I just read a review article on reversing ESCs in the June 2005 issue of Radio Control Boar Modeler mag, and was struck by the very low PWM rates that many of these devices use. Most are in the range of
1000-2000 hz, but about 1/3 are as low as 60, 100, and 500 hz.
60-100 hz seems to be very low, and I was wondering whether this might be ok because the hi-current motors these devices are made for might run efficiently at such low PWM frequencies. ????????
It depends on the inductive reactance of the motors.
There are so many factors involved: the integration of current into the inductor. The differential between supply voltage and backemf of the motor. If your PWM is too slow, the motor speed will be effected by the pulse. If the pulse is too fast, and you don't have a properly tuned LC circuit between the amplifier and the motor, then the motor's inductance will not allow it to use enough current.
Personally, I like to have a higher frequency PWM with an LC circut between the PWM driver and the motor, it tends to be cleaner and more efficient, but high current inductors can be heavy.
Yes, I understand what you've said here, but it doesn't really answer the practical question regarding why so many off-the-shelf ESCs can get away with using such a low PWM frequency. For whatever reason, it must be effective for the intended apps + motors.
OK, from a very basic level, PWM is basically switching power at a fixed rate with a variable duty cycle. As the frequency is reduced, the filtering properties of the inductance is proportioanlly minimized and the power surge of the pulse is applied becomes more pronounced. Now, either the motor will start vibrating with the pulses, or it has to have some serious inertial dampening like a flywheel or harmonic balancer.
Uh. Yeah. He understands PWM. His question is how can they sell something off-the-shelf which you'd expect they would intend to be able to work with the vast majority of off-the-shelf motors, but uses a frequency that suggests it won't.
Exactly :). You don't catch many robotics companies trying to sell h-bridge boards that use PWM frequencies of 60-100 hz, etc. Most people even dislike hearing their motors singing at 2-4 Khz, and prefer PWM freq somewhat higher.
Wouldn't this be more efficient in most cases? Outside of the groaning noise and possible vibrations, what's inherently wrong with the lower PWM rate? Seems to me like that would be easier on the controller. Since they have such extremely low Rdson values, I'm thinking that most heating is going to occur during switching so the lower PWM rate should dissipate (noticeably?) less heat in the ESC module. I would also think it would waste less energy heating the motor coils (by pumping up a field) instead of actually getting work done.
Sidenote: I was looking at some sites out there and came across some interesting claims by ESC manufacturers. Sounded like audio-phool claims to a large extent. For example I see that an ESC will make my motor run faster (higher RPMs) than a resistance type control can (blanket statement). Understandably, the batteries will last longer (at speeds less than full throttle), but the motor will not run one RPM faster (nor will the battery last 1 second longer) at full throttle provided that the resistance type speed control actually works right. In fact since the Rdson of a rheostat can be very nearly 0, it beats the mosfet ESC in efficiency at full throttle.
I wanted to make the following comment in that thread, but my memory of reading about the model railroad speed controller was too hazy. I think it's come back to me well enough now to post it:
ISTR a Railroad Modelling magazine with an ad or review for a power supply/speed controller which had as one of its features an unfiltered/pulsing DC output. The reason given was that at low voltage the engine would sit there because the power wasn't enough to get over the starting friction in the motor. The pulsing voltage gives a higher peak voltage and current into the motor, starting it turning at a lower average voltage than pure filtered DC voltage, and so the speed would be controllable down to a lower speed. For the ESC's, I presume the pulses of the lower-frequency PWM are long enough to get the motor turning at low duty cycle, whereas a higher frequency PWM at the same low duty cycle would appear to the motor as a low DC voltage, too low to overcome static friction. I don't know how PWM frequency would affect efficiency, but that's apparently not the only consideration. If it's important that your motors run and their speed respond somewhat linearly to PWM at a very small fraction of max speed, you apparently need to look into this. Or have a speed detect on the motor used in a closed-loop operation - but then if friction is too much, it may go into a jittery start-stop operation at low speed (the control loop overshoots to compensate), and lower frequency PWM control may solve the problem.
: I was hoping someone here might actually have had experience using : those low PWM-freq ESCs, so they could tehy use their experiences.
I've used a number of them, from 500Hz to 20KHz. There are some things that make some better than others. You'll find that most of the low cost ESC's will be around 1KHz. I think this is that it is cheaper to use a lower capability processor whose speeds don't allow much better, but often that works just fine. Cheaper motors will run better at those low frequencies. Battery powered drills are usuall in the WAY low PWM frequencies, those motors are NOT the highest quality, at least the ones that I can afford anyway. ;-) When I was racing RC cars that frequency was a hot topic. Lower frequencies delivered more "punch" and the higher frequencies delivered smoother power transfers. Here we were using $20 to $80 UNGEARED motors, so these were pretty high quality units. You'll find that if you try to use a 20KHz PWM on a cheapie Tamiya motor that you won't get the thing to start turning until you are at about 40%, whereas if you used a 1KHz PWM it would go at 10%. This is the "punch" factor. My Escap motors however just love
20KHz and will turn at 5% power - They are better motors with better quality control in the windings and other aspects. The lower frequencies, while they deliver the "punch" faster, can also be a bit rough - Look at how the various frequencies "average" out to the motors and you can see why this all happens. So, in short, the lower frequencies work just fine with motors that are cheap, and if those are what you have, that is really what you want to use.
Periodic "spiking" high supply voltage (as opposed to back emf voltage) that are not smoothed out by the inductance of the motor (or an LC circuit) will have the motor basically turning on and off, pushing it into less efficient operating ranges. This causes more loss in the coil windings.
