SAFETY ISSUES
I have been very reluctant to add this page since it could be construed as being negative comments towards CHR and/or the Safari design. However, I feel that I can no longer stay quiet about these kinds of things. The Owner's Area of CHR's website states that they will update their Service Reports when they receive three notices about a problem. Firstly, I believe that only one incident on an important component should trigger documentation that is available to all owners. Secondly, I am aware of first-hand reports of three incidents on a specific supplied component plus three second-hand reports on the same component in other ships and it still has not triggered any documentation on CHR's website or notices to builders and owners. I'm also aware of another important component with more than three incidents and no documentation. In my opinion this is totally unacceptable, especially when it can lead to an emergency situation in flight. The fact that CHR has only ever issued one Service Bulletin that I'm aware of (buried inside an accident report), speaks volumes about their attitude and communication.
I would sincerely urge all builders and owners to contact me if they believe they have found a safety issue with the Safari. I will endeavour to add their comments to this area and if they wish to remain anonymous in public then I have absolutely no problem with that. The most important issue is to make this information available to *ALL* builders and owners in a timely manner. Even if CHR or a builder is aware of a problem, this does absolutely nothing to help other owners unless the information is shared and made public. The objective should be safety, rather than trying to sweep problems under the carpet.
Since various people have failed to document and make this type of information public, I make absolutely no guarantees that this list is complete ... in fact I'm sure it's not. These are simply some of the issues that I am aware of and I will try to update it when new information becomes available to me.
Disclaimer :
The author is not an AME, A&P or any other alphabet soup related to the aviation industry. Anything posted on this website must be considered an unqualified opinion and evaluated by the reader so they can draw their own conclusions or consult with an appropriately rated professional.
- COLLECTIVE TRIM SPRING SYSTEM
- FUEL TANK BONDING & REFUELING
- EARLY DIGITAL ROTOR TACHS
- ANALOG ROTOR TACHS
- ALARM HORN MODULE
- ALARM HORNS
- DIGITAL TACH "FAILURES"
- ALTERNATORS
- PUSH-TO-TEST INDICATORS
- P&B W31 CIRCUIT BREAKER / SWITCHES
- BREAKER SIZE
- STROBE LIGHT
- SOLENOID BOARD
- STABILIZER RIVETS
- TAIL ROTOR GEARBOX
- STARTER SWITCH WIRE ROUTING
- CONTROL RODS
- FUEL TANK STRAPS
- TAIL ROTOR / BOOM CONTACT
- TAIL ROTOR SPINDLE
- FUEL TANKS
- ROUNDED TIP MAIN ROTOR BLADES
- CHT GAUGE
- TAIL ROTOR BALANCING
- INITIAL STARTUP
- COUNTERFEIT AN FITTINGS
- PYLON SAFETY WIRE
- VENTED FUEL CAPS
- MA4-5 CARBURETOR MIXTURE CONTROL
- EXHAUST SYSTEM CRACKING
- MAIN ROTOR YOKE CAPS
- NEW "PANCAKE" OIL TANKS
- OIL COOLER CRACKING
- FLOOR COVERING & SUNLIGHT
- ALTERNATOR CONTROL WIRING
- DUAL ELECTRONIC IGNITION SYSTEMS
- SLICK MAGNETOS
- OVER-TORQUED BOLTS
- STAR COUPLER SETSCREW
- CONTROL STOPS
- CLUTCH SPRINGS
- SHAFT RELIEF RADIUS
- MAST TUBE BOLTS
- FUEL TANKS - CURTIS DRAINS
- OMEGA COUPLER BOLTS
- HEAD ANGLE PLATES
- CONTACTOR (SOLENOID) JUMPER
- STAR COUPLER CROSS-BOLT
- CARBURETOR FLOATS
In my opinion, the supplied spring system is both unsafe and unacceptable for flight. As a direct result, I refuse to even consider flying any ship using this system. It would appear that the supplied spring is being used in a way for which it is not recommended by the distributor and it is being loaded to three or four times its design rating. The very predictable result is that the spring breaks. I have talked to three builders who have had this happen and am aware of second-hand reports of three more breaks.
If the spring breaks, there will be an immediate upwards force on the collective's grip end of perhaps 10-18 pounds. Unless the pilot contains this upwards force/movement, it will lead to an over-pitch situation and a resulting loss of Rotor RPM. Even if this collective motion has been caught in time and arrested, the pilot must maintain a large steady downwards force on the collective while trying to maintain collective authority plus throttle control and also while simultaneously setting up for and performing a landing. Those readers with helicopter training can easily envision times where this would be a critical situation. Also, I tried simulating and resisting this kind of pressure for less than a minute and my arm was already getting sore and strained.
For liability reasons I will not recommend an alternative but I have developed an alternate solution which I'm comfortable with. Unfortunately, to my knowledge CHR has still not released an upgrade more than two years after they acknowledged a problem.
