In this issue of the Controlling Interest, Bernard covers the ESC "brake" and several ways that different manufacturers have implemented that function. Bernard also gives us a detailed look at a schulze future-45Ko brushless controller.
It is time for another installment, and this time without quite such a long gap since the last one as before. I realize that we have all been distracted these past few weeks, but I find it more than a little comforting to think about our small aircraft and other means of having fun to help us adjust to whatever "normal" will be in the weeks and months ahead. For those who are wondering, so far the announced cuts at Boeing are not going to affect me.
I haven't gotten much feedback from readers on the last column, especially my question about which of the "buyer's guides" I should do next. So, in keeping with where my flying interests are at the moment, I'll work on both the S400/480-sized (15-20A) brushed motor controllers and the 40A +/-5A brushless controllers, since I'm flying both types right now. These big guides take a lot of time, bench, and flight-testing, so don't expect the first of them before early next year. By the way, I do watch the reader ratings, and am glad to see that those who choose to vote generally are finding my ramblings of some use.
I also want to say thanks again to Bob Kopski, Electric Power columnist in Model Aviation (the Academy of Model Aeronautics' magazine) for plugging A Controlling Interest again in his November 2001 column.
This time I think I will cover the brake function of our ESCs - in particular, the different choices we have in enabling/disabling them. For the first time, I will have a brushless controller as the ESC of the month. It is one I have been flying with off and on for nearly a year now, and it has had its software updated once - a schulze future-45Ko.
Terminology: Low Voltage Cutoff
Let's look at another acronym. Way back in my very first column in August of 1999, I discussed the various functions that the device we call an electronic speed control can perform. This, of course, includes several things besides electronically controlling the motor speed! In that column, and in a couple of others since, I discussed the "BEC/motor cutoff function". By calling it "BEC/motor cutoff", I was including both the battery eliminator itself, the circuitry that powers the radio system from the motor drive battery, and the motor cutoff, which is the circuit which tries to assure that the drive battery is never drained to the point that radio power is lost. This is usually done by shutting the motor off at some predetermined fixed or calculated voltage. Recently, in some e-mail thread or other (I think it was on an E-Zone discussion board) someone referred to this latter function as the LVC or "low voltage cutoff". This describes what the circuitry is trying to do very well - cutting the motor off (or down) before the voltage to the receiver gets too low.
I liked referring to it this way so well that I posted I would steal this idea for use in a future column. So, from now on, I plan to refer to the "motor cutoff function" as the LVC or Low Voltage Cutoff. As well as making the functional distinction clearer, it will give me a clear place in my data tables where I can put information such as the cutoff method, the voltage at which it occurs or how it is calculated, whether you can restart the motor, and how to do so.
For those of you who would like to look at them, the columns on battery eliminator circuits and their advantages and cautions are the ones from October 1999 and November 1999.
Puttin' On the Brakes - Or, Rather, Taking Them Off
As I've written before, there are many ways that various speed control designers choose to match the throttle function's operation with the transmitter stick's throw, which range from the very simple to complex. Recently, the number of ways the brake function operates has been going that way, too. A few years ago, an ESC had a brake function, or it did not. If you needed a brake, you bought an ESC that had one. If you didn't want it or need it, you bought one without.
Now we have hard brakes, soft brakes, delayed brakes, and even proportional brakes. We have brakes that can't be disabled at all and several variations on disabling them, which range from cutting wires, to those that require programming the brake function every time you power up. In the next few paragraphs, I will go into some of the different approaches to disabling brakes so that the next time you go shopping for an ESC, you will have some idea of what the choices are and perhaps what sort of approach to brake selection you prefer. Along the way, you'll probably figure out what I prefer too.
I'm primarily a sport power flyer, so brakes in ESCs are not the most important feature of the controller to me. I usually don't pay too much attention to them, except when I'm working on a review, or when something about them stands out. None of the planes I'm currently flying benefits very much from a brake and none are currently equipped with folding propellers. However, I have found that having a brake on, for example, my much mentioned LT-25, is not a problem. The airplane does glide better with the prop stopped than it does with the prop windmilling. This is helpful if I decide I want to soar a bit, but is a pain when I'm trying to land.
Besides wanting to be able to use a windmilling prop to help the plane land, another reason to disable the brake is if the plane is equipped with a reduction drive, like a belt drive. Some brakes come on so hard and fast that it sounds as if the belt is going to break when the prop is stopped so suddenly, though this hasn't actually happened yet. Consequently, I find it advantageous if a brake-equipped ESC provides a way for me to disable the brake.
As I mentioned, some ESCs that have brakes provide no means to disable the brake at all. This is often done for simplicity in the design of the controller. Oddly enough, many of the controllers whose brakes can't be disabled at all also have fixed rather than variable endpoints. Examples include the Astro Flight 215D, the FMA Direct units, the WattAge units, and older brake-equipped units from Castle Creations for both conventional and brushless motors.
