Rotax 912 Ignition Problems.

[Kyle Communications]

Dan I know you have done a lot of checks as per emails. M61A1  Mick has posted his issue...this is more than likely similar to yours. This would certainly give you the symptoms you are seeing. Do the test with the multimeter on AC volts on the red wires this will confirm you are getting power or not and you dont need to pull the engine out

 

[waraton]

1 hour ago, Blueadventures said:

Bert Flood will sell and about $280 plus postage.

I only need the plug bit!!!!

[dan tonner]

WOW!  I think we're ready to ask NetFlix or Prime Video to produce series for us!   So much to talk about.  So little time.    😃

 

Bob - your last post describes the perfect scenario for instantaneous failure in both ignition systems - shorting across the two charging coil leads - the exact reason for the Rotax SB 912-026 in December of 1999.    My first "test" at the hangar will be a close look inside the wire mesh sheath over the charging coil wires ...followed by checking the resistance between the connectors of the two "red" leads.   And, if the stator were never replaced, my engine has had 23 years reach the point of collapse.

 

Then I'll look for a squirrel in the exhaust pipe.

 

Mark: if it weren't for Wikipedia I'd still be hung up on CRO    but I'm still trying to get my head around:     the square Root of the Mean of the Square.  I get over my head quickly with electronics.  Fortunately, I had no problem with the chart of multimeter AC Voltage vs RPM from  Blueadventures - which was essentially confirmed by you. 

 

Thanks to all...keep safe,  fly safely.    I'll post my findings.     (JG - this was your thread......still reading?)

 

CanadaDan

 

[JG3]

Yes still watching. Want to hear the result of this mystery.....

[Kyle Communications]

Dan the quickest and easiest way to confirm there is something wrong with the charge coils for the CDI or a issue with shorting inside the harness is to use a multimeter and do the test for voltage as you spin the engine over with the starter..this will at least give you some better direction to look at

 

 

[dan tonner]

Well. gentlemen; I have a plan for arrival at my hangar sometime this week.

Below is a summary of some of your suggestions and my additions, deletions and thoughts regarding what to expect. 

Should you notice flaws in my reasoning (...my wife tells me from time to time that this is possible...) please jump in.

Once I've completed the "trials", I'll update you all here.

As you will probably notice; I am leaning heavily towards "stator failure" as the culprit with my engine and may be looking for a good used P/N 996 539   or     P/N 888 675.

Stay safe. Fly Safely.

CanadaDan

 

Engine checks in search of the cause(s) of sudden engine stoppage.

 NOTES:  Mag switches were checked for proper operation.

              Fuel system was checked for quality, supply and flow control; carb bowls properly full.

             The result of any test may preclude the need for any number of the other tests.

A)  Measure AC Voltage across each red lead to engine ground at cranking speed. (Leads disconnected from modules.)

·       5 – 7 VAC on each lead:   Specification measurement;  All OK.

 

NOTE::  during B & C tests, I will flex the harness at pinch points. Reading changes will represent failure.

 

B)  Check resistances across each red lead to engine ground:

·       3.2 – 4.5 OHMS on each lead:   Specification measurement;  All OK.

·       0 OHMS:   Short to ground  (failure)

·       INFINITY OHMS: broken lead wire  (failure)

C)  Check resistance across the disconnected  red wire leads from the charging coils:

·       3.2 – 4.5 OHMS:  one lead wire is shorted to ground (failure)

·       INFINITY OHMS:   broken lead wire with no grounding or shorting of leads. (failure)

·       0 OHMS:  shorting of leads or possible grounding of BOTH lead wires. (failure)

·       SMALL RESISTANCE:  (half of 3.2 – 4.5 ohms)  All OK.  (measures resistance from lead through the first coil to ground then back through the other coil and second lead.)

D)  Attempt to start the engine while an assistant sprays fuel into the intake end of the carburetor(s). Any detonation indicates the presence of spark. Explosion and fire indicates the need for insurance.

E)  Ignition on; module(s) connected.  Engage the starter with the top plugs removed, connected and  grounded while an assistant looks for spark at the gaps. Results will also depend upon which modules/sparkplug coils are connected.

F)  Look for rats & squirrels in the exhaust pipe.

 

 

 

[IBob]

Sounds like a plan, Dan.