Update: I have now seen a newer ship that has the single cot spring replaced with two more robust springs. These springs are routed the same as the original spring / cable with springs on opposing ends and using two cables. The cable for each spring passes through the center of the opposing spring. I have not seen a Service Bulletin or any other documentation on this change.
As shown on the construction prints, the fuel tanks are mounted on a rubber pad and connected using MIL-DTL 6000 hoses. Thus the tanks are electrically isolated from the frame (as measured on a factory assembled and delivered Safari) and this goes against all aviation standards and practices. If just the frame was grounded during refueling, this would not provide a reliable ground path for the fuel tanks which is where one really requires it. I believe the simple solution is to run a bonding wire from each tank to a stud on the frame and the continuity of this path should also be verified with an ohmmeter. This should not be as big an issue for those builders that have chosen to use properly fabricated steel braided lines.
A helicopter can generate a LARGE static charge during flight and it can remain charged if it lands on something like a trailer that is insulated. Therefore it is important to verify that the craft has been properly grounded before refueling.
There is also the issue of fuel containers (i.e. jugs, cans, etc.). From what I remember, static electricity is generated by the flowing of the fuel,especially against "plastic", and that is why most refueling hoses have wire conductors and metal nozzles. Although some of the high-quality "plastic" fuel containers contain a special conductive component, I would assume that most of the cheaper ones do not have this feature. Thus just the filling and emptying of the containers poses a potential hazard. I believe that metal cans help to address this issue, especially if they are connected by a ground strap. Physically placing them on the ground before a transfer also helps to equalize any residual charge ... while filling any fuel container, it should be placed on the ground and NOT on an insulated surface such as the bed of a truck. Also note that this static issue is why there's often a warning to keep a metal filler spout in constant contact with the tank/can.
I found that Canadian
Tire carries the Briggs
& Stratton WM521 steel Jerry cans for ~$40 Cdn including
the "V" spout.
Update: It looks like Canadian Tire no longer carries these fuel
cans but it is possible that some stores may have some old stock.
I'm glad that I purchased mine when I saw them in stock locally.
One of the big concerns with aircraft fuel is contamination (particle or water) and that's why we check the gascolator. When refueling from cans or drums, I've heard of lots of "tricks" that MAY help to reduce any water content but some are dangerous and of dubious merit. For the small cost, I think it is best to get a special purpose funnel that has a filter, a proven water separator and is built from electrically conductive material. Such a device is available from Mr. Funnel.
The digital rotor tachs supplied by CHR have a fixed high and low alarm point and the first batch(s) of these had the wrong value set for the high alarm trip-point. This could lead to an overspeed of the rotor system without any alarm being given. I've heard of owners using a "chirping" alarm to maintain their throttle setting at the top of the green and this could be very dangerous with an alarm setpoint that is too high.
I would recommend that all owners of the digital tachs either test their high alarm against a known calibration standard or consult the factory to determine if they have one of the affected units.
There have been numerous reports about the inaccuracy and reliability of this tach. I specifically asked one owner to compare his tach to his high alarm horn. In the morning it registered 510 RRPM while that same afternoon it registered 495 RRPM. That's a 3% drift within a couple of hours based on just two samples.
Based on the reports that I've received, I do not feel that this is a reliable and accurate gauge to be trusted. Although I do have one of these tachs, I've chosen to have it upgraded to digital electronics since I don't believe in installing a gauge that I know I can't trust.
There are two ways to connect the CHR-supplied alarm horn module; specifically a P-lead input version and a Hall-effect sensor version.
To my way of thinking, a P-lead connected version has a major shortcoming in that it measures Engine RPM and not Rotor RPM. The specific purpose of this alarm module is to warn of high/low rotor speed conditions. In the event of a slipping clutch, it is possible to get a high alarm due to an engine overspeed, or no alarm at all, even though the rotor speed could be low. Conversely, during an autorotation the low alarm horn will go off regardless of Rotor RPM. In the event of an engine failure, there is no useful information from this module, but the low alarm will be triggered and distracting the pilot.
I have heard second-hand reports that there has been some problems with the Hall-effect sensor version of this module. Specifically, they relate to noise pickup by the sensor and a lack of reliability. This needs to be very carefully checked, especially if one is using the horns as backup for an analog tach that they don't trust. Specifically, one should be able to use an internally amplified sensor and I would choose to use shielded wire for the sensor input rather than the non-shielded wire supplied with the harness.
Before considering using this module, I think its important that a pilot fully understand its limitations. There are some wiring techniques that will improve its usefulness but I strongly believe this module gives false and misleading information if wired into a P-lead.
There were no specifications supplied with my unit to state the specific conditions and RPMs where this unit will trigger. My testing of the module I received indicates that the setpoints were set at 460 RRPM (low) and 502 RRPM (high) with a 2 RRPM deadband.
I have also heard reports that the alarm horn sounder(s) are not always audible, especially in a running ship with headsets on. While these sounders may be audible during installation, that's not when they're required. I think it is critical that builders test these sounders under actual conditions before they even start to consider trusting them. As a minimum, the startup test for the sprag clutch and needle split should give some indication as to their viability. Interestingly, the R22 is not considered airworthy if the horn system is not operational.