On these ESCs you can't disable the brake. From top to bottom: Astro Flight 215D, WattAge IC-30A, and Castle Creations Griffin 40
The first means of disabling a brake that appeared was by physically changing something, either by cutting a small jumper wire attached to the ESC, such as the now unavailable Viper Models MD-52. Or, as is the case of the Jeti microprocessor based controllers for brushed motors, the brake can be disabled by installing a little jumper plug onto the two pins on the side of the unit provided for this purpose. Aveox uses a related approach in their 260 series controllers. There are four DIP switches on board these units. Switch number 2 controls whether the brake is enabled or disabled.
On these ESCs you electrically disable the brake. From top to bottom: On the Jeti JES-350 you install the jumper on the pins on the right side. On the Viper Speed Demon 202, you cut the blue jumper wire on the left side. On the Aveox H260 you switch the brake on and off with the DIP switch indicated by the knifepoint.
Other methods of disabling a brake depend solely on the software in the ESC. These fall into a couple of groups: those that can be programmed once and remember the brake selection until you change the programming, and those which require you to select which brake mode you want every time you power up the system.
Programmable ESCs that store the brake setting handle this function several different ways, from the simple to the complex. For example, the Kontronik Star family of controllers I wrote about in the May 2001 column has six selectable modes. The brake settings range from "off" in Helicopter mode (of course), and "motor plane mode", to a quick hard brake in "competition mode", to a proportional brake in "car mode". The brake characteristics you get depend on which mode you select. Aveox took another approach in the 60 and 160 series of brushless controllers. In these, functions like the start (off) point, the full throttle point, and the brake engagement point are separately programmed and stored. To disable the brake, the brake point is programmed to be above the "off" point.
The simplest approach that I know of right now to programming the brake to be either on or off and having that setting remembered, is used by Jeti in their sensorless brushless controllers. To toggle the brake from on to off or vice-versa, you simply power up the system with the your transmitter's throttle stick in the full-throttle position. Wait five seconds for the unit to beep four times. Next, bring the throttle stick to "off". The mode is now changed. It will stay that way until you go through the same process again. It also lets you know which mode it is in every time you power up by beeping once if the brake is enabled and twice if it is disabled as an "I'm armed" signal.
Two controllers that can be programmed to disable the brake: the Aveox L60 and the Jeti JES 40-3P. Both remember the selection until re-programmed
Finally, there are those controls that allow and require that you select whether the brake is on or off every time you power the system up. This is selected, in every case I know of today (please correct me if I'm wrong), by the position of your transmitter stick when you power up the system. For some reason, German makers implement this option. Orbit, Kontronik in their Sun series of controllers, and schulze in both the slim series of brushed motor controls and their future series of brushless sensorless controls work this way. I have heard that the Czech MGM Compro controls work this way as well, but I have not had one to look at, so I can't confirm this.
For all of these, the brake is enabled if you power up the system with the throttle stick at the "off" point, as we all expect to do. To disable the brake, using any of these controllers, the system is powered up with the throttle stick in the full throttle position. (This also defines the full throttle endpoint.) The stick is then brought back to low to define the "off" point and arm the controller. Schulze adds a third choice: power up the system with the stick in the middle to select "gearbox/belt-drive mode" which gives you both a softer throttle response and a soft brake. The start point is then defined by pulling the stick to off, and high is set 0.6ms of pulse duration higher.
Two ESCs that allow and require you to select the brake mode every time you power them up: schulze future-45Ko and Kontronik Sun 4001
I don't really care for this approach at all. In order to fly with the brake disabled with these units you have to remember every time you power up that you have one of these units in your plane, and that to disable the brake you have power it up with the stick at the high throttle point. I suspect most folks just leave the brake on, and most of the time this is probably fine. However, for the schulze controllers in particular, the default brake action is quick and hard, which leads to my worry about the belt on my Modelair-Tech H-500 in the LT-25.
I'll bet you didn't know brakes were so complicated, did you? Well, designers of speed controls each have their own likes and dislikes, just like the rest of us. Some things are more important to one than another. Therefore, we see different solutions to the same problem. At least it makes things interesting.
Speaking of schulze controllers...
Speed control of the month: schulze future-45Ko sensorless brushless controller
I think the first time I heard of sensorless brushless controllers, the ones I heard about were schulze units. In addition to their world-class chargers, schulze have been leaders in introducing sensorless controllers for brushless motors. Being pioneers means, among other things, that they've had time to refine and improve their product several times. In their current sensorless line, the futures, they are up to version 15 of the software. By the way, schulze is the only ESC maker that actually tells you what the software version is on the label. They have also added a number of selectable modes to cover a very wide range of motors and applications.
