Jump to content

RPM Rollback


Runtoeat

Recommended Posts

  • Replies 65
  • Created
  • Last Reply

I'm a fly on the wall and appreciate the thoughts here.  We kicked things around today before flying to the hangar queen CTLS.  We ended up not going.  We figured this is bigger than any of us after we pulled the heavy maintenance and talked to Brett at LEAF.  We thought that there was just one trigger coil running the modules.  Today looking at the diagram and talking to Brett, we found that there are four coils - 2 to each module - and if I recall the conversation correctly, these provide a sign wave which the modules use to determine crankshaft speed.  This problem will require someone with diagnostic equipment and knowledge of the Rotax ignition and fuel system to straighten out.  It is beyond our abilities and we do not want to find that we guessed wrong once someone is in the air.  We did kick around the step where we would pull the soft start off the starter relay but reasoned that this is a "non event" during the engine idle so it would provide no result.  Now, with Tom's idea that there could be a 12v. leak and Corey's idea to touch the soft start to a 12v source while idling, this might provide some data.  If applying 12v. takes the modules back into the soft start mode with retarded spark, and the rpm demonstrates the 1600 to 800 drop we intermittently see, and if we have smooth running, which we shouldn't based on Roger's comments...but....if we do get the similar rpm drop, this would be the failure state we might be experiencing.   Not sure if I mentioned it but when the idle has drifts down to 800, with the throttle lever at full stop, we can advance the throttle to bring the rpm back to 1600.  Then, once at 1600, the throttle can be returned to the stop and 1600 is maintained. Could both idle jets be temporarily picking up dirt simultaneously and slowing the engine down and are we then bypassing the idle circuit by advancing the throttler which brings rpm up?  I believe the next step might be to find expert Rotax repair locally.  I have found Rainbow Aviation's state by state Rotax repair facilities and we will probably start calling to see if there is someone who may have dealt with similar problems on the Rotax and if not, if there is a person or facility with diagnostic equipment to analyze this. The "ah, ha" for me here is to realize that our Rotax engine electronic ignition is fairly complex and trouble shooting and fixing problems involving this, if this turns out to be the case here, requires technical training and equipment.  Guess I'm rambling on here.  Thanks for the comments.  If you think of something, please let me know

Link to comment
Share on other sites

Two things:

Have you tried a new set of plugs? I've had issues with an aircraft slooooowly losing RPM at idle with old plugs, but when throttle is advanced then retarded, RPM would stay up for a bit, then creep down again.

Use another tach or stroboscope on an ignition wire with RPM readout and verify your tach isn't failed. Note that the thick silicone insulation does make it hard for the clamp on stroboscope method to read correctly.

Now that said, your engine won't stay running at 800 RPM if it was really 800 RPM and firing at advanced timing. No way.

Link to comment
Share on other sites

This is pretty basic, but have you checked the throttle cables end-to-end?  If they were somehow binding up or slipping somewhere, it might cause weird throttle/engine behavior.  It would probably run rough unless both cables were having the same issue, but it's easy to check and eliminate.  I didn't see the throttle cables mentioned in your original post.

Link to comment
Share on other sites

Corey, you are right.  The old adage....KISS.  I mentioned the broken insulators to my mechanic early on but he acknowledged this and said the plugs weren't broken and we moved on during our inspection. This should be revisited.  Your comment about the engine not running smoothly at 800 rpm with the spark advanced is a point toconsider. You mentioned elsewhere on the forum that the Rotax uses "pulse width" signal for timing? Would this be the variable voltage versus engine rpm which the trigger coils send to the modules?  Do the modules use this variable voltage signal to incrementally increase spark advance in the +3 degrees ATDC (start) to -30 degrees  BTDC (run) range?  Or, do the modules provide just two levels of spark, i.e., 3 degrees ATDC (for start) and 30 degrees BTDC (for run)?  If the spark is incrementally raised, perhaps there is a small enough change occuring at idle which causes reduced engine rpm but not enough to cause rough running and this might be the problem?  Or, if there are only two levels of spark advance and there is "a jump" from -30 to +3 at idle, then this probably would cause rough running and the suspicion that the ignition is a cause might be ruled out here.

Andy, the cable have been checked and appear to be OK.  Thanks.