I think there may be two issues with these sounders:
1) Physical location. If they're buried someplace like in the
console or under the seat then they will definitely be muted.
Perhaps a better location is behind the headboard.
2) The units may simply not be loud enough. Only testing can
verify whether a specific sounder has sufficient volume for a
particular installation and there are a lot of different modules
available.
Note that if one has ANR headsets, then the testing should be done with the ANR system activated.
As an alternative (or supplement) to using sounder modules, tones can be directly routed to the audio system.
I am aware of several reports of digital tach "failures". In the one case where I know the details, this was the direct result of an alternator failure which also took out the radio. I have first and second-hand reports of three "failures" by CHR, but no specific details. I have been in direct contact with the original developer of the electronics for this tach and he has stated that he has *NEVER* received a "failed" CHR digital tach for repair. We discussed the electrical protection built into this unit and it does have reasonable electrical protection within it, including a fuse and several other related components. Perhaps someone is replacing these fuses without asking the developer why they may be blowing. How is he supposed to analyze / fix the issue unless someone informs him of possible problems and the details?
The CHR recommendation appears to be to install both the digital tach and CHR's alarm horn module as a backup indicator of RRPM. While this may be a short-term solution, I think it would be more adviseable to find the root cause of the issue. Note the previous discussion of the alarm horn module's limitations and also note that this dual warning system installation can result in two horns going off simultaeously which I think would be very distracting.
Our hypothesis is that these "failures" have more to do with the craft's electrical system rather than a fault within the tach. In the CHR supplied wiring diagram, there is/are no master fuse(s) shown and it is unclear if there is any form of overvoltage protection ... certainly none is obvious. The electrical system has to cope with various things such as low batteries, huge load dumps when the starter is switched off, transients, etc. etc. While simple systems may work in some situations, there is a good reason why electrical engineers add various forms of protection and isolation.
This is not the place for an electrical course, but I do think it is important for the builder to carefully analyze their electrical system. If they don't believe they're qualified, then it should be relatively easy to find someone to assist them. As part of this analysis, the issue of electrical failures should be carefully considered as should the use of alternate or backup power systems since these tachs will not work without a source of electricity.
Update: I have again been in contact with the developer of these tachs and it would appear that the newest versions of these tachs now have their input protection circuitry modified, including the use of a self-resetting fuse. The previous versions could blow their internal fuse with transients above 21.4 volts on the input power supply. While I commend the developer for these enhancements, I also think this just reinforces my belief that any builder who has experienced digital tach "failures" should carefully investigate their electrical system for anomolies.
Builder's have a wide variety of alternators to choose from, including aviation certified externally regulated alternators through to various internally regulated alternators of questionable origin. My CHR-supplied solenoid board contained a voltage regulator module without any markings or documentation and thus I have no knowledge of its specifications or limitations and am unwilling to use it.
Internally regulated alternators are known to have the
possibility of either shorting or "running away". If
one uses the basic CHR supplied wiring diagram with these and
just eliminates the voltage regulator then there are two
potential failure scenarios:
1) A short to ground will burn out the shunt and all electrical
devices will shut down
2) An over-voltage will probably also burn out the shunt but will
also feed an unkown high voltage to ALL electrical devices that
are currently turned on. The cost of the smoke will probably be
high.
I've chosen to use an internally regulated alternator, but I've also chosen to add ANL fusing and to provide independant over-voltage protection circuitry.
These are a neat way of providing an indicator lamp, a test circuit and a dimming option all rolled into one simple unit. However, there are a couple of potential problems with these:
1) If they're turned to full dim then they won't be seen in daylight. I intend to add this to my daily inspection to verify that they're turned full bright. Also, they should only be turned to dim when flying at night and the pilot is acknowledging their warning.
2) Panel location. I've seen pictures of several consoles where these indicators are mounted very low and/or out of the normal scan area. Good ergonomics would place these in an obvious place within the normal scan area.
3) The bulbs are not very bright in daylight and it is possible that they won't be noticed. One solution is to replace the bulbs with flashing LED bulbs in order to draw more attention. For those that don't know how to (or don't want to) make these replacement bulbs, they're available from VK Support Services.
My panel has a mixture of yellow and red PTT indicators so I obtained several different kinds of bulbs in order to see whether I could make these more effective. Here are my perceptions on the various tests that I've performed so far:
Yellow indicators (Oak Grigsby):
Red indicators (Dialight and Oak Grigsby)
For now, I will probably use the yellow LED bulbs in the yellow indicators and the 700 mcd white clear LEDs in the red indicators. There are also some much brighter LED bulbs which I'll probably obtain and do some more testing, but one of the problems with many of the brighter LED's is that they have a very limited viewing angle. I also have plans to add a Master Caution indicator which will be a red flashing LED with a clear lens; the push-to-test function will actually be an acknowledge/reset function.