As I wrote in the September 2001 column, one of the challenges to sensorless brushless systems is that the controller has to infer the motor's rotor position from just a little motion in order to properly start the motor. It also has control over the motor's timing. With the large variety of motors out there with two poles, four poles, and six poles (at least) in their rotors, as well as different construction approaches that respond differently to timing, it is a real challenge to make a sensorless controller that can cope with that wide variety. The schulze futures, with the "super high performance algorithm" or "SHPA" go a long way toward being able to run most any brushless motor. I have driven several very different motors, from Aveox, Jeti, and Hacker with my future-45Ko and it starts and runs all of them reliably and with only an occasional hesitation. It seems to help to go immediately to about 1/4 throttle stick, rather than to trying to ease on power.
The family of sensorless controllers includes, at last count, 32 different models for sport/aerobatic/scale airplanes, FAI competition, pylon racing, helicopters, cars, and boats! The future-45Ko is a sport/aerobatic/scale airplane controller rated for 45 Amps of continuous current, and is opto-coupled for lower RF noise, which also means no battery eliminator circuit (BEC). It is rated for 6-17 cells. Even though it has no BEC, it does have a low voltage cutoff (LVC) function, as many non-BEC controllers do, which is a self-protective measure. The LVC here prevents the voltage going to the microprocessor in the ESC from going too low and "crashing the computer" with unpredictable results. The LVC cuts off at about 5.6V. However, it can be immediately reset by bringing the throttle back to "off", and then advancing it.
The "-Ko" variant also has a "finned" heat sink on it, so that it can be run at part throttle constantly without overheating from MOSFET switching losses. This is more of an issue in brushless controllers, because they never get into a state where the MOSFETS are just turned on and left there as with a brushed motor controller at full throttle. A brushless controller is always switching the three coils of the stator of the motor on and off. Even so, brushless controllers also heat up more at part throttle than at full throttle. The big heat sink on the "-Ko" members of the future family keep them quite cool, even with extended part throttle operation. The heat sink only adds about 1/4 inch in thickness and perhaps 1/10 ounce (~3g) in weight, so is well worth it.
The back of the future-45Ko, showing the large heat sink
The "future" series on controllers has a number of selectable modes and functions. Some of these functions are combinations of switching frequency and timing angles that are programmable and remembered until the programming process is repeated. Other functions, specifically the selection of brake on or off mode or "gearbox mode", have to be done every time you power up, just as with the "slim" series I wrote about in my February 2001 column, and as discussed above.
The future-45Ko (and all -bo, -be, -Ko and -Ho versions with a current rating over 25A) have three choices of switching frequency and two timing settings, for a total of six programmable choices. The three switching speeds are 9.6, 19, and 38 kHz and the two timing options are referred to as "hard" (more timing advance) and "soft" (less advance). There are recommendations for which motors each of the six possible combinations are best suited for in the instructions. Rather than paraphrase those instructions, I'll just refer you to schulze's web site, from which you can download their instruction manual. Go to the schulze index, and then scroll down to the bottom of the menu in the left frame on the screen. From there, select "download op-instructions". Next, from the updated list in the right frame, pick future "SHPA" for the most recent instruction book.
I refer you to their instructions also for a description of how to select each mode.
Physically the future-45Ko is fairly compact for a brushless controller at about 2.9 X 1 X 0.6 inches and it weighs only 1.5 ounces, ready to use. However, that weight includes no motor-side wiring as the future-45, like many schulze controllers, has 3.5mm female gold bullet connectors attached directly to the board. The mating plugs are supplied with the unit and installed in the sockets.
The future comes with 3.5mm gold bullet connectors for the motor installed in the onboard sockets. Here one is shown removed.
By the way, Jeti happens to be using the same size connectors on their Phasor motors, so a Phasor can be plugged directly in to the future-45. For most any other motor, the supplied plugs need to be installed on the motor leads, or jumper leads made up. I made a set of jumpers ending in Powerpoles to mate with my Aveox motors, but have installed the bullet connectors directly on the leads of my Hacker B50-13L. I like these connectors well enough that I'm considering changing all my brushless motor to controller connectors for anything bigger than the Astro 010 to these. Care must be taken not to overheat the little spring contacts on the male plugs when installing them. As a minimum, insert them in a loose socket to act as a heat sink, or remove the spring contacts altogether while soldering, and then reinstall them.
Here's another view, showing the short 3.5mm bullet to Powerpole jumpers I made up to connect the future to my Aveox motors
On the test stand, the future-45Ko behaves very well. Startup is generally smooth and with V15 software, the minimum motor power (which used to be an issue with sensorless controllers) is less than 2% of full. The throttle curve is fairly linear with RPM and smooth from bottom to top. But, being linear, most of the effective throttle range is in the upper end of the throttle throw. Response is damped a little bit, but only enough to make it feel very smooth in the air.
