Rotax 912iS

[skeptic36]

Here is the story of Glade Montgomery collecting his new 912 iS powered trike from Florida and flying it home to Washington.

 

Of particular note are hot starting problems, the amazing airport facilities compared to ours and the fuel economy. You will have to skip over a lot of congratultory replies to his posts etc. http://www.trikepilot.com/members/profile/3231/blog-view/blog_831.html

 

Also here is the DeLorme track of his flight https://share.delorme.com/GladeRoss

 

Regards Bill

 

 

[Tex]

Good read, the warm start issue is quite concerning!!! Given they are putting them in some motor gliders and you really want that restart to perform with out any issue everytime. The injected engine should really be free of issues like this... one of the major benefits of going injected, I would have thought, is good starts!

 

 

[skeptic36]

Good read, the warm start issue is quite concerning!!! Given they are putting them in some motor gliders and you really want that restart to perform with out any issue everytime. The injected engine should really be free of issues like this... one of the major benefits of going injected, I would have thought, is good starts!

May not be the motor, it's the first installation they've done, also he had a lot of current draw with his heated clothing and gadgets.

Regards Bill

 

 

[Yenn]

In GA flying there is a well known problem with injected starts. It seems to be a black art that some have conquered and others are struggling with

 

 

[Guest Maj Millard]

Great read, some hard miles flown there, and some good insight into the new 912is....thanks sceptic..Maj...

 

 

[Head in the clouds]

Yes Yenn. Injected aviation engines are notoriously hard to start. Get to know the quirks of a particular one and you might have some chance. This used to be one of the bugbears of the larger piston engine helis in particular, due to their less effective fan cooling, compared with prop-blast cooling.

 

However - I always understood that it was a mechanical injection issue, shouldn't EFI have solved that?

 

 

[Tex]

Yes Yenn. Injected aviation engines are notoriously hard to start. Get to know the quirks of a particular one and you might have some chance. This used to be one of the bugbears of the larger piston engine helis in particular, due to their less effective fan cooling, compared with prop-blast cooling.However - I always understood that it was a mechanical injection issue, shouldn't EFI have solved that?

Exactly... EFI. I thought the ECU works out what is required and does it. Could be wrong... I note he had some success there once by pumping the throttle

 

 

[Graham Pukallus]

I too had problems initially, but found start proceedures on GA injected engines especialy the Bonanza, was to prime for 7 seconds with pump, switch off pump - go for start and slowly advance throttle, worked a treat for me. Hot or Cold.

 

 

[Head in the clouds]

I note he had some success there once by pumping the throttle

Which is odd isn't it? Pumping the throttle goes back to standard carburetors which have an accelerator pump to squirt a small stream of fuel direct into the manifold to slightly enrich the mixture as you increase throttle. But with EFI there is no accelerator pump, so in theory pumping the throttle shouldn't do anything at all.

 

I too had problems initially, but found start proceedures on GA injected engines especialy the Bonanza, was to prime for 7 seconds with pump, switch off pump - go for start and slowly advance throttle, worked a treat for me. Hot or Cold.

Yes, they all have their oddities. I had a Kawasaki KH4 with (IIRC) a turbo IO-540 and if it was still hot from a previous flight the quickest way to get it started was go and have a cup of coffee and let it cool down. If, instead, you tried to start it hot you'd fail and then have to wait until at least after lunch and also go and get the booster leads...

 

 

[fly_tornado]

Being a euro developed engine you assume that they are running the engine as lean as possible to keep the emissions low, too lean and it won't start.

 

 

[poteroo]

I too had problems initially, but found start proceedures on GA injected engines especialy the Bonanza, was to prime for 7 seconds with pump, switch off pump - go for start and slowly advance throttle, worked a treat for me. Hot or Cold.

That works with a Continental fuel injected motor - but if you do that to a Lycoming injected engine............. plan for a cuppa or two before you try again.

 

I started on Lycoming IO-540's in 1967, in the tropics, and the hot rule was : NO pump, throttle open approx 20-30%, mixture ICO - then crank it over and slowly advance the mixture as it catches. Then with your 3rd hand - retard the throttle so you don't overcome the brakes. One trap is that rookies keep going in too fast with the mixture as the engine begins to catch, and this ensures lots of coughing, black smoke and it often stops. The rule as I understand it is: starve a Lycoming - feed a Continental.