P&B W31 CIRCUIT BREAKER / SWITCHES
I've talked to one builder who has had three of these fail mechanically but I don't know if this was due to a bad batch of them or just a result of their use in our application. They are a mainstream aviation breaker from a reputable manufacturer. However, I talked to a very experienced avionics technician and he indicated that he will not use these devices, even in experimental aircraft.
My biggest concern is for the Master switch and the possible inability to turn off all power or the unexpected turning off. The inability to turn it on would also be a very frustrating experience. Luckily, my master switch is a simple mechanical DPDT switch, although this choice was originally made for a totally different reason.
When I was installing these devices (lighting circuits only), I did notice one unusual characteristic. There are two nuts on the mounting lug and this arrangement is often used in electrical items to slightly alter the installed height of the device. However, when the lower nut is not snug against the main body of the breaker, the mounting lug and toggle can slightly rock in relation to the body. I chose to snug the lower nut against the body in order to prevent any of this movement.
One builder has experienced a quirk with his breaker-switches whereby he can't turn them ON if the cabin has gotten very warm i.e. the thermal trip feature is preventing engagement. I've also found that even on the bench, several of mine are very sensitive in the ON position and can easily be tripped OFF with just the slightest bump. If one is installing these units, then I think it would be prudent to test the OFF/trip "feel" for all of them and then use the most "solid" ones for the most important functions.
There is interesting information about a Beech AD and the internals of these breakers located here.
I believe the 5 amp switch/breaker that is supplied for the navigation lights is under sized if wired per the construction prints as these three lamps are shown as 2 amps each in the Whelen catalog. Another Whelen document shows the forward position lamps as 1.9 amps each and the tail position lamp as 1.8 amps for a total of 5.6 amps. If the instrument lights are wired as per the POH to the navigation light circuit, then this breaker is loaded even further. Normal aviation practice is to not load a breaker to (or beyond) it's rated capacity ... perhaps 80% is a more realistic target.
The supplied strobe lamp is not aviation approved, therefore it would *NOT* be legal for night VFR use in Canada. While this strobe may provide some level of indication, it certainly is nowhere near as bright or effective as the aviation approved units that I've looked at. From the data I've seen it would appear that the supplied strobe is a single flash clear unit with 50 effective candela while the helicopter requirement is for a red unit with a minimum of 150 effective candela (400 effective candela red or white for fixed-wing aircraft). Switching the lens on the supplied strobe from clear to red would further reduce its intensity to only 10 effective candela according to the data sheet!!! Adding a blocking plate in the forward direction (as I've seen on a factory assembled craft) would also negate a strobe's legality since the requirement is for an anti-collision strobe light to project light 360° around the aircraft's vertical axis.
I am aware of one solenoid board that shattered while it was being cleaned, but I don't know what product was being used. I'm led to believe that LEXANŽ is known to have some chemical incompatibilities (especially ammonia which is used in many cleaners) and is also prone to stress-cracking if not properly machined. The first concern with this issue is that there may be a bare battery wire flopping around if the board breaks and the builder hasn't installed silicone boots on the connectors. Probably a bigger issue is that the CHR-supplied contactor and wiring is such that any touching by the contactor case to a grounded structure will activate the coil and have exactly the same effect as the master switch being turned on. The only way to stop the flow of electricity is to either remove the short or a battery cable.
I've chosen to replace this board with phenolic material, changed the contactor to an isolated case model and also installed silicone boots on the battery and control wires / terminals.
I have heard of issues with the vertical stabilizer rivets becoming loose and/or coming out. Most of these were during the initial testing and before the tail rotor had been fully balanced. I have also seen pictures of ships where the number of mounting tabs has been nearly doubled. I have done some testing with an air hose and feel that the factory supplied tabs alone are not adequate ... primarily I saw a drumming action that could lead to weakening of the rivets and/or stab sheeting. Alternative solutions might be to add additional tabs during construction or on a completed craft to use some kind of an adhesive in the gaps between the tabs to adhere the stab to the boom.
The horizontal stabilizer sheeting has the potential of a slight "rocking" motion since it is essentially a flat sheet that is riveted to a round tube. I chose to add a bead of Hysol adhesive (mixed with a bit of flox) between these two surfaces when they were riveted together in order to create a bit of a flat bed. I've also been told by a builder that he popped the inner rivets on the spar due to vibration before the tail rotor was dynamically balanced. This was witnessed by CHR personnel and they commented that it wasn't uncommon to see this.
Both stabilizers should be carefully checked during the pre-flight, both for "smoking" rivets and any possible cracking. Note that there is a bulletin available from CHR about the inspection of the horizontal stabilizer.
Example of smoking rivets
I have been in contact with a builder who had a concern about what would happen in the event of a tail rotor failure. This was precipitated by a conversation about an accident investigation and it was hypothesized that in the event of a tail rotor failure that the bolts holding the gearbox in place could be sheared and the whole tail rotor and gearbox assembly could depart the machine. With the sudden weight loss of ~37 pounds at such an extreme moment arm, the results would not be pretty. This builder has wrapped a cable around the gearbox and then secured it to the frame as a preventative measure. Thus in the event these bolts are ever sheared, there should not be a significant change to the CofG and the machine can still be flown and landed without tail rotor control.
