Model Helicopter Governors and the Futaba GV-1
There are two types of governors, one you elect and one you buy….sometimes both if you are rich! Here of course, we will be concerned with the latter. A governor is nothing more than a speed controller. There are two forms of governors. One can be used to control maximum rpm (overspeed governor) and is a protection device. The other type can be used to maintain a constant rotational speed of a shaft. The “constant speed governor” will be the topic of discussion in this article. Since I have used several types in my models, it will be my pleasure to pass some of these experiences along to you.
One does not need a governor to successfully fly a model helicopter in a very competent and professional manner. Then again the same can be said of the top of the line radios and many upgrades. Where the governor shines is in the reduced setup time, since pitch and throttle curves along with the many switchable flight modes can be put aside if desired. These are no longer the major issue in order to achieve a fairly constant rotor rpm for most situations. The bottom line is that a governor will not make you a better pilot if you already know how to properly setup your helicopter. Now that the honest stuff is out in front, I would like you to know that I own a few governors and have had quite a lot of fun with them. The technology seems to be advancing with a corresponding increase in performance (features) and accuracy.
The term governor may be new to some and is synonymous with constant speed controller with respect to our models. Don’t confuse this with a speed controller for electric helicopters. In our normal day to day lives these non model heli devices are to be found on electrical generators large and small, full scale airplane propellers, turbine helicopter engines, diesel engines or the infamous lawn mower to name a few.
All governors basically work the same. They must have a means to sense changing rotational speeds. It might be in the form of cooling shroud air pressure, flyweights, electronic pickups etc. As a simple example, this can be accomplished through spinning flyweights opposing a spring tension which holds the throttle open. At the governed speed the flyweights will overcome the spring tension through the now higher centrifugal force and move the throttle towards a closed position. Should rpm drop so will the flyweight force, with the now dominant spring opening the throttle to maintain the governed speed. The governed speed can be adjusted most easily by varying the spring tension. From this governed picture a tug of war can be envisioned between these two opposing parameters. Here lies the difficulty since this can be transmitted into governor hunting or torsional oscillations. “Hunting” is not a good situation in a model helicopter since it may be further complicated by the additional requirements of the anti torque (tail) rotor through gyro stabilization. There are ways to get around this, the cheapest of which is called “dead band”. This is a term used to describe the rpm band width where the low rpm undershoot correction begins and the high overshoot condition is checked. All governors have this in various degrees.
Every engine has a different throttle response time. It takes time for the inlet gasses to accelerate and increase airflow into the engine. Engine port timings, compression ratios, tuned pipe setups and even the carb design along with mixture settings play a role here. Rotor inertia will surely have an effect since a “heavy” rotor system will act as an rpm accumulator or governor damper with respect to small outside influences. These are but a few factors. The point I am trying to get across here is that this is not like turning a light switch on or off, and that each installation will be different. Keep this in mind for later on in the article.
Model helicopter governors are much more sophisticated than the above flyweight example. Rpm sensing methods take the form of rotating magnets or optical pickups. There are different locations to sense rpm. I have seen magnets mounted to main gears, start shafts and engine fans.
Lets take the case of the main gear first. Our sampling rate is low due to the slower shaft speed. Ask yourself this, what is going to happen if the main gear strips in a crash? Meltdown! If your clutch has any amount of slippage matters can quickly become confusing to the governor. The lesser evil might be to mount the magnets to the start shaft, but all helicopters don’t necessarily have one. If we were to mount it to the clutch bell, any clutch slippage would cause things to get pretty “dicey”.
Now if I were to pick the best sensing method it would be to directly monitor engine speed. There is two other units I know of one of which uses a production optical device mounted to sense the engine fan speed rather than a magnetic sensor. The optical method offers the best sampling rate due to its higher resolution alternate pie sliced pattern, however at what point this becomes overkill I know not. But we are going to find out!
Futaba, a company with substantial resources has recently released a new governor called the GV-1. Like all model helicopter governors, clever software is buried deep within the electronics which surely jumps through many “hoops” like a well trained dog. Too many hoops can certainly screw up the dog. Simply stated a governor does not know if it is right side up and climbing with a large positive collective increase or falling upside down with a large positive pitch increase. One is aided by gravity while the other is opposed. So the throttle response time has changed due to changing outside influences….another reason for dead band or lag time. The governor must not get ahead of itself and chase its tail. We don’t want the governor to respond to every small gust of wind or power setting change and induce torsional oscillations, especially in a hover. So by having a dead band, small changes are ignored in favour of a steady engine torque. Certain situations can however be anticipated or predicted, and provisions designed into the software for proper corrections which help to narrow the dead band under these specific conditions. One might conclude that additional information as to how fast the collective is being moved and where it is at and being moved to, might through smart programming yield a benefit. The means is there in the receiver whether it be the collective or throttle channel. We could actually move the lower dead band point up with collective pitch increase so as to reduce rotor droop. The reverse would be true in the case of lowering the collective, the upper point of over shoot correction would be reduced thereby preventing overspeed. We would be moving the whole dead band temporarily as needed for the appearance of a smaller dead band width. We have effectively shifted this band as is normally done in many full size helicopter governors, which will allow for a further increase in overall accuracy. Think of it as a governor assisted acceleration/deceleration mix based on rate of change which is referenced to collective stick position. So now, even after a governor is performing well, modelers still have to be protected from themselves by adding electronic safety features. There are times when we do not want inadvertent governor engagement. I shall leave that one to your imaginations! A lot of time goes into designing, but often it is too easy to pick the thing apart while not fully understanding the hurdles involved. While I am no expert on electronics and software design, I do have a healthy respect as to what is presently available in our hobby. I handle governors in my line of work and have a solid appreciation of what many common place add ons (sometimes design after thoughts at that) do to correct for a governors shortcomings. The response time of model helicopter power train system is not easy to predict as compared to a full scale machine considering all the random (radical) and varying parameters encountered.