Link to comment
Share on other sites

Hi Roger.  Mag drop is normal - around 50rpm.  EGT is normal as shown on Dynon.  Sure wish this CT was near to you so you could pull your hair too!   I like the idea of introducing 12v. while @ 1600.  Thinking about Corey's comments on running at advanced spark while at 1600 rpm?  Would you think the engine would be rough?  I'm thinking, if the spark is advanced in the two step method, then after the engine reaches 250 or so rpm during start, the modules will fire the engine with retarded spark but will advance the spark when rpm reaches 800 or so.  So, I'm thinking the engine may run smooth at 800 rpm with advanced spark.  Introducing 12v. would test this theory.

I'm going to work on my old Force / Merc outboard engine with points and condensor today - ahhh, back to simpler times.

Link to comment
Share on other sites

I talked to John today and got more back story. He has something for you to try. I believe you may just have vapor lock and gave him a test to try to rule it in or out.

Do this first and quit wrestling with the electrical for a while.

 

Dump all the 91 oct. and put 100LL in it and go fly it.

 

Link to comment
Share on other sites

On 8/16/2017 at 8:10 AM, Runtoeat said:

You mentioned elsewhere on the forum that the Rotax uses "pulse width" signal for timing?

Pulse width modulation is a technique used to control the width of the injector solenoid on/off state in the 912iS. The longer it is on, the more fuel that is injected.

As for the ignition system, I'm not sure if I said it uses PWM. If I did, I used the wrong term.

The ignition system is powered by a couple of coils in the magneto generator. There are small tabs on the flywheel that pass by the trigger coils, which are hall effect sensors. This illustrates the idea quite well:

Hall_sensor_tach.gif

 

If we make the magnets larger, the sensor will stay on longer. It is the falling side of the signal that triggers the spark in normal operation, but it is the rising side that triggers below 600 RPM and during soft start.

Link to comment
Share on other sites

Roger, thanks for giving us some suggestions.  I think John has a plan to drain his wings and fill with 100LL.  The only thing I might add is his fuel flow and pressure do not show change on the Dynon during the low rpm.  The flow and pressure remain constant and normal.  We will look at fuel vapor lock per your recommendation but feel that there is good flow and pressure at the low idle.  We are also going to introduce 12v to the modules during a "normal" 1600 rpm idle, assuming that we can get a consistent 1600 rpm idle. Not sure what this will tell us but perhaps it will result in the rpm drop from 1600 to 800 that is intermittently occurring and this will mean the modules may be involved  If the 12v. and/or the 100LL doesn't provide us with results, John will need to continue to look for a solution to his engine's rpm rollback.

Corey, thanks for that neat graphic.  Can you provide me with a layman's explanation of section 3.22 (page 306) of the heavy maintenance manual I am attaching?  I'm assuming that the graphic here is showing typical oscilloscope readings at various rpm's.  It appears that a voltage is generated by the coils which varies with engine rpm.  I am assuming the modules process this variable voltage to determine when the spark should occur?  Is this the basis of what a Rotax repair center would do if we were able to take the CT in for trouble shooting?  Would the engine be run at various speeds while the trigger signal is analyzed?  Can the trigger signals be monitored "real time" for each module during a "event" (i.e., a "rollback to 800) to see if there may be incorrect voltage being generated per the rpm that is occurring?

Rotax ignition diag1.pdf

Link to comment
Share on other sites

@Runtoeat PWM is a control method, hence the term "modulation". For ignition, this is entirely fixed; the width is only affected by the speed of rotation of the flywheel and the size of the physical tabs. In the new flywheel, the tabs are a little bit larger so that the retarded timing side fires after TDC. Again though, we aren't modulating anything.

Here's another way of looking at it: what we're doing is compressing the graph into a shorter time period when we increase RPM. We're NOT changing the ratio of "on and off" during each cycle (called duty cycle in PWM).

The graph you are looking at in the manual is what the trigger coil signal should look like when you hook up an oscilloscope to the circuit. A properly functioning system should show one pulse per cycle, with the voltage at around what the manual shows in the graphs.

Honestly, while I am glad they have those graphs, they are showing 500 and 6000 RPM. I am not certain if the voltage climbs linearly with speed. It would be a lot more helpful if they included either a note saying that it does, or have a table of RPM to voltage. I have an o-scope, but I've never hooked it up to the ignition system.

