?

Bye          --+--
Kay R. Fisher  |
---------------O---------------
================================================================================

    re .46:
    
    The size of the inductor required depends on the switching speed.
    
    Yes, I HAVE seen advertised controllers that feature:
    
    "Filtered output"
    
    The circuit I drew is so trivial that it is more than likely most
    controllers have this feature. The diode can be replace by a
    syncronized switch, which probably exists already if it has braking.
    
    							John.

    I am about to become a electric R/C car owner. I have ordered a
    Turbo Optima Mid SE which is currently on backorder. I am now
    running a Vanning (gas), which is a "gas". However, not to take
    anything away from the Vanning, I also recognize the merits of
    the electric cars and am looking forward to the Optima. 
    
    I read nearly all the discussions pertaining to electrics but
    never came across any information on the following:
    
    * What are the practical differences of 7.2v 6 cell vs. 8.4v
      7 cell packs? Both are available in 1200, 1400, 1800, and
      the new 2000 mah current ratings. I'm not sure if a 8.4v
      pack will fit my car (all advertised ref to the car talk
      only about 6 cell) but if it does, what does the additional
      1.2v buy me? Besides the additional cost, and added weight,
      wht are the other tradoffs, if any?
    
    Al

    Well, the higher voltage (at a given capacity) gives you a higher
    top speed, more wear and tear on your speed controller (depending
    on what type it is this might not be significant), a warmer engine,
    and (at top speed) a shorter running time.  It'll probably go further,
    but run down faster (note, only at full speed).
    
    I have no direct experience, as I use 2500 mAH, 12 volt Gates Cyclon
    Monobloc batteries, but then Tycho doesn't have to by ROAR (or
    whatever) legal, and he has yet to win a race....  :+{
    
    Willie

    
    I have an Optima Mid (by brute $$$, it's the equivalent of the new
    SE), and if you plan to race, you'll need the extra volts to be
    competitive.
    
    I recently bought an 8.2v 1900mAh pack from Ande's Hobbies in
    California for $44.95.  The pack is made by Shinwa using matched
    Panasonic cells, not Sanyos.
    
    I have several motors, but my hottest is a Twister 702.  Using a
    1200 mAh pack, I was limited to 4-5 minute runs.  I can now go full
    tilt for a good 6-7 minutes with the new pack.
         
    If your new to electrics, and the Optima Mid in particular, I can
    offer a few tips based on first hand experience...
    
    1.  When you assemble the car, don't use any grease in the gearbox
    or driveshaft joints.  The only place grease should be used is in
    the ball differentials.  Every other location attracts dirt and
    debris.  After a few weeks, the gearbox will become so gummed up
    that the frictional loss will start slowing you down.  After several
    cycles of tearing my car down to clean the gearbox out, I started
    using a dry teflon lubricant made by Paragon Racing Products.  A
    2 oz bottle costs $3.95 and lasts forever.  Since using this stuff,
    my car has remained nearly frictionless for several months now.
                                                                  
    2.  DO NOT use the solid rubber motor boot that Kyosho supplies
    with the kit, especially if you will be installing a hot motor.
    The stock boot traps heat in the motor, causing premature rotor
    and brush wear, and probably demagnetiziation also.  I found this
    out after killing a $65 motor.  If you stay on asphalt surfaces,
    you don't need any boot.  Every few runs, just hit the rotor with
    a little freon spray and compressed air.  If you must run on sandy
    surfaces, use a foam motor boot found at any hobby shop.  They keep
    out most of the dirt and allow the motor to breathe.
                                                                  
    3.  When mounting a 7-cell pack, use the forward set of battery
    mounting holes, which will keep the weight distribution nearly 50-50.
    
    Good luck with your car.  The Optima Mid is a real ball, especially
    with foam slicks and a hot motor.
    
    Ed.
    

    Thanks for all the info, Ed. This Paragon dry Teflon lubricant,
    is it readily available? Would Tower carry it? it sounds like a
    good bet for lubricating my Vanning chain in particular. What a
    magnetic for dirt that is.
    
    The Optima is still on backorder. Since the seven cell will fit
    it, I think I'll opt for the 2000 MAH batteries by Universal that
    Tower carries. If I remember right, they were also in the $45  
    range.
    
    Al

        I can't speak for the particular brand, but I've found Teflon
        powder in hardware stores. It has some common applications with
        graphite powder, but is non-conductive. It doesn't burn the way
        graphite will, but at very high temeratures it will turn into a
        gummy mess. 
        
        I've machined some Teflon parts, the stuff is easy to work with,
        but VERY expensive. Given the price of Teflon, I wish there was a
        way to powder the scraps, instead of throwing them away. Anyone
        have a use for "Teflon wool"? 
        
        There is also a new oil on the market, that has Teflon suspended
        in it. Tower has it under the name "Slip-it". I think it is
        similar to automotive products like Slick-50 and Tufoil. It is
        supposed to leave a Teflon coating on metal after treatment. 

    
    I had originally tried powdered graphite, which worked fine, but
    was a bitch to apply.  I tried brushing it onto varous plastic parts,
    but it tended to fall off.
    
    I finally got tipped to this Paragon stuff, which is actually volatile
    liquid with microscopic particles suspended in it.  You dribble a
    few drops of the liquid onto the parts, and the volatile component
    evaporates, leaving a fine white powder.
    
    I got my bottle at Bill's Hobby Barn in Sudbury, MA.
                                     
    Ed.