Depends on where your heat is coming from. If it is from breaking magnetic flux building back emf during normal operation or dissipation of power in the coil windings.
It is very possible. It depends on flyback of the inductor. If the output circuit is designed correctly, a PWM signal into an LC circuit is a "power" supply, not a "voltage" supply. If the flyback voltage of the inductor is stored in a capacitor, it can exceed supply voltage. This is how a "step up" switch mode supply works.
Not true. At "full throttle" your motor may not be able to go much faster than some lesser power point. The power/speed curve is not linear.
There is a fundimental difference between "hand held" control of a motor, and computer control of a motor. This is a subtle point in the above discussion that I think was missed.
You don't want your power circuit to provide full power when it is not needed. A power crcuit delivering full power is less efficient. A constant current (at about 50%) capacity going into a motor, is more efficient than full power at a 50% duty cycle.
While a pulsed full power into the motor makes for a good human control (it feels "linear"), it makes for very inefficient motor driving. You spend too much of the time at the least efficient end of the motors power curve. Driving high current into the motors, limited by coil resistence, and creating heat. Driving at a more constant current, well below peak, the motor creates less heat.
As for "start" torque needed, that is one of those things that PID controllers are designed to provide. You don't need to over drive the motor on a constant basis for low end torque, you just increase the power delivered to the motor at startup and back it off when you near your speed target.
Hey Dennis, thanks for the input. What you say here supports what Ben said in the other post, about getting more punch and possibly better speed control at lower motor speeds. These ESCs are made to work with very hi-amp, hi-rpm motors like they use in model cars and planes, which are much different from what we use in robotics in general. Most of the ESCs in the RCBM article are spec'ed for #turns in the windings, ranging from 10 to 19. I wonder if the main factor isn't motor design rather than expense, per se. It sounds like low-PWM freq gives better low-speed performance for these particular low #turns motors.
It still seems interesting that some of the ESCs go all the way down to
60 hz PWM. You may lose smoothness, but at low duty, say 10%, you're sending a long 1.6 msec pulse into the windings, versus short 50 usec pulse with 2 Khz PWM. That sounds significant.
Thanks Ben, I just read through that thread. Very different opinions, I think. Despite what mlw was saying regarding efficiency, it seems none of the ESC manufacturers consider using LC-filters between the ESC and the motors, so the argument is somewhat moot regarding what is more efficient and what is not. However, putting everything together sounds like the lower PWM-freq gives better low-speed control and better low-end torque in the monster current motors, like used in model cars and planes, but as mlw says, this is probably more advantageous regards human control than than possibly regards computer control, as none of the ESCs has PID built-in anyways. In model-control situations, the loop is being "closed" remotely by the human, rather than locally by the computer/controller. ===============
voltage
higher
the last few comments are the primary issue. At lower PWM-freq, you have longer pulses for the same % duty-cycle compared to higher-PWM freq, and this is apparently more important for getting the low-#turn, hi-amp motors to control well and produce good torque at low speeds. At least, this is what I am concluding from these discussions.
I don't have any experience with the specific PWM system, but (depending on the inductance of the motor) using a low frequency PWM can be similar to driving 30 miles an hour by flooring the gas peddle of your car at briefly regular intervals. Yes it works but it isn't the most efficient way to do it. The motor spends too much of its time in it's most inefficient operating mode.
Well, this is really the question. The ESCs mentioned are off-the-shelf products made for a specific market, which is model cars and boats and planes which use electric motors with low-#turn windings, and which draw huge currents - 20 to 50 to 100+ amps. One suspects that the customers want some juice when they push the buttons, else the ESCs would not be around for long. The problem is to understand the situation as given. I think Dennis and Ben must be on the right track.
One of the issues with such high currents is overall power. What the hell kind of power source can deliver 50A, at any significant voltage, for a significant amount of time?
50A at 1V is 50W.
50A at 10V is 500W.
50A at 12V is 600W. FYI, that's 10 common home light bulbs, or a microwave oven.
It is exactly at these currents that "power" be managed over "voltage." Now, the torque at the motor shaft is directly proportional and linear with respect to the current applied to the motor.
Since the ESCs are used with model cars and airplanes, they're usually powered by NiCad packs, prolly AA cells, and they're pulled flat in just a few minutes. So for the exactly 3 minutes and 10.43 sec [whatever !!!] before the batteries give up, they're pulling maybe 8.4V
50A = 420W, or 9.6V * 50A = 480W. Given life is so short, I imagine the guys running the cars want some real performance.
I've found this information is a book "Build your own combat robot, by Pete Miles and Tom Carroll"
NIMH to convert to 6 minute capacity multiply by 0.92, then multiply by 10 to get 6 minute current.
For example a 6.5AH Panasonic pack of 10 D (part number HHR650D) works out to have a 6 minute current of 60A, the terminal voltage drops down to 10.3V.
Ni-Cad to convert to 6 minute capacity multiply by 0.9, then multiply by 10 to get 6 minute current.
For example a 4.4AH Panasonic pack of 10 D (part number P400D) works out to have a 6 minute current of 40A, the terminal voltage drops down to 10.3V.
SLA to convert to 6 minute capacity multiply by 0.33, then multiply by 10 to get 6 minute current.
For example a Powersonic 12V 17.5A (part number PS-12180) works out to have a 6 minute capacity of 58A, the terminal voltage drops down to 11.5V.
Both the NIMH and Ni-Cad have a rated max current of 100A, the voltage at this current output is 7.9V for Ni-Cad and 10.3V for Ni-MH. The SLA have a rated max current of 175A, but it's voltage drops down to 9.78V at that current output.
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.