August 25 2007, Kim Doherty on electric basics (www.stunthanger.com)
A Primer on Building an Electric Stunter
I will try to answer in brief some of the questions that keep coming up and give some guidance on the design and construction of an electric stunter. I am not going to directly cover how to choose the various power train components (this has already been done quite adequately by Mike Palko and Dean Papas among others) but rather look at the nuance of making those decisions. Everything I am about to say is strictly my own opinion. It is however for the most part based on fairly extensive research and development of my own plane “SHOCKWAVE” and the continued development program that we have followed over the past year since the 2006 W/C’s.
Where to start?
I think you want to start by defining the performance envelope of the plane you wish to build or modify. Are you looking to conduct some “relatively” inexpensive experiments with electric power where the final outcome / performance is not critical or are you designing a plane to win the next W/C’s. Your approach will determine to some extent the total cost, time frame required and resources necessary to develop an electric power train and / or model.
I do not think that you will meet with much success or satisfaction if you try to modify an existing airframe to achieve top level competitive results. The weights of the various components and the options for mounting them and being able to achieve a proper balance point without adding even more weight for balance suggest to me that this is not the most successful approach. I also believe that there are economies of scale involved in the performance of the various components and that you will be more successful at building a larger model (.60 size or larger) than a smaller one. You can of course build a “.40” size electric model and it could fly very well. If you are looking to optimize each component then the larger the model the better.
It would be unfair of me to suggest that to develop a highly competitive electric model is either easy, non complex or inexpensive. “SHOCKWAVE” took a team of two electrical engineers, two world class F3A pattern flyers, a full CNC machine shop, a world class furniture maker, the expertise of one of the top component vendors, a custom Bob Hunt lost foam wing jig, and my self 1400 hours to design and construct. From the first phone call to discuss the possibilities to the day the plane was put in the box for the trip to the W/C’s in Spain took just over a year and $4500.00 U.S.
There is no question in my mind that you can achieve a high level of satisfaction with an electric plane for significantly less money and less time. I do urge caution though as making uninformed decisions could lead to costly errors and less than satisfactory performance.
So let’s get down to the business of designing and building an electric powered airplane.
Look carefully at the first paragraph under “Where to Start” above. You need to do this and refer back to the outcome as you move forward since there are literally dozens of motors, ESC’s and batteries to choose from. You must also be realistic in what type of performance you are looking for and your personal ability to achieve it. So here is the first and most important question you must answer.
How much can my plane weigh?
That is a fairly nebulous question given that planes light and heavy have over time performed quite well. Knowing that the power train would not be light I approached this by looking for an example of a large highly refined stunter that could fly competitively at a fairly high all up weight. I remember Paul Walker saying that he liked his Impacts when they weighed around 63 ounces. Given that there could be up to seven ounces of fuel in the tank I settled on an all up weight of 70 ounces. Within that I would build my plane. (final weight was 70.5 ounces)
How much power is required?
If you want open class, expert level performance that makes no concession to gas powered models, you should look at providing the capability to generate somewhere around 8 to11 watts for ever 1 ounce of model weight. You might not use all of it but you will want to make sure that you could if you wanted to or had to.
“SHOCKWAVE” has been flown on a low power setup putting out just over 500 watts to a high power setup putting out just over 750 watts. All with the same motor, battery and ESC combination. The difference is in the propeller, line length, lap speed and the rpm chosen.
Extensive bench testing of the power train was conducted with computer based data acquisition using the Hyperion E-Meter. This allowed us to monitor and capture data such as volts, amps, rpm and temperature of all power train components with each of the various prop and rpm combinations. This is where you can make or break your system. Without testing (especially with a new unproven setup) you may end up short of amps, over current, blowing up batteries, melting down motors etc. It is cheap insurance and easy to do. You do not have to be an engineer to evaluate the data.
Selecting the power train components
Now that you have made the power decision, you can begin to select the various components that make up the power train. To begin to understand the relationship between volts, amps, watts, rpm and temperature you should look at purchasing a software package such as MotoCalc (motocalc.com) or other similar software. This software lets you play “what if?” with hundreds of components without the risk of damage or unnecessary expense. It will however give you only an approximate assessment of the performance you can expect. You should also avail yourself of articles written by Mike Palko and Dean Papas on the subject of choosing each component. These will provide you with the technical knowledge you need to make some educated decisions. While you are at it, you should begin to frequent the various electric flight forums on RCgroups.com especially the batteries and chargers forum and the power train forum. At some point prior to beginning to lay out the physical structure of the plane you will need to purchase these components so you can confirm your analysis with some real world testing.
