Dafydd Llewellyn

The problem WAS the starter solenoid switch. You don't want to cut power to other than the starter circuit. Yes, I suppose one can use one of those isolators in series with the starter solenoid, and shut it off immediately the engine has started as a standard procedure.

Does the 170 have a header tank? Will the fuel necessarily flow to it if the aircraft is not flown wings level, zero slip or skid? I'm not familiar with the J 170 fuel system. As I said, it was grapevine, not corroborated.

I've heard (grapevine, not corroborated) that this accident, whilst not total fuel exhaustion, may have been fuel run-out on one tank - i.e. fuel mis-management. Does anybody know of any findings being released?

You mean, a "first-life" engine. But NOT the el-cheapo 0-320-H with the ball-stud type valve rockers, that Lycoming made specially for Cessna. An 0-320-E, yes.

The limiter rating needs to be less than the current rating of the starter leads (look up FAA AC 43.13-1 Chapter 11 for the current ratings). If you look at the copper cross-section of a 150 amp limiter, it's a lot smaller than any starter cable - less than 10%. The impact current when the starter contactor first closes is several hundred amps, but it only lasts for a fraction of a second, which the limiter can handle. I don't know whether it will protect the starter motor in all circumstances, but it will prevent the starter cable from setting fire to the aircraft. The trick is to find the smallest limiter that can withstand the normal starter usage. In my case, naturally I've never had another problem since I started using them, so I don't have a series of experiences to draw on. The only sign of the problem in the Cherokee whilst my wife was flying, was that the fuel quantity indicators etc all went to zero. It wasn't until after they landed that the burned small from the battery box became evident. The whole exercise cost over $2000 (new starter, new ring gear), all for the sake of a lousy $15 starter contactor.

I had this on my PA 28; it destroyed the starter and melted a terminal off the battery (whilst my wife was flying circuits in the aircraft with an instructor). Since then, I have installed a current limiter in the starter circuit; it's essentially a heavy-duty fuse - I've found 150 amp rating to be sufficient for starting a Lycoming 0-320; 100 amps would probably suffice for a Rotax or Jabiru engine. You won't know it's happening in time to use a manual switch. see http://www.larpro.com.au/larpro/product/4091 Airworthiness approval needed to fit this, of course. It goes as close to the battery as possible.

Don, CAO 95.55.1.5 is a sub-set of CASR 21.191(g) - which is the "51% rule" portion of the experimental certificate regulations. So RAA does have a de-facto experimental class; however since RAA does not use the certificate of airworthiness process, these aircraft do not get experimental certificates, as do VH-registered equivalent aircraft.

Do, please, look at a typical design standard before you decide how it should be re-designed. Here's one; it's an extract from CS-VLA. they are all pretty similar. In my lexicon, SCAT is a four-letter word. INDUCTION SYSTEM.doc INDUCTION SYSTEM.doc INDUCTION SYSTEM.doc

Yes, looking at possibilities now.

Yes, multi-point EFI sidesteps all this difficult airflow business; but the certification difficulty of it means it won't happen quickly, I'm afraid. So I think some attempt at sorting out the airflow problems is a necessity until the EFI can be done. After all, there are people on this thread who want lots of development testing before the product goes on the market; that's surely true of any EFI solution. I like the idea of using the existing carbie as the throttle-body for an EFI system; that reduces the "hidden function" issues - one could use it to start the motor, thus testing it at every engine start. I do suspect that the root cause lies largely in the existing airbox. May know more in a week or two; guessing gets one only so far.

Thanks, Geoff; I need to digest all this info. I'm currently working with a 2200, and naturally the 3300 is a different kettle of fish.

Yep; the PA 28 was even better. There's no such thing as a "good" aeroplane - but some are worse than others . . .

You're talking about a Marvel-Schebler (Facet) updraft carbie - an MA-3 I imagine - aren't you? With a typical nose-mounted Bracket filter, feeding a selector valve bolted directly to the carbie? And feeding directly into the normal 0-200 "spider" manifold? Yes, that setup is normal for them, and works fine. I'm not sure how even the spread of EGT is, for the 0-200, never had a chance to play with one. Quite a different arrangement than the one on the Jabiru, though. The Jab installation might well benefit from going to a selector valve right at the carbie, though; it's one of the things I'd like to try.

Yep - that's another clue. Ta.