Sheet #16 of the construction prints show a 1/4" hole centered at the back of the pilot's collective to allow passage of the starter button wire. In my kit, the stubs on the collective pivot tube supports were actually 2-1/2" long instead of 2" as shown on sheet #20. What this means is that if one is to center the hole for the wire based on the collective, the pivot tube support stub actually encroaches on the hole and will pinch or rub on the nylon feed through tube and the actual starter wire. Depending on the ship's wiring, this may unexpectadly engage the starter or more likely result in a short circuit and burned wire if this circuit is not fused between the power distribution bus and the starter switch.
There has been an incident where a control rod broke and this appears to be a relatively clean break in the center (ie. non-tapped) area of the rod. There was initial speculation that this may have been the result of the anodizing process that the builder used and/or a material flaw. A laboratory analysis would seem to indicate that the break is not the result of anodizing, but rather an existing crack that slowly progressed and finally broke through. Since there was an area of discoloration on the crack, this is a graphic example of why the daily inspection should be thorough and include a visual check for any areas of discoloration and/or possible cracks. Also, light colours tend to make this kind of crack more visible.
Update: Before performing any chemical processes on any critical control piece, the builder should read the May '07 issue of Experimental Helo. I need to re-read this article and do some more research of my own, but it would appear that any of the normal aluminum processes (anodizing, Alumiprep, Alodine, etc.) will significantly alter (i.e. lower) the fatigue life of the base metal.
In an NTSB report, it was stated that CHR has changed the aluminum rods to steel rods but it is not clear whether this is in reference to the straight control rod modification or also applies to ships with walking beams. Since CHR has chosen not to document this change on their website, it is probably prudent for all owners with aluminum rods to contact them for clarification.
On some ships, the pitch control rod may rub on the frame just below where it exits the cabin through the firewall upper rib. This only occurs at some extreme movements, but it is an area that should be checked. The author makes no recommendations as to whether the use of UHMW tape, frame reliefs, rod bending or just inspection is the proper remedy.
I am aware of one builder who had a fuel tank strap break at the bend for the clamping bolt. I do not know whether this was from over-tightening the clamping bolt or possibly from too tight a bend radius and metal fatigue. I leave it up to the builder and/or CHR to evaluate whether the simple bent tang arrangement satisfies their requirement for a safe retension system ... I know it doesn't meet mine and I've chosen an alternate solution. In the meantime, this is an area that owners should add to their pre-flight checks.
I have also talked to a different builder who had a bolt break on the right rear tank strap. This bolt was used both to hold the tank in position via tensioning the strap and also as an attach point for the cross brace. His recommendation was to use separate straps for the hold down and cross-bracing but he has since gone to a much different design.
I have also heard second-hand of yet another problem that occurred with the fuel tank straps. When one strap lets go, the tank will start moving around and the second (and final) strap will be under a much larger load and more prone to breakage.
I'm led to believe that a very agressive right pedal stomp could cause the tail rotor to contact the tail boom which would then require the spindle, output shaft and blades to be replaced. This has only occurred once during factory testing of extreme conditions and should never occur during normal flight. I've flown an R22 in 40 knot gusting conditions and know that one often has to use a lot of pedal motion. However, the key to avoiding stomps is to immediately recognize deviations and then apply corrective inputs smoothly and quickly.
There is a Service Bulletin available from CHR in the Owner's Area of their website that addresses a possible issue if an owner is to dis-assemble / re-assemble their tail rotor assembly. It is possible to install the spindle backwards and this is *NOT* recommended under any circumstances. More information is available in the bulletin which all owners should review. Also note that some builders may have a .PDF drawing of the tail rotor assembly that should only be used to illustrate the various parts and *NOT* as a re-assembly document since not all parts are shown and various orientations may be ambiguous.
I had heard from another builder about a leak in a fuel tank and I did a very low pressure test on my tanks. The result was that I found one pinhole leak in the welds on one tank and five pinholes in the other tank. After re-welding these holes, I did a higher pressure test at 5 psi to see whether there were more leaks ... ten more pinholes were identified!!! Based on my results, I would strongly advise that all fuel tanks be pressure tested before painting them. I have also discovered a very significant leak in the welds on my oil tank.
It is my understanding that the factory has changed the welding technique on these tanks and now performs a pressure test on the fuel tanks after they're constructed. I'm not sure at what time they implemented this procedural change and if there is any doubt, then it's probably best for the builder to perform a pressure test just before painting and installation.
It is my understanding that the older rounded tip blades have a potential problem with the absorption of humidity through the vent hole. As I remember, this has been identified when the machine is in a high humidity area and there is a considerable temperature change such as overnight cooling. My understanding is that the humid air then enters the blade through the vent hole, condenses and cannot escape until after the blade has been re-warmed such as by being left out in sunlight. The negative effect is that the balance of the blades is affected and there can be considerable vibration.