 

As to the Rotax 912iS - doesn't seem anyway a pilot is going to change events if there's an electronic ICU in the system. One for the electronics people?

 

happy days,

 

 

[Spriteah]

Yep. The YBLT Cessna is 180hp injected. Same idea. Run pump, look for bounce of needle (2 seconds). Crank, as she fires advance mixture to rich, then some throttle. Im tipping if the fan stops at a thousand feet your going to muddle that up!!!!

 

Jim.

 

 

[facthunter]

Those GA systems are what I call"dribble" systems and are pretty crude in the priming phase. You can flood one and light a fire under the front easily or "cook' a starter motor anytime. I think they would be less of a problem in flight. If they are feathered they will turn over usually by the action of the prop, once you crank it a bit to start the unfeathering. If you develop a system that works well stick to it.

 

A digital system should fire up straight away, if there is enough volts left after the starter takes some. You can usually leave them for weeks and they fire easily. Poor hot starts? Something wrong with the sensors, or system logic? Nev

 

 

[fly_tornado]

Does anyone know if the 912iS has a 4 wire o2 sensor?

 

 

[Tex]

Which is odd isn't it?

Yes, they all have their oddities. I had a Kawasaki KH4 with (IIRC)...

Very odd. As above must be electronic... I note they begin start relay on the 35a line then switch to the 15a after certain RPM. Maybe it is staying in the 15a... My skill with electronics is breaking them, even if I haven't touched them, so it beats me.

 

The Kh4 is a great helicopter... I love their look and seating config. Wish there was one around here to fly.

 

 

[Guest sunfish]

I'm planning on a 912 iS and will visit Floods shortly. Airmaster Constant speed prop as well.

 

All information I have from users indicates that it is a honey of an engine and it's being tested in a really demanding environment (Ghana) with poor fuel, hot conditions, humidity, etc. Cruise fuel flows of 12 l/hr have been mentioned.

 

The engine has no Oxygen sensor but it has EGT probes on each cylinder so it seems to me that the Rockwell Collins ECU does some pretty smart leaning.

 

The starting technique for an EFI engine has nothing to do with the old low pressure GA engine techniques, it should fire up first time every time.

 

His hot starting problems sounds to me like a peculiarity of his installation, possibly vapour lock. There is a technical suggestion out there that the engine needs a fuel rail bleed jet and return as specified by Rotax (but often not fitted) in 912 ULS engine installations. That allows the fuel pump to purge any vapour or bubbles very quickly.

 

Alternatively, maybe he has a voltage problem, but that would affect cold starts too.

 

 

[Downunder]

EFI engines run return lines, I thought ? Shouldn't get vapour lock with that?

 

 

[fly_tornado]

EFI engines generally don't run 02 sensors above 1/2 throttle. the CHT, intake air temp and MAP sensors tell the engine most of what it needs to know to set the correct fuelling.

 

 

[bexrbetter]

I can only presume that the Rotax is a very low pressure system to avoid fire risk and therefore can suffer vapourising.

 

Hot or cold start problems are virtually unheard of in modern automotive EFI systems but if you ever saw just how far and fast they spray fuel you might be concerned having those pressures in a plane.

 

 

[Guest sunfish]

Read the rest and I was wrong. Sounds like an electrical or software issue.

 

 

[rankamateur]

if you ever saw just how far and fast they spray fuel you might be concerned having those pressures in a plane.

Isn't the day that the pressure matters, the same day that you flick the fuel master tap in the cabin, and starve the problem?

 

 

[skeptic36]

May not be the motor, it's the first installation they've done, also he had a lot of current draw with his heated clothing and gadgets.Regards Bill

They have found the problem, below is written by Glade Montgomery:

This is somewhat of a postscript.

 

It was not a quick process, and (since my Revo is now diagonally clear across the country from its manufacturer) it definitely required a bit of cooperative effort on my part. Regardless, we finally found a solution on the 912iS warm-start problem that developed about mid-way through my big cross-country trip (you may read above for details).