Link to comment
Share on other sites

Thanks Corey.  You answered my question.  I wondered if the "O" scope would be a tool that could be used to see if the trigger coils were outputting correctly.  If I'm not simplifying things too much, I'm thinking there might be a Rotax procedure where the mechanic might be able to trouble shoot the trigger coils with a "O" scope to determine volts vs. rpm and  a table of values to compare the findings with.

Roger, thanks for the call today.  I am attaching the shot of John's Dynon during a "low" rpm event.  Phil apparently took this when the Dynon showed 960 rpm.  Now, since our conversation, I must believe you when you tell me that a Rotax engine cannot run at this low rpm without shaking it's self to death.  Few have your experience and knowledge of this engine.  But, still.......after looking at digital displays all these years, it is very hard to set the display aside and ignore it. Then too, Phil and Dave have thousands of hours in aircraft and I need to ask them "what did you feel when the Dynon showed the rpm drop?"  Thinking back, I'm not sure if the drop was felt by them or if they responded only to the visual "drop" shown on the Dynon. I know they did not feel any roughness, which you say is not possible if the engine was at 890 rpm.  I need to sit in the CT and try to experience what Phil, John and mechanic Dave saw and felt.  Along with this, I want to see if  we can install a optical tach to capture prop speed and get accurate engine rpm after correcting for the 2.43 reduction.  I don't think any of us here have your sense of adventure and are ready to jump in and go.  The increase in rpm from 4800 with intermittent rpm roll back rpm of 4500 (indicated?) was eliminated and we now see stable static 5000+ rpm  with no rpm rollback after the new ignition switch was installed.  With good full throttle power it is tempting to buckle up and go but we will need a little more data to understand the intermittent low rpm we still "see" during idle.

59978c18b0347_johnsdynon-890rpm.jpg.69a349e9c43844b67328fd6e7e4f2e50.jpg

Link to comment
Share on other sites

Update:  Yesterday we were able to fly the CTLS back to our home airport and today we were able to get into a detailed study of the wiring and the connections at the modules.  We found that the connection of the soft start wire and the solenoid wire blade connection at the starter relay was suspect and reworked this with new connectors.  We removed the modules, inspected the 6 pin connectors and the wiring.  We applied dielectric grease and reconnected the connectors, after not finding anything amiss here.  We performed the resistance test for the module wiring called out in the Rotax Heavy Repair manual.  Next, we removed the new ignition switch and inspected this.  We discovered that some of the wires connected to this switch were loose and in poor condition.  Even when the screw was turned tight, some of the wire were still loose and could be moved.  We attached blade lugs which could be torqued tightly by the screws. For test purposes, we made a temporary harness which supplies a temporary constant 12v. to the modules soft start circuit whenever we choose to do so,  After completing the inspection and repairs to the wiring, we proceeded to our testing:

Test #1.  Pull the 12v. soft start wire from the starting solenoid and start and run the engine.  John, the owner, had not started his CT without the soft start and was amazed at the "explosive" start of the engine without the Soft Start.  Less than 1 turn and BANG!  Off and running.  We continued to run the engine up to and above 5050 rpm static with no roll back.  We pulled the throttle back to idle and the engine settled into a 1800 rpm idle, WITH NO ROLL BACK TO 800 RPM!  We sat and played with engine at various rpm.  Solid with no rollback.

Test #2. Keep the 12v. power to the modules pulled from the starter relay.  Connect the external 12v. source to the soft start wire with momentary switch.  Start the engine and run to 5000 rpm.  All steady with no rollback.  Introduce 12v. to the modules soft start circuit.  Rollback occurred which reduced the 5050 to 4500 rpm.  Remove 12v. source.  Steady 5000 rpm.  Repeat introduction of the 12v.  Rollback to 4500 rpm.

Test #3.  Repeat Test #2 but introduce 12v. to the soft start circuit while idling at 1800 rpm.  Steady 1800 rpm.  Introduce 12v.  Rollback to 850 rpm.  Remove 12v.  Back to 1800 rpm.  Repeat 12v. to soft start.  Idle reduced to 850.  Remove 12v.. 1800 idle with no further rollback observed.

