Rotax 912 side-effect while chasing radio problem

[IBob]

As well as the preflight mag check, I am in the habit of running another mag check just before shutdown. This is something I got from another member here and the aim is to be confident that both ignition circuits are still working. (It's not a very good check, as the modules will be warm, but that's another issue).

I normally do this post flight check at a lower speed of 3000RPM (instead of 4000RPM), and while my engine always ran on, it was noticeably rough during the mag check at this lower speed. I never thought much of it: I know the engine requires both ignitions for optimum firing and thought this just a symptom of less than optimum firing.

 

Recently, however, I was chasing a minor radio issue I have (interference, but only at WOT) and started at the plugs and leads with the intention of working back. On inspecting the leads to the lower plugs, I found signs of rubbing where they pass down between the heads. There was no damage to the leads or outer sleeve, but it was quite obvious that they sometimes rubbed or vibrated against the metal of the heads : hardly surprising, as they are unsupported where they pass down through this narrow gap.

In order to eliminate this as a possible cause of interference, I made a couple of square rubber grommets to guide the leads between the heads. I also inspected and/or replaced the cable ties on all plug leads at the heads, with the aim of keeping the leads away from the metal of the heads.

 

The result of this is that (while I still have my WOT radio interference) my mag check at 3000RPM is no longer rough. So either removing and refitting all the plug caps has improved a connection (but they were all on tight when I removed them) or carefully routing the leads away from the metalwork has improved the ignition at this slower RPM.

[skippydiesel]

Just guessing but you may have a damaged high tension (HT) lead.

 

By securing the lead(s) in place with the home made "square" grommet you may have reduced/eliminated arcing at a break.

 

Just a thought!

Edited May 25, 2021 by skippydiesel

[IBob]

That's certainly a possibility, Skippy.

Unfortunately, since what I was trying to locate (radio interference cause) was something quite different to what I now have (smoother 'mag' check at 3000RPM) I wasn't focused on the latter.

So I can't be absolutely sure now, but it seemed to me that rough running occurred with either 'mag'. Had it been notably one 'mag', I'd have gone looking for the cause.

[RFguy]

I wonder if insulation on plug leads is not what it used to be. 

it's a fair voltage to isolate. plug gap in air of 0.8mm would require about 3 to 4 kV in air,

but pressure increases breakdown voltages and temperature reduces it :.

CC at room temp, breakdown ~ 24kV. If the CC was at 500k, the breakdown will be something over 12kV.

 

In air, should be able to jump at least 5mm (non pointy electrodes), pref 6 or 7.

so, as you can see, plenty of reserve to track across objects or breakdown the air.

 

More likely to track I think. (track- make way along something). which is why the porecelan insulator is long.

Build up of burned carbon on the plug lead  insulators would be a issue, it would just track on the carbon like it was copper....

 

 

 

[RFguy]

Something zI want to do Ibob is a monitor for ignition system. IE just something to remove some of the guesswork in case of some problem. 

 

Needs for me, 12 HV clamp probes likely 3D print something up. Or possibly print a single round plate that can sense all the the leads as they exit the dizzy.  You can see the difference between no, low, med, high spark, and also, spark, but not getting to the plug spark.  if the probes have some resolution, you can pick compression and rough mixture.

 

Using a small microcontroller , it would be trivial to count sparks, so if they were going missing every now and then, they could be caught. Also you'd pulse extend if a  plug wasnt firing  so that a single misfire could be caught on a telltale LED... etc etc probably just 1 single chip for the whole thing. the devil is in the mechanical design for the probes. 

[IBob]

Interesting idea, RFguy, that's for sure.

 

My 912 has run faultlessly since new, so I'm not actually looking for any sort of ignition problems. I'm just a bit surprised, now I'm focused on it, that there is nothing much to prevent the leads to the lower plugs from making contact with the metal of the cylinder heads. As you can see here, they are threaded through some sort of (heatproof?) white sheathing that just floats between the heads.

All I did (at this stage) was pull off all the plug caps....and they were all tight.....put them back on, make and fit a couple of square grommets to hold the lower leads centred where they go down between the cylinders, and replace/adjust some cable ties on both upper and lower leads to try and ensure no contact with engine metal. (These are older pics, so not shown here).

Edited May 25, 2021 by IBob

[skippydiesel]

IBob - congratulations on what looks to be a very nice/clean set up.

 

Questions:

• Its a bit hard to see but why such a big oil cooler?
• It looks like you have used some form of "corrugated" metal hose for part of the radiator cooling system - please  give your rational ? product name? supplier?
• Whats with the cut down muffler ?