Hopefully now that I have you thinking in different “theoretical avenues”, the design of the GV-1 will make some sense. This unit receives its rpm information from a magnet mounted to the engine driven fan which is mechanically direct coupled to the engine under all conditions. Engine sampling rate is 1:1 since one magnet can be used, while the second is meant for fan balancing purposes should you desire. One magnet must be reverse mounted to the other since the system is polarised. The mounting of the magnetic pickup system will vary from ship to ship with mounting brackets supplied. The system offers enough flexibility so that an alternate mounting position on the starter shaft is certainly possible although not mentioned in the manual.
Now to have a look at the little jewel. The GV-1 has lots of little gizmos in the way of features which allow you to tailor install according to your equipment or personal preferences. Considering what a governor is supposed to do may cause it to appear as a complicated device, in so far as all the bells and whistles are concerned. This is not to put you off but only a warning that if you do not fully understand the systems “software” setup, expect difficulties and possibly hazardous operation. I had no problem understanding the manual for the most part, but someone less experienced with how and why the throttle servo position is used for safety features along with the various transmitter switching modes, might have trouble if in a hurry. Don’t be. The programming method is similar to that of a computer radio using two function and two data buttons. These buttons are small and a nonconductive tool (plastic screwdriver) is supplied for convenience. The unit itself has a liquid crystal display for viewing the setup menu features.
The device has several inputs. The throttle, two aux channels and the signal from the magnetic pickup feed into the unit. You do not have to use the aux channels but I’d recommend using at least one for turning the unit on and off rather than relying on the low collective stick to do so. In other words it will function fine with a five channel radio but with reduced features. Another option is to use the idle up switch to prevent the throttle from reaching the governor cutoff point at low stick. One does not want the governor to kick out while performing an inverted hover! This alternate method of setup was not mentioned in the manual. The use of a second aux channel input will allow the feature of remotely changing the governor speed setting if the first channel is only a two position switch used to turn the governor on and off. Up to three speed settings can be stored with the proper radio transmitter/ receiver. The unit has a mixture control but when used it reduces the speed settings to two. Throttle hold will override the governor off if the throttle hold set point is lower than the governor on point.
On the output side of the GV-1 is the socket for the throttle servo. An additional mixture output socket is available should you need it to correct mixture related governor problems which would plug into a mixture servo connected to the carb. The unit has its own nine point mixture curve.
The system is light, small and consumes very little battery energy (40ma). This is similar to an idling servo with no load. While the magnetic sampling rate is lower than the optical pickup method used on another governor the GV-1 has a much better response time. The specs claim a dramatic improvement in the locking bandwidth over past units. This is not surprising considering the custom chip and tiny SMT construction inside the GV-1. As is common with model governors should a system component fail, as would be the case of a failed pickup or dislodged magnet, the normal throttle signal will be allowed to pass through to the servo.
The display will show rpms both set and monitored at approximately 95% of top stick in either engine or rotor speed plus it will allow viewing of your selected gear ratio which is part of the setup process. Lots of other goodies can be viewed which aids in the setup process. One such feature is the low voltage function (battery fail safe), which when enabled can bump the throttle down when at 3.8 volts and below. The bump down range (throttle opening) is user selective/adjustable, it can also be reset for a 30 second normal operational period from the transmitter. Each low stick reset will give 30 seconds to land with normal governed throttle operation. Personally, I’d turn the governor off and land. The unit also has a voltage display function. Even though by now you think this setup process might be pretty time consuming, it is not and really quite simple after understanding the terms in the fifty page manual. I have seen VCR documentation just as aggravating!