I have been in contact with one builder who was having a lot of problems with getting reliable and accurate readings on the CHT gauge supplied in the instrument kit. After doing some research, it would appear that there are several potential issues with the supplied gauge. This instrument is of the non-amplified type and is not compenstated for ambient temperature changes. According to the documentation it is calibrated for a temperature of 75° F, but ... that compensation is at the reference (or cold) junction of the thermocouples. This means the actual gauge reading is offset by the ambient temperature at the junction of where the thermocouple leads join the extension cables. Since these junctions will possibly be outside the craft (at least for cylinders 1 & 2), at an outside air temperature (OAT) of 0° F the gauge will read 75° high and conversely at 100° OAT the gauge will read 25° low. This is before even considering the effects of wire routing and interference on these very low level signals. Even though this gauge has visual redlines, the reading could be off by 15-20% and the pilot must be aware of the offset effect to determine the true temperatures.
Per CHR documentation, the tail rotor must be dynamically balanced within four hours of the initial startup.
When performing initial testing, the drivetrain must *NOT* be run with the tail rotor and associated drivetrain installed but without the main rotor blades installed. Ignoring this restriction can cause a transmission failure that may not be covered under warranty.
Several RV builders have received non-AN fittings that were ordered from Aircraft Spruce even though they ordered standard AN parts and paid the appropriate price. The finish is a brighter & shinier blue and there is no AN, MS or AS markings on the part. As such, there is no knowledge about cast vs. forged, alloy type and general suitability for aircraft use, especially in areas such as fuel and pressure lines. I actually received a couple of these and decided not to use them since they were to be used on my oil pressure sensor.
There is an interesting article about AN vs. Industrial fittings here.
When I received my transmission, all of the bolts / castle nuts except for the three that need control rod ends had cotter pins pre-installed. Also, most of the drilled head bolts also had safety wire in them. However, the four bolts that hold the pylon to the mast tube did not have safety wire installed. Another builder received his transmission with safety wire on these bolts, but I also have seen a couple of factory ships that did not have safety wire here. Although the safety wire routing is a bit unusual, I believe these bolts should be safety wired.
Factory craft without safety wire and a builder's craft with it
I noticed that a couple of ships had a vent tube attached to the fuel tank caps even though these are supposed to already be of the vented variety. In talking to the owner of one of these craft, he told me that he had experienced two venting failures of similar caps when he flew fixed wing craft. Although it is a builder's choice whether to add a vent tube, I think all builders should be aware of this. It is very easy to detect a problem during construction but the concern is after the craft is operational. Based on the Safari's fuel line plumbing, I would expect a venting problem in one cap to show up as uneven fuel levels on the gauge ... essentially the fuel being mostly supplied from one tank only.
Update: CHR now has placed a notice to builders on their website about this potential issue.
MA4-5 CARBURETOR MIXTURE CONTROL
The mixture plate supplied with the kit which changes the pull angle on the carburetor's mixture lever was designed for a MA4 carburetor typically found on O-320s. Although it will fit and work with the MA4-5 carb commonly found on O-360s, the lever is considerably different and there is a potential problem. Since the plate does not fit tight to the lever and is held by a single bolt, it is possible for the plate to rotate on the bolt if it not completely tight at all times. Also, the lever on the MA4-5 carb is angled and if the nut holding the control cable were to loosen, the plate and lever will rotate due to gravity. In either case, any movement in flight would cause the mixture setting to change from the typical full rich setting towards the full lean cutoff position.
I have been informed by another builder with a stock CHR exhaust system that after just a couple of hours of operation he had signs of cracking on the weld at the sides of the exhaust "can" where the tailpipe is welded to it. He also noticed that the welds on a factory ship had cracked at this same point. I believe this is due to the mounting arrangement on the stock system whereby it allows fore-aft movement of the tailpipe but essentially no side-to-side or vertical movement. Considering the engine mounting, I would think that there would be fore-aft movement due to thermal expansion but engine vibrations would be primarily side-to-side and possibly a bit of vertical. Left undetected, the worst case scenario would be that the tailpipe breaks off in flight and heads towards the tail rotor.
I am aware of a main rotor yoke cap that broke as a result of a backfire during engine startup. I would thus recommend if there is a backfire that the engine be shut down immediately and the drivetrain be inspected with particular attention being paid to the yoke caps.
It appears that newer kits are now being delivered with what I describe as a "pancake" oil tank ... essentially two ends from fuel tanks welded together. I understand that CHR has also decided to change the crankcase venting arrangement but I can find no reference to this on their website documents and I understand that owners of this new tank do (did?) not receive any written instructions about these changes. Note that the Construction Manual(s) and Prints describe the old style tank and venting arrangement which are not correct for the new tank.
