(ET) Lots of Amps.... Saving the motor.

[Larry Elie <lelie ford com>]

Remember this item?

----------
From:   Randy Kanary[SMTP:rkanary kanarysweb com]
Sent:   Tuesday, October 12, 1999 12:52 PM
To:     Larry Elie; elec-trak cosmos5 phy tufts edu; 'Lee Richardson'
Subject:        Re: (ET) RE: 600 ?

;Yes, that was yours truly! Yes, the traction motor can draw that much, if
;some safeguards are circumvented, and yes sustained loading at that level
;will evaporate the fuse link! I spoke to Mr. Gunn when I made my last 
parts
;order about "ratings versus real world experience".If the power pack is
;capable of sustaining the load, and armature rotation is still
;occurring,rated current draw, and therefore, horsepower and torque ratings
;can be exceeded. Don't try this at home! Unless of course, you don' t mind
;fixing it later!EV you later!

;RJ Kanary@ Bandi Bros. Inc. ATRA ® Member Shop
;Member TRNi 1&2 Since 1998
;ASE® Certified Master Auto Technician Since 1992
;Member Tech Line Associates Since 1987
;rj kanarysweb com


Time for a mini-lecture.  Since we have some EE's on the list, you guys 
can ignore this but I 
think it wise to explain how electric motors usually 'burn out'.

Long.

All magnetic-field electric motors work by trying to generate large 
magnetic fields across small 
air gaps using magnet wire.  Permanent magnet motors only have to do this 
for the armature, while 
other motors do it for the field as well.  The magnetic field achieved 
depends on lots of things; the 
geometry of the system, the gap between the field and armature, the 
quantity and type of steel used 
for the armature, and, most importantly for this discussion, the number of 
turns of wire and the current 
carried through the wire of the armature.  Many of these parameters can't 
be adjusted too much; the 
type of steel is going to be a low carbon steel; 1006 or 1008 are among 
the best for a motor for magnetic 
saturation, and they are cheep, so they are used.  The geometry factors 
are governed by both the 
size of the package (into which temperature may figure) and how good of a 
job of fabrication and 
bearings that can be put in the motor.  As the air-gap between the 
armature and field is reduced, the 
magnetic field increases.  Obviously, you can't get the gap to 0.  In 
reality, for large motors (or generators 
for that mater) you don't get very close.  A typical 'good' gap for a 
motor of the size of the ET is about 
0.8 to 1.0 mm (~.03 to .04").  Before someone says why not bring it down 
to .001", remember, this is 
the average gap over a large area, and has to include all bearing slop 
etc.  I doubt the ET's motor is 
that low, but have not miked one.  Anyway, there are small motors with 
gaps of 0.4mm, so it might be 
done.  Regardless, the designer has to pick a gap and stay with it.  

The next, and I maintain hardest part, is to figure out what type of 
winding will be used.  Trade-off time.  
The wire has a finite size.  In some exotic applications, like an 
accelerator, people use square or even 
hexagonal windings, which can carry several percent more current in the 
same area, but for every motor I 
have seen round wire was used.  Again, for exotic applications, one might 
use silver for the lower resistance 
per foot, but every motor I have ever seen used wires of electrolytic 
copper.  So we now have a fixed material, 
but the wires must be insulated.  There are dozens of insulators, or 
'varnishes' used, but in reality, this is 
cost driven.  Low temperature varnishes are cheaper to make and use than 
high temperature varnishes.  There 
are also multi-coats, but these add a tiny bit to the diameter, and add 
lots of costs.  The highest temperature 
insulation I have seen in a motor is a triple coat of Kapton, which flows 
at about 250C.  This wire cost 
over 10 times what a single coat of one of the poly coats, and I'm sure 
the ET didn't use it.  Now, back to 
the trade-offs.  The resistance of the wire depends on it's gauge, or 
diameter.  The diameters in principle are 
continuous, but in reality, the gauges are fixed by what are known as 
'half-gauge' sizes.  The designer is 
forced to used these incremental values of diameter and by definition 
incremental resistance per unit length.  
Smaller wire will give more turns, and the field is proportional to the 
product of current times number of turns, 
so a small wire may sometimes win out in power density, if you have a nice 
high voltage available.  However, a 
high voltage requires thicker insulation, taking up more space.  Sigh.  
You can't win by just increasing the voltage.  Regardless, eventually, the 
number of turns and room temperature resistance of the coil are eventually 
fixed.  
Now we are part way there.  BTW, for this discussion, I am assuming a 
'single' winding.  In real life, you may have 
multiple windings in order to match the resistance, not to mention 
multiple taps for different speeds.  Life isn't simple.