 

[IBob]

Skippy, what you see there is (or was) all standard ICP Savannah issue.

 

The lower black radiator is coolant. The upper silver radiator is oil. Both these radiators hold the temps within specs within the ICP cowl design. I do blank off part of the oil radiator in winter, but all the radiator is required in summer.

 

The metal hose also came with the kit. It's used for oil, coolant and parts of the fuel system. I know there is some debate about it, but I figured if it was failing in use we'd have heard about it. The downside (as used here) is that it requires a short piece of 'rubber' hose and two hose clips to terminate it. The upside is that it is very compact and can be hand formed into quite tight bends without any flattening or collapsing. At the engine, it allowed a very compact installation, especially for the larger 'hoses': I had heard the cowls were a tight fit, but this wasn't the case at all.

 

The muffler is also ICP. It's compact and does the job, but various others have suggested it is a poor design, potentially reducing engine power. Also the muffler and downpipes don't fit together as neatly as they might, and a very competent LAME who took a look at it here made some scathing comments about the setup. I believe ICP may be working on a new design.

 

 

[IBob]

As for the installation, I went through a heap of cable ties, running and rerunning stuff until I was happy with how it all sits. It took a while.

A lot of where to run the various bits, and just how to secure them at various points against vibration and chafing, came from studying pics provided by the (Australian) agent who supplied the kit.

I found these invaluable in all sorts of ways as the build proceeded. I would encourage anyone building to source as many pics as they can to reference this sort of thing. In my case, I had an old laptop in the workshop, allowing me to view the agent's pics, also access pics on the site and other places.

[skippydiesel]

Thanks for all that IBob.

 

For the same engine (Rotax 912 ULS) I find it quite intriguing that there are so many variations on installation/fit out eg your oil cooler looks to be almost three times the size of the one fitted to my aircraft. Yours is in a forward position directly in the air blast from the prop. Mine is mounted low on the firewall/engine bulkhead, in the exit airflow from the cowling - works perfectly. Your coolant radiator looks to be the same size as mine but is mounted at an extreme angle to what I would imagine the air flow (in /out) to be.

[RFguy]

On three rotax installs I know, and one I know really well, the stock oil cooler is insufficient for aussie hot days. (OAT 38+, long climbs MaxAUW ) 

The cooling ability becomes very sensitive to cowling design and neg pressure with the marginal cooler.

Edited May 25, 2021 by RFguy

[IBob]

Skippy, I can't say I'd thought much about radiator sizes, but I guess it's a matter of what works within some constraints of where you can put it?

While I'm not privy to the ICP design process, at a guess I would say that when they remodelled the cockpit and nose and went from motor bed to ring mount (for the XL which went on to become the S) they probably sketched up a cowl, then worked out how best to pack in the necessary bits.

And it's come together pretty neatly, including the airbox.

The coolant overflow bottle is tucked way down in the back, but there's an easy workaround for that (thread fine neoprene tube down the bulkier overflow tube to add or remove coolant).

As mentioned, there are questions in some quarters around the muffler, but it works okay on the large number of builds since then.

[skippydiesel]

12 hours ago, RFguy said:

On three rotax installs I know, and one I know really well, the stock oil cooler is insufficient for aussie hot days. (OAT 38+, long climbs MaxAUW ) 

The cooling ability becomes very sensitive to cowling design and neg pressure with the marginal cooler.

Well its part of the pilots responsibility to manage the impact of weather on his/her aircraft .

 

When hot weather (above 30C) is forecast, I arrange to depart at first light.

For my personal comfort, I prefer to land every 2 hrs (max 3).

So the second takeoff is still in the cool of the morning.

The last landing in potentially 40+C is perfectly doable and by then (somewhere between 10:30 & 13:00,) I am due for  bit of a feed, a drink and a walk.

If another departure is planed/desirable, a short hop after 16:00 hrs usually sees temperatures down again. If not,  leave the next morning.

I rarely fly more than 4-5 hrs in a day - 6 hrs is big for me - so rather than fight the elements (bigger oil /coolant radiators) I just work around it - seems to pan out rather well.

[RFguy]

Like I said, I suspect the rotax cooling for OAT > 38 is very sensitive to cowling/neg pressure performance and design.

I spent most of summer flying the rotax around in OAT > 35 in the middle of the day. required management. alot of it. 

So, nothing a HUGE oil cooler doesnt fix. 

If I put a 914 in my Jab, it will have a HUGE oil cooler . and blinds so I can get reasonable warm up times in winter. 

 

IBob , what's your warm up time like, annoying or fast ?

