Hi Fred, I don't want to seem argumentative,
but I don't think that's true, because... The motor must dissipate whatever power is not converted. This is a function of both voltage and current (and mechanical loading). With short duration pulses, the average power delivered to the motor is relatively low. If the motor is not overspeeding, (and not overloaded) then it will be within its power range. The motor has thermal mass and the temp of the motor (and wire) won't respond to the short duration current pulses. But again, it's easy enough to test out! :mrgreen: I don't know if the CVs can limit the voltage, but as you say, they can limit the output, so that max throttle corresponds to a shorter pulse, reducing the average power delivered to the motor.:thumb:
cidchase
I don't like to argue either. The only reason I am posting instead of PM is because the reasons for motors getting hotter on pulse power are not intuitively obvious, and worthy of discussion. I had to review why all experiments end up with the same results - our DC motors run hotter on pulse power.
Let's start with generalities with filtered DC. A motor is a non-linear device. The current drawn increases with the load, to a maximum when the motor stalls, appropriately called stall current - all without any change in the supply voltage. The limit on stall current is the resistance of the armature windings. Most motors will let the magic smoke out very quickly at stall current - Tortoise and a few special-design (usually very low torque) motors have a safe stall current rating.
The voltage controls the RPM. Voltage and current are almost independent of each other in normal operation. As you increase voltage, motor RPM increases, but current increases very little, if at all (there is a very slight added friction from the higher RPM). As stated, current is a function of load, RPM is a function of voltage. At a certain point, a motor receiving an excess voltage, but within its current limits, will fly apart from too high an RPM - but it won't overheat.
How can this be so? BEMF (back electro-motive force) provides some of the answers. As a permanent magnet motor turns, it also acts as a generator, generating a voltage opposite in polarity to the input. Since the impedance (resistance) of our power supplies are normally very high, the back current is negligible for the sake of our discussion. The BEMF voltage effectively reduces the voltage at the motor. If the power supply voltage is increased, the motor increases RPM, which increases BEMF to establish a new equilibrium.
The second part of the answer is that the coils in our motor serve as an inductor - an inductor which increases impedance as the motor turns faster.
This is why average current is highest when the motor is stalled. The increased impedance from the motor turning has disappeared, and so has the back emf. Actual voltage across the motor is at a maximum, and impedance of the motor is at a minimum. I know, technically, back emf and inductance are part of the same phenomena, but it helps to separate them in this discussion.
As you said, energy not transformed into rotational motion is lost as heat in our motor. The formula for energy lost in a resistive impedance is:
Energy=(current)squared times impedance times duration or
Power=I(squared)R where energy is power over a time duration.
In a linear circuit, E (voltage)=IR so one can write P=IE (for a linear circuit!). We can't use that substitution for our motors because we don't have values for the BEMF to calculate the real voltage at the motor.
Now, to look at our non-linear motors on pulse power. Heat is proportional to the current squared times the resistance (some power is transformed to motor rotation). Let's assume we are running at a slow steady speed on pulses with about 25% duty cycle (the pulse is on 25% of the time). If the pulses were truly rectangular, the average supply voltage would be 1/4 the pulse voltage.
Heat is proportional to the current squared, so a pulse generates around 16 times the heat (4 squared) as a steady state voltage equivalent to an averaged 25% duty cycle pulse. But that heat is only being generated 1/4 of the time period, so the average effect is only 4 times the heat.
Open frame motors dissipate heat far better than an enclosed motor. The downside is that their lower efficiency (larger magnetic air gap, magnetic flux concentrated at only 2 points of the rotation) means more heat produced to begin with (shown by the higher operating current).
A can motor is much more efficient, but the can enclosure limits heat dissipation. Still, the iron core of the armature is a great heat sink.
The coreless motor has no iron mass, is enclosed, and has just plastic to hold the armature together. Great efficiency, terrible heat dissipation.
A small motor has less mass to dissipate the heat into.
In the '60s and '70s, when research was being done to get our open frame motors to run slowly and smoothly on various forms of DC, pulses of different shapes were tried and applied. Pulsed power was at its best in overcoming the cogging of our motors at very low RPMs. Filtered DC was found best for maximum power with minimal heat. Eventually, it was determined that superimposing pulses on a DC base was the right answer. At very low RPMs, the pulses dominated and provided tie-creeping speeds. As RPM increased, the benefits of pulses disappeared. So in the better throttles, the DC base was steadily increased until it overwhelmed the pulses.
Pulse Width Modulation is very easy to implement in today's electronics. A power-handling electronic switch (triac, SCR, etc) is controlled by a timer. The timer adjusts the pulse duration, and the SCR conducts or is shut off. The pulse shape can be adjusted by introducing capacitance, and the pulse frequency can be part of the timing. On old power packs, "pulse power" was simply half wave rectified AC from our transformers. Pulses were 12 volt sine wave, 50% duration, with a 60 Hz frequency. Motors running on this pulse power ran hot and noisy, and could only achieve 1/2 speed.
As pulse generator circuits became easier to implement, pulse frequency went up. Today's DCC decoders "silent drives" use a much higher pulse frequency and much shorter duty cycle to get rid of the noise and some of the heat. The double or quadruple efficiency of today's can motors also help keep the heat reasonable. Coreless motors are still iffy on DCC due to the heat of PWM.
I hope this makes sense. It took me way too many years to understand it properly.