Now the geometry is 'fixed'.  Suppose one now has chosen a diameter that 
yields n turns with a total room 
temperature resistance of R.  R has to match the application; for a 36V 
nominal ET, an R of .9 ohms 
would be drawing 40 amps, at room temperature.  In case you hadn't 
noticed, I say room temperature a 
lot at this point.  There's a good reason for that.  When you pass current 
though a wire, it warms up.  
Fine, but as it warms up, it's resistance changes.  The formula for the 
resistance change for metals is;
((Vt/It)-Rrt)/Rrt=Temp.Coefecient*DeltaT(Centigrade).  For Copper, the 
generally accepted value of the 
Temperature Coefficient is 4.07x10-3.  If you were using the coil as an 
RTD, you could re-write this as; 
DeltaT=(V/I-Rt)/(Rrt*4.07E-3)  and compute the temperature of the coil 
very accurately if you can measure 
the V and I accurately, and know the R at rt accurately.  Now, we know 
that Vt/It is the effective resistance 
of the coil at t.  As the temperature increases, R increases.  Lots.  For 
the .9 ohm rt example at 40 amps 
above, as the coil reaches 100C (a delta t of 80C) we get a resistance of 
about 1.2 ohms.  Not too much?  
Well, remember, the ET is basically a voltage source (and one that drops 
with charge... but that's another 
story) so for our 36V example, we now have only 30 amps available.... just 
by warming up the motor.  Now, 
I don't know the temperature rating of the insulation in the ET, but I'm 
going to guess it to be no higher than 
160C, beyond that cost gets prohibitive.  For the 160C case, delta T at 
140C, R goes up to 1.4 ohms.  Now 
our 36V system will only deliver 25 amps, way below the 40 we got at room 
temperature.  Although these 
values are low for the example, in a continuous mode the % drops are real. 
 All is not lost; by putting 
out less current, the system is slightly 'self-healing'; you can't put out 
enough current to do as much work 
or to heat the motor quite as fast.  However, motors aren't usually 
designed quite that well protected.  The 
power the motor puts out is a selling feature.  You design to put out as 
much as you can, even if only for a 
short time or 'duty-cycle'.  Then you put a thermal interrupt in the 
circuit at a temperature just before the 
insulation melts.  This is how the ET gets away with these high currents.  
Without the interrupt, eventually, 
the insulation will melt, and adjacent windings will touch.  The 
resistance drops quickly, calling for more 
current, locally heating the motor even more and eventually burns out a 
wire.  The other failure is a brush 
will overheat by the same method.  The motor stops.  For quick 
'short-circuit' loads, they put in a fusible 
link as well; don't forget the contractors also have maximum current 
ratings.

What does this mean?  First, the current in the motor will always depend 
on both the battery voltage and 
the temperature.  People who set records for electric cars speeds often 
cool the motors with dry ice.  If you 
want to really pull a heavy load, do it with a cool motor and a full 
battery.  The GE should do better on a 
cold day, unless the temperature is low enough that the batteries aren't 
delivering properly, but that too is 
another subject.

I hope this wasn't too long or boring.  I have designed a few magnetic 
field devices, and this is about the way I 
go about it.  The GE isn't bad.

Larry Elie