 

First, it turns out Rotax was purely innocent. The problem was outside their product.

 

More specifically, it was in a spade connector on a relay within the outside wiring.

 

The fuel-injection engine has two fuel pumps (redundancy in virtually everything). Until the engine is actually running, a different and separate circuit is required to keep those fuel pumps powered. And, it's not like in a carburated engine where a fuel pump might merely assist. For fuel injection, at least one fuel-pump running is essential. Anyhow, this separate circuit is engaged via a little relay. The symptom I had was that, when attempting to start warm, within fractions of a second of the cranking action beginning, both fuel pumps would shut off (there are separate lights for each fuel pump, which makes the shut-off rather obvious). With both fuel pumps off, further cranking was futile.

 

It turns out it's this relay that's involved in keeping the fuel pumps powered pending engine start, and one of the wires connecting to this relay, via spade connector, was not tight. The female-end of the spade connector was just a little slack. The fix, once this was discovered, was as simple as pulling off the female end, squishing it a little tighter with pliers (so the parts that ultimately clamp on the spade are nearer together), then re-inserting the spade (don't think anything sexual here, please). I tested afterward, and all was perfect. I am happy.

 

Evolution Aircraft contracts with Powrachute in Michigan for direct manufacture of the Revo. This company does superb work. With a tremendously complex aircraft, this is the one (and only), directly definable and operational defect I have encountered. It appears. when someone at Powrachute attached the seven spade connectors to this particular relay, that person failed to notice this particular female end was not really and truly tight when pushed upon its adjoining spade (again, no sex please).

 

BTW, this was the first time Powrachute ever built a Revo with a 912iS, which encompasses monumentally more external wiring as compared to the standard 912s.

 

Though the consequence was frustrating, I have to tell you I am really impressed that this is the "worst" problem I've encountered, from the very, very fine work done on my Revo.

 

I salute Powrachute for their excellent work, for the pride they obviously take in their work, and I further and even more strongly salute Larry Mednick, his dad Phil, and the whole crew at Evolution Aircraft for the intense dedication they have to this sport. I could hardly be happier. They afforded me a great adventure, along with further and continuing sublime enjoyment in my incredible aircraft.

 

 

[bexrbetter]

Isn't the day that the pressure matters, the same day that you flick the fuel master tap in the cabin, and starve the problem?

Well having actually seen how much and how far they spray, by the time you consider it and do it you already have a very serious problem. Some systems do come in a double pipe setup, outer sheath for safety.

 

Low pressure systems are fine it's just that none of the newer, lighter engines have low pressure systems so your back to square one, ie; older heavy, and not very efficient engines.

 

 

[skeptic36]

More on his "warm start" problem:

 

As most of you know, my awesome new Revo was built with Rotax’s latest and greatest: the 912iS fuel-injection engine. It’s an incredible power plant. Purrs like a kitten, and my fuel economy is off the charts. Regardless, there have been some kinks to work out. Of course, this is not unexpected, and the folks at Evolution have been entirely dedicated in their commitment to assist.

 

Regardless, there is a party with whom I am rather unhappy. It’s the North American Rotax distributor, Kodiak.

 

Let’s first take the warm-start problem. This was real-time described in my big-trip blog:

 

http://www.trikepilot.com/members/profile/3231/blog-view/blog_831.html

 

To summarize, the problem began about mid-way through my cross-continental trip. If I stopped for fuel, I could not startup (and resume flying) for at least an hour. If I started first thing in the morning, let it warm slightly, then had reason to momentarily shut down before actual takeoff, I found myself needlessly stuck on the ground for at least an added hour. When I got home, I found my flying here likewise compromised. Stop for a bathroom break. Stop to exchange passengers. Whatever! No go again for at least an hour.

 

It was extremely frustrating. A plethora of theories were successively proposed, investigated, and dropped. Lacking their own 912iS expertise, folks at Evolution (in particular, Larry’s dad Phil) worked hard to pump information and guidance from Kodiak, which was somewhat painstakingly passed through to me. By such means, I was directed to try a series of expedients. Diagnostics. Changing components. Putting in new wiring and novel components.