If I can attribute the performance of “SHOCKWAVE” to one aspect of our development program it would be the extensive testing that we carried out. I used a Hyperion E-meter with the computer interface to capture real time data such as rpm, volts, amps, and temperature. This information proved invaluable in determining the validity of each of our component choices. Once we came back from the W/C’s we started using the Eagle Tree Micro Logger to record real time in flight data. Remember, once you launch your plane, you are going to fly to the end of the selected time even if your battery has exploded and the magnets in your motor are falling out of the can. Even at the low end, this is expensive equipment and a healthy dose of caution will serve you well. At a minimum, I would recommend both the E-Meter, a good digital multimeter and an IR temperature gauge..
Since you have already picked a number of watts per ounce of airplane you must know how much you think your airplane complete with power train will weigh. “SHOCKWAVE” was built to a weight budget of 38 ounces fully painted without any power train components and 70.5 ounces ready to fly. You will then need to make some educated ( personal knowledge combined with other sources such as MotoCalc) guesses as to what prop, diameter, pitch, number of blades and rpm range will be required to produce the level of performance you are looking for. Each motor will have a specified rpm capability signified by the “kV” (thousands of rpm per volt) nomenclature in the specification sheet or listed in MotoCalc. Realize that you will not be able to achieve 100 percent of the rated rpm under load. While you are playing “what if?”, keep looking at what the amp draw will be. It is all too easy to pick a motor that could seemingly provide the required rpm but might require the ESC and battery to operate at amperages that produce temperatures close to those on the surface of the sun. As well, the battery must have a high enough “C” rating (ability to give up its charge during a given period of time) to be able keep up with the current demand. It is in this area that you MUST do your homework and perhaps learn some new concepts if you are going to end up with a successful setup. You will notice that I have been somewhat vague about precisely which motor to use for a particular application. The reason for this is that there are so many possibilities and each of them could be made to work. Electric motors do not have as narrow a power output potential as do IC engines. The final power output of an electric motor is very much governed by the battery you choose, the prop you intend to swing and the rpm you intend to run.
There are however some general principles that you may wish to consider. Motors with more winds on the armature will turn lower rpm’s and produce more torque. Motors with fewer turns will turn higher rpm’s and produce less torque.. Not all motors weigh the same amount for the same power handling ability. Not all motors are made as well as others. Some are more electrically efficient. Some motors provide built in fans for cooling. (I strongly recommend a built in fan) Some have three ball bearings on the shaft some have two. Larger motor cans have more rotating mass further from the axis of rotation and produce more torque. Oversized motors run cooler since they are not pushed to their limit. In general, you get what you pay for. IMHO at the current time, the best motors are produced by Plettenberg.
As I see more and more posts on electric flight it seems evident that many people believe that there is only one motor to use for all types of control line stunt flying. (the AXI 28/26 – 10) While it has been flown quite successfully by Bob Hunt in his particular setup, it is NOT the motor for ALL applications any more than the Plettenberg Orbit 30-12 that I fly is. There are dozens of very good motors available to choose from. If you insist on copying a particular setup I would suggest that you research all of the specifications of the plane you are copying. Choosing to copy most of a particular design but not something else such as airfoil thickness or empty weight could lead to reduced performance.
The electronic speed control (ESC) is the heart of the electric power system. The ESC controls such things as the rpm, rpm step size, rpm accuracy, speed of response to rpm changes and management of Governor Mode.The ESC also provides us with “Brake” functionality to stop the prop prior to landing.
The motors that we use are called brushless out runner motors or rotating can motors. This is due to the stator (the part with all of the windings) remaining stationary and the housing or “can” rotating around it. There is no segmented commutator as there is with a brushed motor, so there is no physical way to switch or “commutate” the electricity to spin the motor. This function is carried out by the ESC by electronically switching the current on and off to the wires that are wound on the stator. The ESC switches the current on and off very rapidly to control the speed of the motor. The motor is either “all on” or “all off”. It is never partially energized. This switching produces heat. We run our motors (control line stunt) at a constant part load and hence more switching work (and more heat) for the ESC vs. running flat out. This is why I chose an ESC that had a heat sink built into it. In the 43 degree C daytime temperatures of Spain I was glad I did.