Rather than chatter about the governor, I am going to take you briefly through the installation on a gasser. Unmentioned was gasser compatibility but the manual also didn’t say incompatibility! The only installation problem was to figure out how to mount the magnetic sensor. I was going to use the start shaft (and I may yet) but thought better considering that the damped clutch might screw up the GV-1. So the decision was made to use the fan by fabricating a bracket to secure the pickup to the engine and fan chamber metal floor. A hole was drilled in the fan chamber floor so that the pickup would fit. The angle shaped bracket is riveted to a plate which was secured to the engine crankcase. Screws could be substituted easily enough for the rivets. Since the fixed wing version of the newer G-23 has the high tension coil mounted to the engine and the heli version doesn’t I decided to use these hard points for the sensor mount attachment. The job took some time since we tried different methods of securing the pickup. The above mounting is very rigid and will maintain the set gap adjustment between the sensor and the fan. Because the gasser has axial crankshaft chuck this gap might change in operation. What this means is that the sensitivity (gap setting) should be checked at both extremes of chuck. The GV-1 has a read out for checking the proper flux strength. The max strength is 97% which I managed to get under both conditions after some tedious work. It will however work as low as 60%. When installing the magnet we were careful and drilled the proper hole directly under a plastic fan impeller fin for the lowest possible flex point. There have been reports of magnets being thrown and I really don’t want to pull the engine again. The magnet was secured with aluminum epoxy which by its nature of being nonferrous will not interfere with governor operation. We carefully sized to the hole for a .003″ pinch fit on the tiny magnet. We also used laquer thinners to clean the bonding surfaces. I shall be very upset if it ever comes loose.
With everything mounted and plugged in, the final GV-1 setup parameters were entered into the units nonvolatile memory. Gear ratios, throttle end points, switching options and rpm settings were quickly and efficiently programmed into the unit. Read the book and fully understand what you are doing before you do it thereby avoiding any surprises, before running up the machine. We were using a Miniature Aircraft tachometer for inflight pitch adjustments. The purpose of the “topping check” is to make sure the rotor speed is not dragged down due to lack of engine power through too much top end pitch. If the rotor droops at max stick the collective top end pitch must be reduced in order to have constant rotor speed. Alternately we want to extract all the power from the engine by having enough pitch. What it all boils down to is finding where the rotor droop occurs and then backing off the top pitch a bit so as to just remove it.
Well flight time finally came and with the governor set internally to 1500 rotor rpm, off we went. The below tests were of the unit’s governor section only without using the overspeed cutout feature. We engaged the unit and tached the machine using a Miniature Aircraft tachometer to 1530rpm both at low stick and light on the skids. It stayed within 20 rpm. Next we picked the machine up into a hover and the rpm increased to 1610 and was very steady. Translating into forward flight was smooth and the rpm stayed at basically 1500 rpm at all power settings. Good and steady. A normal approach was made and the rpm increased to 1680 rpm then dropped to 1600 in the hover. We tried linear and nonlinear throttle curves, adjusted the mixture slightly and always ended up with the same figures. Next we set the throttle curve to the pre- governor values and flew with the governor off. Hover was 1500rpm, forward flight varied up to 1550 and the normal descent was at 1540 rpm. The governor deadband is about 140 rpm wide under the worst temporary condition. The amount of time the governor was operating in the off speed area should be a consideration since we don’t spend all our time descending. The area that matters to me is the difference from the hover mode to that of forward flight which is just over 100rpm.. There was absolutely no indications of governor hunting which made me very happy. Rapid power increases from the hover to max stick followed by a quick reduction in power yielded a very clean governor response with no hunting or over shoot. This is the best model helicopter governor I have used so far, but accurate pitch/throttle curves can result in more constant rotor rpms. Just to be sure of our results we mechanically changed the throttle geometry to be absolutely linear…..90 degree servo/carb arms with half stick at 100% unmodified servo travel. Off we went for some more tests. The same overspeed condition occurred during normal descents, but now the rotor did not increase when entering the hovering mode.
The biggest difficulty for the governor was on approach and if this overspeed condition was seen as an irritant one could simply turn the governor off at this point. This condition was mentioned briefly in the manual. The beauty of this device is in the fact that it will save time by not having to set the pitch/throttle curves perfectly both mechanically and electronically. Adjustments to the top pitch will still have to be made in the case of large atmospheric changes in order to get the most from your engine. There are some benefits over my non governed machine. The rpm is more stable in the hover and up top using the GV-1. My mechanical gyro can be set higher without wagging the tail due to smoother engine torque.
With the glow engine the descent overspeeding issue will be less. The glow engine hovers at half stick with a much larger throttle opening than the gas. The gas engine also has more torque. With the two stroke engine governor design the engineers had to avoid the region where inherent engine instability occurs at lower throttle openings by not allowing the throttle to operate this area all the time. This restriction is why the deadband is so much wider under this condition. Simply modifying (lowering) this one condition and calling it a gas version would surely benefit the G-23 engine.
While my ungoverned helicopter out performed the GV-1 in some regards, I have seen many models around that this unit should be fitted to. I see numerous beginner to intermediate flyers fumbling along with radio adjustments, to which this unit might be a “god send”. For the 3-D types I think this unit will be very popular. For the person well versed in pitch and throttle curves, the GV-1 governor becomes a very interesting and flexible toy, second in priority to the purchase of a piezo gyro. I’ll be pleased if you’ve enjoyed this and hopefully I have given you an insight into model helicopter governors.