In essence, with the new style tank the engine conversion no longer uses the 3/8" vent line shown in the plans as going from the #1 cylinder drain-down crankcase hole to a fitting on the oil tank. Instead of just three plugs, all four of the drain-down holes in the crankcase are now plugged. The conversion now relies solely on the main large oil return line (going from the accessory case to the oil tank) to provide both an oil return path and to vent the crankcase. The new tank is then vented to atmosphere in the same way as the original tank.
Update: I have now seen a factory installation of the new style tank where they have used a vent line from the #1 cylinder drain-down crankcase hole. This vent line then went to a "Y" fitting in the line between the #3 cylinder valve cover and the oil tank.
If a builder has received a new style tank without current written instructions, I'd recommend that they contact the factory just before plumbing their tanks to discuss the venting du jour philosophy. Failure to properly implement this undocumented venting arrangement can result in a blown crankcase nose seal and/or a major loss of oil as discovered by other builders through the school of hard knocks.
The oil cooler supplied in the Safari Kit is known to have an issue with cracking around the oil hose bosses when used in it's original application. I am now aware of a Safari owner who has experienced it. This area should be included in the Daily Inspection and if any weeping of oil is found, it should be investigated for possible cracks before just tightening the fittings. Further tightening of a tapered NPT fitting in a cracked boss will just make the problem much worse.
I am aware of two builders that have experienced problems with the floor covering (i.e. carpet) when exposed to bright sunlight on warm days. Specifically, one experienced "melting" and the other experienced "burns". In talking with the factory, one builder was told that this is a known problem when the craft is parked with the tail towards the sun and it would appear that the bubble creates a magnifying effect. Besides using appropriate material and considering the parking direction, it is possible that a bubble cover should be considered.
While the wiring diagram on Construction Print #48 will work under most conditions, I think there are some serious shortcomings, especially with the wiring for the alternator control circuit. I am aware of several builders who have had problems in this area and regardless of flight safety issues, it causes both time and money to correct problems that can be generated as a result of this wiring.
I've commented above about the reliablility of P&B W31 Switch / breakers ... I do NOT believe they should be used for either the master switch or the alternator control switch. I chose to replace them on critical circuits with high quality toggle switches and Klixon breakers. With that being said, there is a bigger problem when following the provided diagram ... it is possible to have the master switch OFF, the engine running and the alternator control switched ON. This is definitely an aviation no-no and can result in over-voltage conditions or spikes on the power bus regardless of whether one is using an internal or externally regulated alternator. Normal aviation practice is to use either a Cessna-style split switch with a mechanical lockout or some other form of electro-mechanical lockout.
One of the implementation problems that may be encountered is that most existing Safari's will require a retrofit. I'm adding the following diagram as an example of a simple implementation which *MAY* correct this issue ... it is up to the individual owner to determine if it meets their needs. It uses separate switches and breakers with the key feature being a two pole master switch (i.e. DPST). In order for the alternator control to receive power, both the master and alternator switches must be in the ON position.
DUAL ELECTRONIC IGNITION SYSTEMS
If the builder chooses to use dual electronic ignition systems, such as Lightspeed units, then they should carefully analyze their electrical system for single points of failure. Some of these systems absolutely require a source of electrical power and will result in an engine-out condition without it. There are many alternatives for backup power and I'm not going to recommend one technique over the other ... I'm simply pointing out that it's essential to consider the implications and possibly make wiring changes.
Like many people, I do not have a paid subscription to Unison's service letters and have to gather information wherever I can find it. There is a mandatory service bulletin from Unison that effects certain models and serial number ranges of their magnetos ... there may be others, but these are two that I'm aware of. They can be found on the Lycoming website: SB583 (Unison SB2-08) and SB584 (Unison SB3-08).
This has now been extended with Unison's SB3-08A.
While removing the rod ends in my tail rotor in order to loosen them up (a step omitted by the factory during assembly), I discovered a bolt in the tail rotor slider cross that was severely over-torqued at the factory. This had been done to compensate for thinner than normal washers between the rod end and the forks in the slider cross. I chose to replace the bolt/nut and use thicker washers, but it does raise concerns about all the bolts and nuts that have been pre-installed by the factory. It is very obvious that a torque wrench was not used in accordance with AC43.13.
The star coupler on the tail rotor gearbox input shaft is held in place with a setscrew and although the Input Housing drawing shows a keyway, I don't remember seeing one. I have talked to one owner who had this setscrew repeatedly loosen on him which could lead to a loss of the tail rotor rotation. His rememdy was to remove the setscrew and apply a bit of Loctite to it ... it hasn't loosened since then.
When I took out my setscrew to add the Loctite, I noticed that it came out very easily and there appeared to be no indications of Loctite or other secondary locking mechanism.
The various stops for the cyclic (8) and collective (2) each use a bolt and a jam nut to set the stop and then lock it into position after adjustment. This kind of arrangement does not provide the two levels of securing that is normally used with aircraft bolts and has been known to come loose on other helicopters resulting in incidents. One could possibly use a toothed lock washer under the jam nut or removeable thread locker (normally blue).