Yes, I got that - means you were running on air that is probably around 25C warmer than the outside air - so the benefit you saw was wholly from increased vapourisation of the fuel spray. The difference between what you saw and what Dlegg saw is valuable info.

Well, I still don't know how to initiate a conversation on this thread; but I'm quite happy to collect you from Trevor Bange's strip. If you're flying a Savannah, you should be able to use my tiny strip, weather permitting (not advisable in a strong westerly). Right now, it's cluttered with building trusses, though.

Negative entropy, in fact!

I see you're based in Warialda; I'm near Clifton, QLD. If you're up this way, why not drop in and see what we're doing?

Thanks, JJ. The picture, thanks to Dlegg, JJ and Deb, is that there is a mixture mal-distribution issue; and it is significantly reduced if the induction air is warmer. Flow-straighteners in the duct upstream of the carbie help a bit, but are not a complete fix. However, a note of caution: The result Dlegg had, in reduction of the average CHT and EGT, from disconnecting the carbie from the airbox, is almost certainly due to the change this makes in the carburettor bowl venting; the air pressure in the airspace above the fuel in the carburettor bowl has a major effect on the overall mixture; there were some bad experiences in the work-up of those engine installations until it was realised that the bowl vent location had to be chosen so that selecting cold or hot air did not alter it. The Skyfox Gazelle installation initially had an enormous problem with this; the RPM drop from selecting hot air was about 2000 RPM. Once the correct vent location was found, the RPM drop disappeared. The Jabiru carburettor jetting is set so that it works with the factory bowl vent setup; when you disconnect the carbie from the factory airbox, this jetting is not going to work the same; the carbie bowl vent line would need to be plumbed to a flush vent in the side of a smooth-walled piece of tube that supplies air to the carbie, downstream of any air intake filter - and even that may not give exactly the same effect as the original. Dlegg evidently got some enrichment, but not so much as to stop the engine. I'm assuming, of course, that these experiments were done on -19 registered aircraft; it would be illegal to do that sort of thing on a -24 registration. Here's what I am starting to see from this: 1. Since there is no change in the airflow ducting downstream of the airbox when carby heat is selected, the reduction in EGT spread due to carby heat can only be that much more of the fuel spray is being converted to vapour before it gets the the end of the plenum; vapour goes where the air goes , but liquid droplets do not follow the air around corners very well. It should follow that aircraft in cold climates have a worse mixture distribution than those in warm climates. 2. Some, but probably only a minor part of the mal-distribution is due to air swirl in the flow coming from the airbox. How much of the effect is due to the system upstream of the carbie, and how much of it is inherent in the plenum / carbie system, may be something that can be examined in the test cell, by varying the induction air temperature (this is measured in the cell) on an intake that does not use the normal Jabiru airbox system. 3. Some variation in the distribution due to power setting is to be expected due to the deflection of the flow coming out of the carburettor by the throttle butterfly, especially at part-power settings; the butterfly should have least effect at full throttle. 4. Carburettor lateral tilt is a known contributor to mal-distribution; however its effect should be to move the fuel spray from side to side. It should be possible to learn something about the shape of the carburettor spray pattern by observing which cylinders show a rise in EGT, and which show a fall as the carburettor tilt is altered. This is also something I can investigate in the test cell. I'm very grateful for the information you have provided; it comes very timely, because I can add a few bits and pieces to the test cell setup, and maybe get some useful information as a result, which I will pass on. This is "wildcat" research; you never know what it will produce. There's never time for this in the commercial pressure of a certification exercise; I'm having a ball, doing this in my retirement. In the meantime, it is very hazardous to fly with the carbie heat system disabled. The principle of the carbie heat system is to ALTER the air temperature in the region of the throttle butterfly, sufficiently to move the temperature out of the icing zone. That can be achieved, in principle, either by normally running on cold air, and selecting full hot air when necessary; or by normally running on full hot air, and selecting cold air when necessary. Some engines object to being fed hot air when it is not needed; you can induce detonation that way, in an extreme case. So whilst it's a most valuable piece of test data, I don't advise adopting it as standard practice. There's a problem; we need to find a proper answer to it.

Thanks, Deb.