In two of the three brake modes, it is self-adjusting for both endpoints (and consequently the range). In the default "brake on" mode, the initial throttle stick position is taken as "brake", and the full throttle point is set when you advance the stick to full throttle the first time. When power is first applied, the future-45Ko will indicate the timing mode it is in with the appropriate number of beeps (e.g. four beeps for Mode 4), followed by a single beep two seconds later. Advance the throttle after this single beep to full throttle to set the high point. The motor will hesitate just a bit to let you know the upper end has been set. The brake comes on hard and quickly when the stick is returned to the low point in this mode.
The "brake off" mode is selected by powering up with the throttle stick high. As soon as power is applied, the system will beep to indicate which timing mode as described above. Two seconds later, it will beep twice more, indicating selection of "brake off" mode. This also sets the full throttle point. The "brake off" is then set by bringing the stick down to the bottom. The controller will beep the motor once to tell you it is armed.
"Belt drive" mode is selected by powering up with the throttle stick in the middle of the range (pulse duration between 1.4 and 1.67 ms). Again, the timing mode is indicated by the appropriate number of beeps, followed in two seconds by three more beeps, which indicates belt drive mode. You then can select the brake to be off by advancing the stick and holding until you get two more beeps. Next, pull the throttle stick to the low point to set the start point. The system will beep once to tell you "I'm now armed". From here, a fixed range of 0.6 ms takes you to full power. If you want the brake to be enabled, just pull the stick to the off point after the three-beep indication. The unit will beep once and be armed with a fixed throttle range of 0.6 ms.
In belt drive mode, throttle response is slowed down, especially the initial run up to full power. Also, the brake comes on more softly than in the default brake mode.
I find this mode programming every time you power the system up to be a bit too much to remember, as I mentioned above in the general brake disabling discussion. But, it does allow the maximum of flexibility in use.
Shutdown with a loss of signal is decisive after a brief delay during which there may be a brief burst of increased throttle. The brake engages or not depending on which mode you selected upon power up.
For flight tests, the future-45Ko went into my workhorse LT-25 behind an Aveox 1114/4Y on a 3.6:1 Modelair-Tech H-500 belt drive. Power is 10 cells of various types and my current prop of choice is the 14X7 APC thin E-prop. This combination draws in the upper 30 A range at full throttle.
First, though, I had to complete a range check. This airplane has a Hitec Super-slim receiver in it, and it is commanded by a Multiplex Cockpit MM transmitter. The receiver is at the back of the cabin, above, and behind the entire power system. On the ground, range checks showed only a moderate decrease in range with the motor running slowly as with it off - a good performance.
In the air, the future-45Ko behaves just as the bench tests would indicate - very well indeed. Startup is reliable and fairly smooth. The modulation of power is without surprises throughout the throttle range. Because it is a linear throttle, most of the usable stick travel is in the upper half of the stick's motion. The controller is cool to the touch after 8+ minute flights of mixed aerobatics, cruising around, and shooting touch and goes. The heat sink clearly does its job well.
Landings with the prop turning over slowly to provide a little extra drag are easy to do. I should note this was not so easy with an earlier software version, where the minimum input power was over 10%. At that setting my LT-25 simply would not land. Now, with the super high performance algorithm or "SHPA", this has been cut to less than 2% and landings with the prop just ticking over are simple.
The only problem, really, comes when I power up the plane with the throttle stick low as is my habit, then take off. This leaves the brake on and I usually remember that only when I pull the throttle all off the first time in the flight and I hear a loud "clunk" from the plane as that big prop suddenly stops. I keep worrying that I'm going to damage the belt, though so far it has held up well. As I've remarked before, if all of the controllers you have are those that require you to power up with the stick high to disable the brake, then you probably won't have this problem. But for me, well, maybe it's the beginnings of old age setting in...
Overall, I like the future-45Ko and look forward to using it in an application that uses more of its capabilities.
Oh, one other detail: Some time ago, I made a statement that no brushless controller could also be used to drive a conventional brushed motor, then a couple of readers pointed out that the schulze future series of controllers could in fact do this. I mentioned that in a subsequent column, though I wondered why anyone would do this. Apparently, schulze agree because at the very end of the instructions it says, "brushed motor mode is no longer available from production date 23.8.00 because of new improvements".
For weights, measures, and such, see the tables below.
In my next column, I plan to continue the brushless theme with a look at the second-generation Jeti brushless controllers. What other topic I'll cover hasn't popped into my head yet.
Work has begun on a 40 Amp brushless controller "buyer's guide" as well as one for S400 brushed motor controllers. I could also be doing one on 2-5A controllers for low power park flyers. Which one is of most interest? Let me know at abcawley at earthlink.net (this is a new address) or abc_quiet_flight at operamail.com.