I am aware of one owner who noticed that the engagement of his centrifugal clutch seemed to be changing. Upon cleaning and inspection, he discovered that three of the four springs that press the shoes outward were broken. This was on a Safari with just over 200 hours.
I am aware that an NTSB investigation revealed that an improper relief radius had been machined into a tail rotor shaft. I have also been in contact with another builder who discovered that there had not been reliefs machined into his main rotor shaft. The purpose of the relief radius is to distribute stresses to try prevent stress risers which can then lead to failure. Since parts are often made in batches, I have no idea if other shafts are in the field that have similar machining errors. I am not aware of any Service Bulletin or recall as a result of this.
ANY shaft that is removed for overhaul, inspection or other purposes should be carefully inspected to verify that proper reliefs have been machined into it. If the owner does not feel qualified, a good machine shop should be able to measure and verify this. As to whether this warrants a precautionary inspection ... I leave that opinion to CHR and to the individual owner who is responsible for maintenance.
There are four bolts at the top of the main rotor mast tube (through the pylon) and four bolts at the bottom that connect it to the transmission housing. I am aware of a builder who has sheared one of these bolts on a couple of different occasions. I would recommend that checking these bolts be added to the daily inspection list ... since the safety wire may hold a sheared bolt in it's relative position, it will be necessary to physically touch the bolt head to see if it can be moved. Perhaps we (or CHR) should be using NAS bolts in these locations ... they're about 25% stronger in both shear and tension.
It is hypothesized that these bolts carry the majority of the weight of the machine during aggressive manoeuvers such as quick stops and the flare at the bottom of an auto. Although there are seven bolts holding the transmission housing to the frame, those bolts are attached to members that have a fair amount of vertical give to them whereas the pylon is held in place by the pylon rods which effectively create a triangulated structure that is much more rigid.
The Curtis drains that are used to sample the fuel employ an O-ring to seal in the closed position and prevent fuel drips. These are known to leak on many new Safari tanks, sometimes quite a bit. Whether it is due to the original manufacturing process for the tanks, the drilling and tapping of them or the combination is up for speculation ... regardless, it would appear that these tanks get a lot of "swarf" in them which then settles into the drain reservoir (the lowest point) and subsequently cuts the O-rings. Since the O-rings are not available separately from Curtis, their solution is to replace the valves if they start to leak (~$10 each). While I have heard of several cases of this, plus a first hand experience, I have not heard whether the gascolator drain is subject to the same issues.
The instructions that accompany the Rexnord Omega coupling for the tail rotor drive indicate that the capscrews should be torqued to 204 in-lbs or 17 ft-lbs. That may be appropriate for specialty capscrews, BUT CHR has chosen to supply AN style bolts with drilled heads rather than the Rexnord capscrews with the locking patches. This means that the supplied instructions specify a torque that is nearly three times the maximum specified in AC43.13. I am aware of two builders that have sheared bolts while trying to meet the torque specification in the instructions. If they had been able to reach this value without shearing the bolts then the bolts would have been considerably stretched and it's possible they would fail in operation which could lead to a complete loss of tail rotor power.
One builder was told by CHR personnel that there was supposed to be a clarification sticker from CHR included with the coupler's instructions. I am not aware of any builder that has received such a sticker or notice ... certainly I didn't. I believe this is another case where the factory knows there is a serious error but does not feel that it's necessary to notify the builders about it.
One of the things I noted when installing the yoke caps is that the head angle plates I received will require some machining. The seating area for the washers under the four bolts encroach onto the curved inner radius of the head angle plates. Thus the washers are at an angle and if one is to try snug the bolts down they would not be getting accurate torque readings. The bolts would also have all of their force being applied to one side of their heads.
It's my understanding that some kits were delivered with a stainless steel jumper between the master contactor and the starter contactor. I do NOT recommend this and I've received a report of a jumper that got so hot it started to melt the LEXANŽ board that the contactors are mounted on. One table that I've looked at indicates that 304 stainless has ~40 times the resistance of copper! This jumper should either be a piece of #2 wire with proper lugs or heavy copper (i.e. 1/8" x 5/8"+). Brass might also work but I prefer to use copper.
On my kit, the forward half of the star coupler that is attached to the tail rotor driveshaft is held in place with a single AN4 cross-bolt. I was very careful to undersize the hole in the driveshaft and then ream it to size for a resistance fit. After four hours of running it was noted that this hole had started to oval and there was noticeable movement of the star coupler on the driveshaft. Similar to another owner who had noted this, I chose to remove the star coupler and add a lot of red Loctite under it during re-assembly. Time will tell whether it will prevent further wear and movement.
I've been told that newer kits may have a keyway on this part but I have not personally seen one of these.
It should be noted that there is a Mandatory Service Bulletin regarding carburetors with floats that may be leaking which can lead to an excessively rich mixture and possibly engine stoppage. The bulletin is available here.
Last updated: September 18, 2009