    I checked the Tower Catalog. They list a few Paragon products of
    which, something called "Liquid Bearings" sounds like what you
    are talking about. Quantity is not mentioned but the price is
    $3.65. I'll swing by Tom's Hobby Korner to see if he stocks it.
    
    Al

>    Does anyone have a car with a 6-cell pack that would be willing
>    to try a Gates Cyclon Monobloc battery (6 volts, 2500 mAh) for a
>    comparison?  I've got a spare from Tycho's battery change....
>    
>    Willie

    I used to run a cox BMW 3.0CSI electric racer with 4 2.5AH 2 Volt
    Gates lead-acid cells.  The car was originally run with a 7.2 Volt
    1.2AH NiCad pack.  The increase in weight from the cells was
    slightly offset by the extra voltage, but the car still suffered
    some from the weight.  On the whole, however, I liked it a lot
    more than the NiCad incarnation of the car.  Run times were, of
    course, MUCH longer, and the added weight really helped the car
    corner a lot faster.

    Randy (8^D)

    Randy,
    
    	I was wondering if anyone had tried it!  I used 3 of the 5 AH
    individual Gates cells on Tycho in his original incarnation, and
    was thinking about going to 6 of them for 12 volts, but I got
    scared... :+)  Glad to hear it works!  The monobloc batteries at
    2.5AH and 6 volts are available thru Radio Shack (of all places)
    for $10 each, and are somewhat lighter than the originals.
    
    Willie
    
    PS.  I was wondering what you were talking about for a moment there,
    	but I'm willing to drop the voltage_on_cell_cases subject. 
    :+)
    

    Better late than never.  Having run an RC10 for about 2 years with
    a resistor speed controller, not the 3 speed type but a wirewound
    with a slider (similar to Parma kit) I switched to a Futaba MC108B
    (thyristor (I think) with full power relay) then to a Futaba MC112B
    (FET type).  This is what I found!
    
    The main reason for the switch was to reduce weight and lower the
    centre of mass.  I also wanted to mount the cells longtitudinally
    to reduce roll moment.  I had suspected my car was loosing out by
    being some 6-8 ounces over weight, compared to a rivals Ultima.
     BTW I invariably beat the Ultima 'cos the driver tries to beat
    me by going too fast and usually dumps before the race is over and
    also crashes.
    
    Right away I can confirm that the resistor gave far superior slow
    speed handling than the 108, about equal to the 112.  I think the
    108 had a current limiter and tended to snatch on startup.  As I
    drive off-road speed control can be very critical and a race is
    not run at full throttle all the time.  If it was the resistor and
    108 would win because full speed is straight through.  The current
    consumed by the servo handling the resistor makes some difference,
    but in practise nothing much.  After a race cells are hotter using
    the 112 than the 108 or resitor (suggesting to the ignorant that
    the 112 must be sucking more power out of the cells and therefore
    more 'go' into the motor).  The 112 itself is hotter than the 108
    or resitor.
    
    But what about car performance, I hear you ask?  I suspect that
    acceleration is better with the 112, this may be due to the difference
    in weight.  The 1 second delay to get into reverse with the Futabas
    is a pain when you crash, but then so is crashing.  If you can't
    get thru the race without crashing into stationary objects, there
    is no point in going faster!  The bottom line is that under race
    conditions I can't detect any difference, other than the weight
    distribution benefits in handling, which is very beneficial.  As
    far as motor heat is concerned I havn't noticed any difference although
    I tend not to let my motors get too hot so as not to destroy the
    magnets.  For the RC10 owners amoung you here is a typical setup:
    
    Technigold motor 0-4 degree advance, gearing 14:52, 6, 1.2mAh cells
    charged to 35-40 degree C, wet grass, flat track, 6 minutes duration.
    Ran 18/19/18 laps within 5 minutes (qualifiers), getting pole in
    the A Final and won that by 1 lap (19).  I personally avoid very
    'hot' motors in the RC10 as I don't enjoy replacing intermediate
    gears.  The RC10 is now 2� years old, is run EVERY week at race
    meetings and apart from *andys* front arms is out of the box, including
    servo saver (tie-wrapped together).  Fully ballraced, of course.
     Gearbox took many hours to build, has NEVER been opened up (if
    it aint broke don't fix it) and still on the orriginal diff nut.
    
    I like my RC10, must tell you about the CAT sometime.
    
    Rob,  bbk Racing                                                          
                                                           

    re .49 BTW the cells are SCRs and the guy next to me on the grid
    (also RC10), an old sparing partner often qualifying within a second
    or so my FTD, decided his strategy for the final was to beat me
    to the first corner, then drive WIDE to keep me behind.  He put
    a 17 double (Losi can) in and as the lights changed to green there
    was a very nasty grinding noise and at the first corner my only
    company was an Ultima (3rd on the grid).  After a lap or so I shouted
    'Where's Dave?'.  'Intermediates' was the cry.  One down and one
    to go.  The Ultima stayed behind for about 6 laps (I can drive WIDE
    also) and then his cells started to go off and I pulled away to
    lap him in the closing seconds of the race.
    
    Moral?  Cars only win when they last the distance.  In answer to
    the orriginal note if your cells are charged and your speed controller
    is delivering full power (chech the transmitter trims) then
    transmission is probably the problem.  A 2WD should roll freely
    without a motor.  Front wheels should spin almost for ever when
    given a quick flick, rears should certainly spin for a few seconds.
    
    Rob