In regular practice the ESC would receive a signal from a radio receiver to indicate how fast it should turn the motor. In our application we need to provide a similar signal to the ESC to control start, ramp up speed and duration, flight rpm, brake application and shut off. The ESC processes the signal from the timer / processor and reacts accordingly While most ESC’s will run your motor, not all of them will provide you with a control line specific program. The quality of Governor Mode implementation is also very much dependent upon the manufacturer. This is where a bit of “Black Art” is involved.
A control line specific ESC is now available from Schulze (Schulze 18.46k F2B ESC) which provides an rpm range of approx. 7000 to 12,000 rpm, an rpm step size of 20 rpm and Brake functionality while in Governor Mode to stop the prop at the end of the flight. This plus the Timer / Processor which will be released shortly by AeroLectric (Kim Doherty & Pat Mackenzie) is the setup which is currently being flown by Paul Walker.
When I say battery, I am talking about the whole pack. If you are using Lipo's it would be beneficial to also talk about how the "pack" is constructed. In my case my battery pack is described as "5S2P 4200 mAh". (milli Amp hour) This means that I have 5 cells (each cell is 3.7 volts and has a capacity of 2100 mAh) in series as a unit and two of those units in parallel. (total of ten cells but for the purpose of charging only 5cells) giving me 5x3.7 volts = 18.5volts and 2x2100mah = 4200mah capacity.
Putting batteries in series causes the voltages to be additive but not the mAh capacity. Putting cells in parallel causes the mAh capacity to be additive but not the voltage. Then you need to look at the "C" rating of the battery. For a 4200 mAh battery, a "1C" rating would mean that the battery pack can deliver its full load of power in the space of one hour. The Thunderpower ProLite 4200's are rated at 15C Continuous draw, 24C Burst. So this battery pack can deliver its full charge to the motor in as little as 1/15 of an hour or 4 minutes.
You can build a pack any way you want. You can have the simplest battery : "1cell", (say 2100 mAh) so just one cell producing 3.7 volts at whatever mAh capacity rating you want or you could have a "4S" pack consisting of just 4 cells in series (14.8 volts) at whatever capacity you want (say 2100 mAh). If you needed more capacity but not more voltage you could have a "4S2P 4200 mAh" battery which is still just 14.8 volts but is built of 4 cells in series and two of those in parallel. (total of 8 cells)
If you want to turn a motor at higher rpm then you will need more voltage. If you want to turn the prop at this same rpm for a longer period of time then you will need more mAh capacity. Electric motors are VERY good at generating torque. Larger propellers with higher pitches are much more efficient than low pitch props. You should consider this when you are trying to decide on what combination will do the job.
Your decision as to the number of watts that you need to fly your plane and the selection of a prop, rpm range, motor and ESC will pretty much determine which battery you will need.
Watts = Volts x Amps - this is the expression of the power produced by the system.
Volts = Watts / Amps - this represents the ability of the motor to turn a specific number of rpm’s with a given prop
Amps = Watts / Volts – this represents the ability of the motor to turn that same rpm for a given period of time.
Notice that volts and amps are interchangeable in the equation but your ESC and motor must be able to handle the resulting current draw and heat of your particular choice.
We do not want to be carrying around a lot of excess battery capacity (which translates directly to battery weight) at the end of the flight (best practice would be to use no more than 80% of the battery’s capacity) so you will need to look for batteries that have a high enough “C” rating to permit the battery to yield the required number of amps in the approximately 6minutes and thirty seconds that we need to run.
If you intend to use “Governor Mode” (and I STRONGLY suggest that you do) you need to have enough headroom (reserve capacity) in the battery to maintain the chosen rpm for the entire flight. If you run an electric motor at full speed it is irrelevant if you have also selected governor mode as the ESC will fall out of governor mode almost instantly. So how does that affect us? You will need to choose an rpm that does not drain the battery voltage or amperage below the level required to maintain the chosen rpm for the duration of the flight. This is an area where testing will greatly accelerate your progress.
Timer / Processor
As noted above, the ESC requires a signal to initiate any of its functions. The currently available timers provide very basic functionality. If your goal is just to get your motor running on the test stand or fly casually these may be sufficient. Having flown and competed with an electric powered plane for the last year I can safely say that to be competitive at the Open / Expert level you will need something with more capability. You will want to be able to set your rpm at a specific value without having to guess how much you are changing the setting. You will also want to be able to be able to safely hook up your battery and start or stop your motor. You will want the person helping you or members of the public to be safe when handling your model. You will want to be able to control how long your motor runs. You will want to be able to control all of these without having to resort to removing chips or hooking up a computer to reprogram the firmware.