 

The only thing I would add (if you get no result from your proposed list) is to again measure the charging coil outputs to the ignition modules, with the ignition modules connected.

 

The reason is this:

When disconnected, the charging coils are under no load, they are not delivering any current (aside from the minute amount required by your meter).

When connected, the coils are under load and delivering current as they would in the usual 'live' situation.

 

The best parallel I can think of is of a failing 12V battery:

If you measure the voltage with nothing connected to the battery (no load) you are likely to see 12V and it all looks good.

However, if you put the battery under load, it can no longer deliver sufficient current, and the voltage drops accordingly.

For this reason, the battery test a lot of workshops use (or used to use, I haven't checked for a bit) is momentarily subjecting the battery to a dummy load (usually a big resistor), while measuring to see if the voltage drops.

[IBob]

PS: If you're going to wiggle wires etc, do it in a very measured and progressive manner.

Right now you have a solid fault, which is bigly much  preferable to an intermittent fault.

So if wiggling stuff does make a difference, you need to be very clear which wiggle did it........

[dan tonner]

The promised followup:

First - the summary:  The problem was failed insulation on the two charging wires from the stator to the modules. The insulation over the wires was more like “putty” than “plastic” and at the point where they were zip-tied near the modules, out-of-phase voltage from the two coils seems to have been able to jump the compressed insulation and cancel.  I believe this issue was the reason for SB 912-026 back in 1999 and my engine was never made compliant.

 

Second - the test results:

A)  AC Voltage across each red lead to engine ground at cranking speed (modules disconnected)    5.6 VAC each side.   (when wiggled – VAC fluctuated to 0).

Bob: I also did the test with the modules connected – 4.55 VAC on each side; - as above, when the wires were wiggled, VAC fluctuated to 0.

B)  Resistance across a red lead to engine ground – 3.6 OHMS each side.  Flexing the wires did not seem to make any difference.  (Possibly redirecting the reading across the crossed wires and through either coil to ground?  ie: no change)

C) Resistance across the two disconnected red leads:   6.7 OHMS;   when wiggled, the reading would change to 0 OHMS – indicative of crossed leads.

NOTE:  the 6.7 OHM reading is a measurement down one red lead through its charging coil to ground and then up through the second coil and its red lead.  The resistance with the coils in series is summative:  3.6 + 3.6 = 6.7  + line/connection resistance?      (The resistance would not decrease as I had erroneously speculated.)

D)   I did not perform the "fuel spray" test after I finding the “crossed” leads.  I separated the leads beyond the zip-tied connector.  (no fuel sprayed – no insurance claim 🙂)     I proceeded  to the next test.

E)  I removed the set of plugs served by the one connected module and, using sections of single-strand 12 ga copper house wire, I grounded each sparkplug and reconnected the caps. When the starter was engaged (higher RPM with plugs removed) there was lots of spark, in sequence, at every plug. After connecting the other lead to the module, the second charging coil produced identical results.

F) No rats; no squirrels.   (I didn’t really look….)

 

So, where am I?

     I pushed back the wire mesh sheath and separated the charging wires to a point halfway between the last two clamps on the motor. I cut the wires here and spliced in two lengths of 22 ga red automotive wire and triple shrink-wrapped the soldered connections. I pulled the wire mesh back over the joint and along the wire as far as it would go – adding a  second wire mesh sheath, shrink-wrapped at the joint, to gain a couple of inches of length. I secured the ”new” wire at the second clamp on the motor. The plan was to ensure there would be no movement from the charging coils all the way to the sparkplug coil bracket – especially where the “original” stator wires were still intact. I added two new female bullet  connectors and reassembled the ignition unit.

     Because of heavy rain and muddy apron conditions, I did not roll the plane out of the hangar to attempt a start - and had to travel home (about  200 km away).   I am certain it will start (on the first try) when I get to my hangar - I will only add to this “already-too-long” thread if it does not start.

 

     It is clear I need to either:

·       repair the “original” stator by replacing the red charging wires (and probably the yellow “generator” wires as well) right back to the coils,   or,     

·       replace what I have with a good second-hand “newer-style” stator.

(I am not ready to purchase a new stator for US $1500.00 (plus, plus)...yet...

 

    Thanks everyone for “kicking in” –

Stay safe; Fly safely.