Test #4.  Temporary external 12v. removed and 12v. soft start circuit wire re-connected to solenoid.  All systems now @ "factory" condition.  5050 rpm.  Solid with no rollback observed.  1800 rpm (idle).  Solid with no rollbacks observed. 

Summary.  We can create rpm rollback at 5000 rpm and 1800 rpm by introducing 12v. to the modules "soft start" circuit.  Our testing shows the Rotax will run at 800 rpm when the modules' soft start circuit is supplied 12v.  We assume this puts the modules in the "start" mode and assume that the modules are providing the 3 degree ATDC spark during this condition.  The Rotax operation during this condition is not "smooth" but is jerky and rough but still it is running forcefully.  An interesting sidle note is if the throttle is just touched, not really moved but just touched, the rpm immediately increases to the normal 1800 idle rpm and remains there.

Although we can recreate the "rollback" condition by introducing 12v, to the soft start circuit, we are not totally sure that there are not other reasons why the modules went into this rollback condition.  We believe that a 12v. "spurious"? current may have been entering the modules.  We don't fully understand how this could happen.  We surmise there may have been some bleed thru of voltage at the ignition switch due to poor wiring connections and condition of the wiring.  Maybe poor grounds where ground loops are occurring?  Perhaps this somehow occurred at the hard shell connectors where the modules are connected?  We did not find bare wires laying near any 12v. source where this might have occurred.  Due to the wire supplying voltage to the solenoid and the wire supplying voltage to the soft start are connected to the solenoid, we question how voltage would be going to the soft start circuit but no current apparently going to the starter solenoid?  Perhaps the starter was being activated but John did not notice any operation of the starter and perhaps we did not notice this either when we first began looking into this problem?  As the plane now stands, it is running correctly and we don't seem to be getting the rollbacks. This time, after repeated runnups to 5000 and multiple times going into the idle mode, the rpm remains stable. The Rotax runs noticeably different since installing the new ignition switch.  It now runs consistently @ 5000+ rpm, which is well above the previous maximum rpm of 4800 found since this CT was purchased used with under 300 hours on it.  Now, since going thru the wiring and correcting questionable wire connections at the switch and at the hard shell connectors and at the starter solenoid and tightening all grounding connections, the idle condition has not exhibited rpm rollback and the condition feels stable.  I believe that the plan now is to work the Rotax hard for a few hours, making numerous take offs and flying close to our airport, until we have confidence that this concern has hopefully been corrected.  Wire gremlins can pop up anytime.  Hopefully, they won't return.

Link to comment
Share on other sites

If your soft start is hooked up to the starter solenoid and allowing voltage to the soft start wires after the engine is running then you may just need a new solenoid. 

Most soft starts are hooked up to the solenoids without any problem so that makes the solenoid at least suspect until it gets replaced.

 

Link to comment
Share on other sites

We have checked to see if there is any current at the solenoid end of the start wire when main buss power is "on" and the switch is at "run".  None found. However, there could be other spots along this wire, as suggested by Tom, where current might be entering this wire. We have discussed just replacing this wire.  Actually, this might not be too hard to do if we attach a new wire to the old at the switch end and pull the wire thru the instrument panel. If it is not possible to pull a new wire thru, a new wire could be run thru to the solenoid. There are non-functional "clipped" wires in my CT and I do not like finding them.  I have asked my mechanic to remove these when he finds them.  If removing the old "start" wire on the CT being worked on is not possible, due to the complexities of the wire bundles, I suppose there is the option to leave the old in place and make this non-functional.  As Roger suggests, we have also considered possible bleed thru of current from the 12v. battery "in" of the solenoid to the "start/soft start" blade connection on the solenoid. We have not found any current bleeding thru here. We have discussed either replacing the solenoid and have also discussed not using the solenoid to supply 12v. to the soft start and instead providing for a completely isolated 12v. current to the soft start circuit. This would insure that the soft start would only receive a current from a protected source. For the present time though, Roger's suggestion to replace the solenoid is the easiest to do.  Thanks to all who  continue to provide suggestions.  We all have "skin in the game" when it comes to  learning how to safely maintain our Rotax engine  Most of us understand that although this problem didn't involve your plane or mine, it could in the future. 

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...