A purpose built timer / processor with full functionality including “Electronic Needle Valve” and “Magic Wand” safety interlock with the ability to make use of the Schulze dedicated 18.46k F2B ESC should be available this fall. The prototype will be flying in Paul Walker’s plane at the 2007 FAI F2B Team Trials in September.
Laying out the plane and power train
The largest problem you will be faced with is to locate the battery, motor, ESC and timer / processor in such a manner that you can achieve a proper C of G, provide easy access to the battery for installation and removal and provide access to the timer / processor should you want to change the rpm or flight duration. This is predominantly why I do not advise trying to modify an existing design as even if you manage to find a good location for everything you will inevitably find that the tail moment is too short to balance the package out. It is also why I suggest that you purchase the complete power train prior to drawing a single line on the plan or starting a modification to an existing airframe.
It would be helpful to take a plane of similar size to the one you are going to build, take everything off of it and start placing your new components on the plane (use elastic bands or masking tape). You need to pay particular attention to the distance between the ESC and the battery as it can not be very long. Remember that you can locate the timer remotely (say behind the wing) to act as balance for the battery and motor) Once you have it all layed out and balanced, you can begin to draw the plane.
You may find as you draw the plane that some things want to occupy the same space. It is likely that the battery and the bellcrank will want to be in VERY close proximity to each other. We machined a new bellcrank specifically for “SHOCKWAVE “ to accommodate the 5S2P battery weight.
Ground clearance may be an issue as you could be swinging a fairly large prop. (as much as 14")
Cooling of a large motor and battery is major issue. You can puff a large lipo so easily it will make your head spin and your pocket book scream. You can always cover up some of your cooling vents later. For starters, keep the battery, motor and ESC COOL. The magnets in an out runner are only held in place by epoxy. Too much heat and you will have little magnets flying everywhere. They need LOTS of air flow. Remember that each 5S2P battery costs $220.00 U.S. (when I started they were $300.00 each!) and you will need at least four and more likely six or eight. Anyone for ten?
You will likely end up with a plane that weighs at least 66 ounces and possibly more. “SHOCKWAVE” weighs 70.5 ounces and the wing area and airfoil thickness were based on a plane weighing 63 ounces dry with 7 ounces of fuel. You need to find planes that have successfully competed at this weight and use them as a reference for your own.
The structure of the fuselage will be quite different from your normal design. You do not need any plywood doublers or extra stiffening (much less if you use molded compound curved surfaces) and since the motor will in all likelihood be mounted from the rear, a flat vertical front end surface and a cowl are more appropriate than the regular type of cowl and motor mount system we traditionally employ. The firewall does need to be able to withstand the large torque loads which can be generated by larger props and the application of rapid start sequences or electronic braking. I suggest a laminate of a ¼” balsa core with pine crush-blocks (to accept the blind nuts from the motor mount) let into the core and faced with 1/32th inch plywood back and front all held together with epoxy. This will provide a lightweight former and sufficient gluing surface to ensure proper distribution of motor forces to the fuselage shell.
If you are going to run any electrical wires through the fuselage you will want to provide for these with a hole in each of the formers behind the motor compartment. Don’t forget to allow for a sufficient diameter to permit electrical connectors to go through the hole as well.
How to contain / mount the battery
The battery is heavy. If it ever got loose inside your model it would likely destroy it. You must provide a secure means to mount the battery to the plane while ensuring that the battery is properly cooled and remains easy to install and remove. Most batteries are covered in a tough plastic shrink wrap. You can install Velcro onto this plastic to help secure your battery and / or you can use Velcro straps to provide even more security. Utilize existing fixation points such as landing gear mounting points and the top of the landing gear itself as battery restraints. (assuming fuselage mounted gear)
The wires come off the battery at a fixed point that is not always convenient to where you need them to be. It can also affect the position you mount your battery in.
Ideally, you want to mount your battery such that it is directly centered on the vertical CG and as close as possible to the longitudinal CG. Leave some room for fore and aft mounting as you will not want to be adding even more weight to balance your model. You will find that your choices in battery shape, dimension and weight are somewhat limited.