CanadaDan

[skippydiesel]

YEEEEHA!!!!! well done. Persistence and a little help from your friends - great!🙃

 

Word of caution: sounds like you may have increased the number of "mechanical" connections (bullet connectors) - it is always better to minimise mechanical connections as these reprise a potential failure or raised resistance point .

[Kyle Communications]

Glad you found the issue..that sounds like it will be the problem. It couldnt have been anything else

 

[IBob]

Excellent outcome, Dan. And something of an education for most of us too, I would think.

Hopefully you are able to rework the existing stator.

Go well......)

[M61A1]

5 hours ago, dan tonner said:

The promised followup:

First - the summary:  The problem was failed insulation on the two charging wires from the stator to the modules. The insulation over the wires was more like “putty” than “plastic” and at the point where they were zip-tied near the modules, out-of-phase voltage from the two coils seems to have been able to jump the compressed insulation and cancel.  I believe this issue was the reason for SB 912-026 back in 1999 and my engine was never made compliant.

 

Second - the test results:

A)  AC Voltage across each red lead to engine ground at cranking speed (modules disconnected)    5.6 VAC each side.   (when wiggled – VAC fluctuated to 0).

Bob: I also did the test with the modules connected – 4.55 VAC on each side; - as above, when the wires were wiggled, VAC fluctuated to 0.

B)  Resistance across a red lead to engine ground – 3.6 OHMS each side.  Flexing the wires did not seem to make any difference.  (Possibly redirecting the reading across the crossed wires and through either coil to ground?  ie: no change)

C) Resistance across the two disconnected red leads:   6.7 OHMS;   when wiggled, the reading would change to 0 OHMS – indicative of crossed leads.

NOTE:  the 6.7 OHM reading is a measurement down one red lead through its charging coil to ground and then up through the second coil and its red lead.  The resistance with the coils in series is summative:  3.6 + 3.6 = 6.7  + line/connection resistance?      (The resistance would not decrease as I had erroneously speculated.)

D)   I did not perform the "fuel spray" test after I finding the “crossed” leads.  I separated the leads beyond the zip-tied connector.  (no fuel sprayed – no insurance claim 🙂)     I proceeded  to the next test.

E)  I removed the set of plugs served by the one connected module and, using sections of single-strand 12 ga copper house wire, I grounded each sparkplug and reconnected the caps. When the starter was engaged (higher RPM with plugs removed) there was lots of spark, in sequence, at every plug. After connecting the other lead to the module, the second charging coil produced identical results.

F) No rats; no squirrels.   (I didn’t really look….)

 

So, where am I?

     I pushed back the wire mesh sheath and separated the charging wires to a point halfway between the last two clamps on the motor. I cut the wires here and spliced in two lengths of 22 ga red automotive wire and triple shrink-wrapped the soldered connections. I pulled the wire mesh back over the joint and along the wire as far as it would go – adding a  second wire mesh sheath, shrink-wrapped at the joint, to gain a couple of inches of length. I secured the ”new” wire at the second clamp on the motor. The plan was to ensure there would be no movement from the charging coils all the way to the sparkplug coil bracket – especially where the “original” stator wires were still intact. I added two new female bullet  connectors and reassembled the ignition unit.

     Because of heavy rain and muddy apron conditions, I did not roll the plane out of the hangar to attempt a start - and had to travel home (about  200 km away).   I am certain it will start (on the first try) when I get to my hangar - I will only add to this “already-too-long” thread if it does not start.

 

     It is clear I need to either:

·       repair the “original” stator by replacing the red charging wires (and probably the yellow “generator” wires as well) right back to the coils,   or,     

·       replace what I have with a good second-hand “newer-style” stator.

(I am not ready to purchase a new stator for US $1500.00 (plus, plus)...yet...

 

    Thanks everyone for “kicking in” –

Stay safe; Fly safely.

CanadaDan

f you remove the flywheel and stator, you may find the cables in similar condition underneath  to what you found further up.

If you pick away the potting where the cables terminate at the stator end you will find that the leads are soldered into small holes in tabs.

I used silicone insulated high temp cable. I de-soldered the old ones ad re-soldered in new ones, then repotted with a good quality silicone.

I

It's not much more difficult than replacing the stator and it's waaaaaaay cheaper.

Just make sure that you apply fresh heatshrink and braid then ensure correct routing on assembly so as not to be crushed under the stator plate and good clearance on the flywheel.

Edited November 30, 2020 by M61A1