 

Nothing worked. Making it all the more frustrating, it took a lot of time even to determine if an expedient had worked, because – to tell – the engine had to first be well-warmed, then allowed to cool a bit, though not entirely (turns out calling it a “warm-start” problem is not perfectly descriptive: it always started well if fully warm, and always started well if fully cold; it was the in-between state that was a problem).

 

This went on for a considerable time. I grew increasingly aggravated with each continued failure.

 

Somewhat early on, it was divulged Evolution had used an alternate external wiring plan, for the engine, which it is my understanding had been recommended (and evidently was created) by folks at Kodiak. I am also given to understand the same recommendation was made to (and evidently accepted by) several other North American aircraft manufacturers.

 

Here’s the thing. The standard Rotax 912iS external wiring plan calls for two momentary start switches, which must be simultaneously held during engine start. So near as I can gather, folks at Kodiak evidently guessed American pilots would find that objectionable, and so came up with a plan that uses an external relay in an alternate circuit design, in effort to eliminate need for the secondary momentary switch.

 

When I was given this understanding, the objective sounded sensible, and much of my work (as directed by folks at Kodiak and translated through Phil at Evolution) centered around seeking to make that alternate plan function as needed.

 

The essence of the fault (at least beyond the base level), we duly determined, was that the system was shutting off the fuel pumps after, maybe, 1/3rd a second of engine cranking. When the engine was fully cold or fully warm, this was not an issue -- because, in fact, the engine starts just fine within that short period. Where it’s partially warm, by contrast, a little more cranking time is needed. Since it’s fuel injection, it fully depends on the pumps (or at least one them). No pumping, no injection. Thus, if pumps shut-off before start, you get no start.

 

We fought it, and fought it, and fought it – with ever-increasing frustration on my part. At several junctures I was at my hangar, tools in-hand, while in three-way conference between myself, Phil at Evolution, and someone named Nino with Kodiak in the Bahamas. Eventually, I found myself on the verge of screaming.

 

Finally, during a particular telephone conversation, Phil suggested I look at Rotax’s schematic for the standard 912iS external wiring. You should understand, I am not a schematic genius. However, I do have a bit of background experience (I spent several years as an appliance technician), so looking at a schematic (at least one that’s not extraordinarily complex and/or is reasonably presented) is something I am comfortable with. It’s a skill I’m sure I possess in greater degree than folks at Evolution.

 

Anyway, I examined the standard Rotax-recommended schematic, and compared to the alternate plan that evidently was produced and recommended by folks at Kodiak. With no more than 30 minutes analysis, it was thoroughly, totally and absolutely apparent there was no way (may I say “No way in hell?”) the alternate plan could achieve the engine inputs Rotax intended.

 

Indeed, I could not believe how thoroughly (yes, I will say it) STUPID is that alternate plan. I do not know who precisely created it. I do not know of a certainty it was someone at Kodiak. I am, regardless, extremely certain the person that did it could not possibly have been operating, when doing so, with appropriate intelligence, care or due regard for consequences.

 

There are at least four problems with the alternate plan:

 

1. Engine cranking begins with what are supposed to be emergency-mode engine inputs, as opposed to proper cranking-mode inputs.

 

2. Within a brief moment after cranking begins, engine inputs switch (via that added relay) to half (and only half) of what is supposed to be engine-start mode. Because in this “half” mode the negative-side engine input is missing, it provides a direct and precise explanation for the exact symptom (pumps shutting off) we were for so long fighting.

 

3. Engine inputs remain in this half-engine-start mode for the entire duration of the engine-run cycle. In other words, inputs do not appropriately switch to intended engine-run mode. One visible consequence of this (I did not realize it was not an intended feature in the interface until studying the official schematic) is the dashboard pump lights remained constantly illuminated during flight.

 

4. If both engine generators happened to fail, Rotax’s official schematic has provision for manually switching to emergency-mode battery-powering of the fuel pumps (or, again, at least one of them). The alternate plan has no provision for such manual switching. It might, or might not, achieve needed switching automatically, depending on circumstances.