The problem with mounting the battery is that it does not have any handles or holes to put a screw through or convenient tabs to use to hold it in place. My solution to this is to mount the battery to a plate that has tabs on both edges using both Velcro on the battery and plate as well as two Velcro straps to mount it securely. The plate then provides a consistent interface to the plane’s structure. The plate and battery are slid down to the mounting floor and then back to engage the tabs in slots cut in the motor compartment formers. A single small bolt stops the plate from moving fore and aft. C of G changes are accomplished by mounting the battery further forward or aft on the mounting plate. Since the battery weighs 16 ounces, I can access the entire range of CG positions without the need of any excess weight for balance.
Cooling the battery
Cooling the battery is very important. The battery in SHOCKWAVE “floats” in a sea of cooling air on all sides (with the exception of one narrow edge for mounting). Air is free to flow from the bottom of the battery compartment to the top and vice versa depending on which way the nose is moving. The compartment also receives a healthy dose of ram-air from the scoop on the cowl. My battery never reaches a temperature of more than about 42 degrees C (108 degrees F). We were able to maintain these temperatures even in the 43 degree C daily temps in Spain. Now having just told you that you must cool the battery I must also tell you that it needs to be reasonably warm to operate at its optimum efficiency. So keep it cool, but do not build a refrigerator for it. Actual operating temperatures for a battery pack will depend on the current draw and the “C” rating of the battery to deal with that demand.
Mounting the ESC and Timer / Processor
The ESC can get quite hot and should have access to a good supply of cooling air. In my case the ESC is soft mounted via Velcro to a small floor directly below the battery and the heat sink is exposed to the air through an aperture on the bottom of the fuselage. You also need to be conscious of the distance between the ESC and the battery. With the Schulze 18.46k F2B ESC the total run of connecting wire from the battery to the ESC and then to the motor should not exceed nine inches. Check your ESC’s mounting instructions.
The Timer / Processor can be mounted anywhere as the hookup for this is thin, light low voltage wire. If you have an external switch or electronic needle valve, consider mounting it behind the wing to provide some additional balance weight and to provide a safe place to make changes to your rpm from.
Safety is a key word with electrics. First, large lipos pack a lot of electrical punch. You should take all precautions to ensure that they can not be shorted. You must be careful not to drop them as they are not protected by a rigid outer shell. You need to design a system which protects both you and anyone helping you at the field from injury. You need some sort of failsafe method of rendering the motor dead.
Charging a large lipo requires high quality lipo specific chargers and balancers. I HIGHLY recommend that you use both Thunder Power batteries and the Thunder Power 1010C charger and 210V balancer. Pound for pound these are the best batteries and charging equipment on the planet.
I freely admit that I am sponsored by Thunder Power but it in no way influences which batteries I use. If they were not the best I would not be using them.
A final note on charging: Lipos, if handled properly in a respectful manner will provide hundreds of safe trouble free use cycles. Handled or charged improperly they can become LETHAL. Prior to ever charging a lipo for the first time, you would be well advised to seek out the facts on improper use or charging. Lipos can explode causing fires. A good source of information about lipo batteries is the manufacturer’s data and safety sheet as well as the Batteries and Chargers forum on RCGroups. http://www.rcgroups.com/batteries-and-chargers-129/
Testing and Monitoring
If you want to know what is happening to your setup you will need some monitoring equipment. The minimum I recommend is a Hyperion E-Meter, an IR temperature gauge and a Digital multimeter such as the Fluke 81438. You can't see electricity; you have to test for it.
Large electric planes are much more efficient in power to weight ratio than small electric planes. It is no accident that “SHOCKWAVE “ is the equivalent of a .90 size stunter. That is (IMHO) the sweet spot in the marriage of all the above factors.
SHOCKWAVE : Some numbers
The empty weight of my plane finished without any power train components is 38oz. (1075gm) The weight with every thing except battery is 54.3oz.(1538gm). Total weight is 70.5 ounces.(1995gm) SHOCKWAVE is a full take-apart (two piece wing, stab and elevator, rudder and tail cone, cowl and engine) model. The fuselage is of molded balsa construction and all flying surfaces are built up and fully tapered (including the trailing edges).
Target weight was based on a 63oz plane with 7oz of fuel. The Plettenburg Orbit 30-12 motor weighs 10.7oz (305gm), the Schulze 18.46k F2B ESC weighs 1.5oz.(45gm) The ThunderPower 5S2P 4200 Pro Lite battery weighs 16.7oz.(473gm) The AeroLectric timer-processor board weighs .75oz.(20gm)
I hope this has been of some help in your efforts to embrace electric stunt.