No, from the aerodynamic point of view, they're the exact opposite. And despite their crude appearance, they helped to even out the EGTs in the Motorfalke installation, which I approved under my CAR 35 authority in 2007. No snake oil involved, if that's what you mean; the plenum on the engine into which the carburettor delivers the fuel/air spray, is designed on the assumption that the air goes straight through the carburettor, i.e. the airflow is not moving in a spiral as it goes through the carbie. If that's not the case, the fuel spray will be off-centre in the plenum, which will mean the distribution pipes from it will not get an even share of the fuel spray. This is because the fuel is still largely in the form of droplets of liquid at that point; it hasn't had time to become vapour. They are not a complete answer, but they can improve matters if the airbox is inadvertently generating a swirl in the air as it goes into the induction duct. The Hiclones do look like snake oil to me.

Never seen of heard of them; Ian McPhee made those himself.

Thank you; that's a most interesting piece of information; the engine in the test cell does not have the Jab airbox, at the moment it has a pod filter; I'm hoping to be able to run it early next week (still a lot of instrument wiring to install and test). After it's been run in, in the test cell, it goes into an aeroplane, so we'll be able to get a direct comparison similar to yours. The question of restriction from the induction system can quickly be assessed by putting a manifold pressure gauge on it; at full throttle, you should not see more than about one inch of manifold pressure lower than what the gauge read before the engine was started. I suspect - but do not yet know for certain - that the problem is mainly due to the "difficult" run to the carbie inlet, necessary to clear the corner of the firewall; the "tight clearance" cobra head is something to be avoided, I suspect - but we don't really know until it's been properly tested. That sort of thing has the potential to induce a virtual tornado running through the carbie and on into the plenum - but we don't yet know for sure whether it really does this. We may know that pretty soon; it's something that can be tried. The induction system in the aircraft - tho not in the test cell - MUST be equipped to provide not less than 50 degrees C temperature rise in the air entering the carbie, when carburettor heat is fully ON. That is required at 75% power, and it has to be tested in flight. So any changes there must be tested and approved unless the aircraft is -19 or VH experimental; and if you make such a change in a -19 or VH EXP. after flying off your required hours, you'll have to do that all over again - so anybody wanting to try this needs to think before doing so.

Don't forget to ask Santa for a maintenance contract, whilst you're at it. Same as for any other engine, but more complicated.

Andy, you should really ask Ian that question re fuel injection. He can see, just as the rest of us can, that with the likely future variability of fuel, multi-point FI with knock sensing is likely to be the only viable answer. It would also get rid of the mixture distribution hassle and make life easier in regard to induction icing. However, it's a very major step to certificate; and I think it's on the shelf just at present. But I'm not Ian's mouthpiece. I've looked at a fuel-injected Limbach installation in a Grob 109, and I did NOT like what I saw; it required a split electrical system, for a start, because electrical failure will stop an electronic FI system. To give just one example - what happens to an electronic FI system if the alternator voltage regulation setup fails? A carbie does not give a damn what the electrical system is doing - it just goes on working. It also looked to me like an engine fire going somewhere to happen; there was no way it could be made to meet normal certification standards for powerplant fire protection. Any system simply transplanted from a car, seems likely to have those objections, from what I have seen of them; but those are the minor bits; the major bit is certificating the computer that controls it. One approach is to keep the carburettor and its manifolding etc, as a fall-back; but that raises the same problem as the proverbial theatre emergency lighting system that is never checked; when you want it to work, it very likely won't. What Ian is doing is developing a set of modifications that are, individually, classified as minor modifications, to be approved by putting them all on a specimen engine and running it through the JAR 22 certification endurance test, (once Ian has finished flogging them on his own J230; he has an impressive pile of ruined cylinder barrels already). That will give Ian a bunch of approved modifications that he can assemble as he sees fit. Exactly how they will be applied is still undecided, I think - but one obvious option would be for CAMit to set up an engine overhaul shop, and using repair-by-replacement, overhaul existing Jabiru engines and bringing them up to an identified CAMit standard in the process. They would be Jabiru engines with the CAMit rebuild, and would be data-plated as such, I expect - which I assume must transfer the liability for them from Jabiru to CAMit; but I'm guessing, here. Another option would be for CAMit to apply for a supplemental type certificate for the rebuild - which would definitely transfer the liability. A third option is for CAMit to obtain a full TC for the modified engine and start manufacturing them from scratch. You will appreciate that these are somewhat academic distinctions, since CAMit builds all the bits in the first place - but from CASA's point of view, the differences are real. I think you will in due course see all these stages applied.