[IBob]

RFGuy it's pretty slow in cooler weather. Not that it's something that annoys me.

 

I do have an oil thermostat I bought from SA a while back, but have never got round to fitting it. I would think it would make a difference, though I also remember reading some comments a while back (on this site, perhaps) saying it didn't make much difference, and speculating that the oil tank was also acting as a radiator.

 

I saw under the hood of a factory build Savannah recently and it had an oil thermostat. I don't know if they are now part of the package, or an optional extra.

[RFguy]

i can tell ya it's annoying when you are renting it ! 🙂

 

where is it fitted in the circuit ? gather bypass oil cooler of course. 

Wonder if you cold kiwis adopt a automotive style heating from rotax engine coolant ?

[IBob]

RFGuy, in the 912, oil is drawn from the tank, through the radiator, to the pump. (The oil temperature sensor is located at the pump.)
From the pump, the oil is passed to the necessary places in the engine.
The oil then pools in the bottom of the engine, and is pushed bank to the tank by crankcase pressure (blow by).

 

An oil thermostat bypasses the radiator when cold, so that (some) oil is drawn directly from the tank to the pump, and on to the engine. In installations I have seen, it sits neatly in the upper area behind the gearbox.

[RFguy]

Good solution. I am fan of dry sumps. Did a dry sump conversion on a race car engine years ago. Although we used scavenge pumps.

[IBob]

RFGuy there is a bit of extra pipework involved in fitting an oil thermostat: it's (obviously) not like a car coolant thermostat that just sits in circuit and throttles flow until temperature comes up.
I saw Mark Kyle comment here earlier that he had a coolant thermostat fitted to his 912. It would be interesting to know what results that gave.

[skippydiesel]

I love "fiddling" with engines/mechanical things and am not against oil/coolant thermostat's, cowl flaps/louvers and bigger radiators  etc  BUT you are talking about an aircraft here - Everything you contemplate should have a Cost Benefit Analysis done on it.

 

In this context Cost = weight & complexity and Benefit = the hoped for improvement in performance.

 

In general KISS is the way to go - the more complex a system, the more potential for something to go wrong.

 

Sure we all get satisfaction from fitting that extra gizmo but sometimes the satisfaction is the only real benefit.

[RFguy]

Surely Sean you should add to that comment, "In my opinion, .... "

I am trying to do that more often, also.

[IBob]

Skippy, I agree with you entirely when it comes to KISS: I made a living (though sadly not a fortune) specing and programming deliberately low tech high reliability automation.

And it's part of why I didn't fit the oil thermostat yet.
The other part is that I'm pretty sure if the unit failed, it would fail closed (as it's operated by some sort of capsule containing a wax that expans when heated). Which is the opposite of what you'd want.

It's a Mocal OT/1-92 automotive unit.

Next time the lid is off the factory Savannah here, I'll get details of the unit the Italians are fitting.

[skippydiesel]

39 minutes ago, RFguy said:

Surely Sean you should add to that comment, "In my opinion, .... "

I

Agreed !

[skippydiesel]

31 minutes ago, IBob said:

Skippy, I agree with you entirely when it comes to KISS: I made a living (though sadly not a fortune) specing and programming deliberately low tech high reliability automation.

And it's part of why I didn't fit the oil thermostat yet.
The other part is that I'm pretty sure if the unit failed, it would fail closed (as it's operated by some sort of capsule containing a wax that expans when heated). Which is the opposite of what you'd want.

It's a Mocal OT/1-92 automotive unit.

Next time the lid is off the factory Savannah here, I'll get details of the unit the Italians are fitting.

I have often contemplated installing an oil thermostat  -  I come out of my "day dream" when I remember that they probably have the most benefit in really cold climates  - this is not what we have in most of Australia. I do fit a fixed winter cowl exit flap/reducer in winter & remove it when day temps are consistently trending  north of  25C 

 

At one stagger I designed and fabricated  a cowl flap (opened/closed by linear actuator) - had great fun - never fitted it.

 

The truth is, I usually realises at some stage in my fantasise,  that I now fly RAA and the beauty of these aircraft is their light weight and relative simplicity - why spoil a good thing?

[IBob]

Skippy, the traditional fix here seems to be tape across the face of the radiator...though I did see a S Island build recently where they looked to be incorporating the oil cooling into the coolant circuit with some sort of plate heat exchanger, so requiring only one radiator.

 

My preference (at this stage) would be for a ground-adjustable blind on the radiator that I can set without lifting the cowl. The trick would be coming up with something secure, durable and rattle-free. I haven't given the design much thought at this stage............