 

The actual schematics are simple enough (they really are quite simple) I was able to deduce all of this in the space of about 30 minutes. Again, this is in the absence of me being any kind of schematic genius. I instantly felt very appalled (alright, beyond appalled) a supposedly responsible organization and its personnel had evidently foisted such an irresponsible and incompetent product-- as this alternate wiring plan -- upon the flying public.

 

At any rate, upon realizing this, solution was obvious. It was to rip out all components and wiring as uniquely associated with that (my favorite word for it) “jackass” alternate plan, and replace with wiring and componentry appropriately matching Rotax specs.

 

It did not take much of a genius to figure that, did it?

 

Now, do you want to hear the real surprise?

 

No, it’s not a surprise at all.

 

Of course, and obviously, once wired according to genuine Rotax specs, the pumps ceased to shut-off after 1/3rd a second of cranking. Indeed, they continue pumping – for as long as it takes – to start. Hence, my warm-start problem was totally and thoroughly solved, instantly.

 

Well, almost instantly. It killed a long day of rather hard work to do the entire re-configuration, from jackass-alternate plan to Rotax-specified.

 

I also no longer have bright-green fuel-pump-indicator lights constantly (and annoyingly) illuminated during the entire time I fly.

 

And, I am fully confident that, in the extremely unlikely event both engine generators fail, I can easily (and correctly) switch to battery-pump-run backup mode.

 

Ahhh. Doing it right. It’s not bad thing.

 

In defense of the great Evolution folk, they were relying on Rotax’s duly-appointed North American representative. There is no fault on Evolution’s part. Larry and Phil were terrific.

 

I wish what I just described had been my only problem.

 

Beginning shortly before I left Zephyrhills, one of the Rotax “Lane” lights began to occasionally go into a flashing mode. This is the equivalent of an engine warning light on a car. Since the 912iS has dual/redundant ECUs (referred to as “Lanes”), there are two separate lights, one for each Lane. Anyway, it got worse and worse. By the time I completed my cross-continental trip, Lane A was flashing continuously. The word from Kodiak (again, relayed to me via Evolution) was that it should not be a cause for any real concern. Most 912iSs were doing it, and it was thought to be some kink that needed to be worked out in the monitoring system.

 

I guess that kind of assurance is good enough to get you on your way for your big delivery trip home, but it’s not a matter you exactly want to drop. You expect the matter to be fully investigated, results to be disclosed, and a solution provided.

 

To that end, even before leaving Zephyrhills (that was early February) I employed tools as provided by the Evolution team to extract diagnostic log data from my 912iS system, and emailed it to Nino at Kodiak, anticipating he would analyze the data to deduce specifically what was behind my flashing Lane A. Nothing happened.

 

There was still no report upon completing my big trip, and, in the weeks following, I repeatedly, over and over again, asked for the analysis. Just as many times, I was awarded with zero response.

 

Eventually I was asked to provide another set of logs. Phil, from Evolution, overnighted me the log-extraction dongle. I invested the time and effort to again extract, and again email. Again, no response. Again repeatedly I asked, and again (just as many times) there was nothing.

 

A third time I was asked to re-provide logs, and a third time I did so. Again, no response. No analysis. No evidence Nino – or anyone else at Kodiak – had even looked at them.

 

I finally got irate. Just earlier this week, good ol’ Larry passed on, to Kodiak, a sense of my impatience. Finally, a bit of analysis was done. It does not give us a real and full answer yet, but at least it’s something.

 

Does it sound as though I’m pretty disgusted with folks at Kodiak? Yeah. It’s the understatement of the year.

 

One other problem (this one surfaced before I even saw my new Revo) is the 912iS hesitates on sudden application of throttle. Larry observed this in early testing, informed me, and sought solution via Kodiak. What he passed on to me is that a solution is expected via a firmware update, scheduled for July.

 

In all, I am truly delighted with my Revo. Larry's new Rival wing design is positively sweet. I want to repeat, one-up, I am able to get 85 mph at 2.5 gph. Handling is crisp, precise, light and immediate. The wing is also relatively short, and fits in my hagar more easily. I love it.