ianboag

I will own up to the fact that I habitually flare high. If I do a lot of practice I get better at it, but if I am away for a while I revert .... thump. It's boring - and of course practice is limited by the fact that the bit that matters is only about 10 seconds of a several minute circuit. I completely understand that real pilots "know exactly where the ground is" etc etc. And that there are all sorts of good ideas about "where to look" and "how to tell" and there is some good stuff on Youtube. You can buy an automotive ultrasonic parking sensor - with display - for about $20 on EBay. Range is up to 2.5 m. The one in the picture shows the height and a colour coded display in a neat little unit (about 40mm high) that sits on the glare shield. So I bought one - so far I have tried holding the sensor at various heights and it seems to perform as expected. I expect to attach it to an undercarriage leg using a simple bracket to get the max benefit of the 2.5 m range. I'm not sure how it will go with the ground whizzing past .... time will tell I was just wondering if anyone has tried anything like this.

Like I said. If the gyro dropped the cable, the pilot could expect to land fairly gently. Much easier than dealing with a parachute from 400 ft

Wot? It was a gyro, not a brick! Drop the tow cable and it could descend into the ocean at a perfectly acceptable rate. Given that it was flying at about 20-25 knots in use it would have been a doddle to ditch at low speed.

Solar no. 3G yes. Could be solar I suppose. It runs off 12V DC and draws about 350 mA. I run it off a wall wart feeding a long connection. Uses about 250 MB/month so a Telstra 12-mo $50 5GB data pass works fine. Raspberry Pi (running weewx) + basic Jaycar weather station + 4 Webcams + 3G modem stick + box + pole. I designed the page on the KISS principle. How many would you like? :-)

Now have two working in Oz if anyone is interested ynrm.avmet.nz (not actually Narromine airfield - down the road a bit) yktn.avmet.nz is recent - at Kyneton IB

I had a similar problem running an MGL tacho off the mags. I went through the (not very hard) exercise of setting up something similar to https://www.adafruit.com/product/2349 as well and that worked too .... phototransistor and IR source - pointed it at the flywheel and put a black stripe on it with a permanent felt pen. Eventually I hooked it up to one of the alternator leads and that worked fine. Had to fiddle with the calibration factor as I recall. I don't recall why I changed it - was a long time ago in my pre-Rotax days.

what the man REALLY wants is a gyro ..... reasonable (bit slow) X-country, magic for inspecting stuff and pretty much impervious to turbulence ...

It is a J200. Not a 701, but not too shabby on takeoff. I used to own one. The relevant "Jabiru failure" here is their previous engine which sucked a valve and threw a rod at 8000 ft on a cross country. The later misfortune on the beach is more fun to read about but really has no relevance to what this thread is about.

I see Telstra can do you a 1-year pass giving 5GB for $50. That would comfortably run the system for 12 months. https://www.telstra.com.au/broadband/mobile-broadband/prepaid

I can send you a fully configured SD card (think of it as an HDD) for the Pi. Can provide support to get you going. The minimum data plan here in NZ is $15 for 500 MB. That's plenty. The pix are about 40-50k each and I refresh them every 15 minutes from (dawn-30 min) to (sunset+30min). I think there's something more less equivalent in Oz - I remember getting something like that when I was visiting and got a local SIM to beat crippling charges for roaming data. Yes - wifi can work - or ethernet for that matter. There's a bit more mucking about with NAT redirection on the router to enable inward remote access for maintenance, but it can be done. Before I got the cellular stick working, I used wifi and a portable cellular hotspot - which is actually a router ..... The nice thing about the cellular solution is that the unit is totally self-contained.

That's the one. Before they did the solar/rechargeable thing, the sensors were powered by two AAs with a claimed battery life of something like 2 years. When I made my setup, I picked 5V from the power connector into the box, put a 5v-3v converter in the box (they cost about $5) and ran 3v up the pole to the sensors. The other benefit is that - very occasionally - the FO units get sensor lockup (not that I've ever seen it) and cycling the power is the only reliable way to fix it. So you need the weather station, any model Raspberry Pi, a case for the Pi, a cellular modem USB stick, a (powered) USB hub, n of the 720P Jaycar USB webcams where n is however many you want, probably n USB extension cables for the cameras, a UV-resistant plastic or wood (not metal) box, a pole and you are in business ..... There's a bit more stuff involved if you want to use solar power. The Pi is a bit of a power hog, but there are workarounds that I am still playing with.

The lens has dried off. I guess with that sort of wind, even at our cold temps the water doesn't last long .....

Right now it's raining, the wind is from the west and the SW facing camera has rain on the lens. That may take a reasonable part of the day to evaporate, but (from experience) it eventually does. Proofing the cameras against this is sort of hard .....

It's a Fine Offset 1080 - sometimes referred to as a "Jaycar junker". Can't argue with the $NZ200-odd price - the FO 1080 (and the later 2080) are generally accepted as inferior (accuracy/durability) to the much more expensive Davis units. But I figure that if it gives me the speed to within 3-4 knots, and the direction to within +/- 20 degrees that will tell me what sort of day it is and which runway to use. Wind changes all the time anyway. And if it croaks I'll just lower the pole and replace it - all the bits in the box are off-the-shelf-items. There are zillions of FO 1080s in use around the world. The whole thing lives in a box up the pole - I just feed it 5V DC from a wall wart. Can have as many cameras as you want. I'm working on reduction of the power requirement so it can run off solar. Not this week though. I figured that it should just give the info you need - ie the weather right now and what the sky looks like, rather than give graphs of the last week/month/year and gauges and compass roses and thermo-hygro whatnots and stuff plus it should be usable from a phone. Simple is good ... The software is weewx - a freebie (widely used around the world) for the Raspberry Pi - it can handle most weather stations. There's a similar one called pywws which works equally well I understand. The default report was one of those all-singing/dancing/story-of-your-life things, so I did some radical surgery on it ... There's a similar (free) PC product called Cumulus written by a guy in the Orkney Is. I tried it before I went the Raspberry Pi way - it is very good too. I had done the same radical surgery on the default report as well. The computer side is pretty easy to maintain because I can log in to it remotely .... I'm happy to lead anyone through making one. I'll even send the (configured) SD card for the Pi .... Sorry that the pic is not as clear as it might be. Wind stuff is at the top - other sensors and cameras are half way up. The Pi, cell modem and weather station console live in the box at the bottom. 5V comes through a hole in the wall ....

The way it works is that there is a computer in the station. It is a Raspberry Pi Linux board . Every 15 minutes it reads the weather, takes the pix then builds the whole web page and uploads it. There are no smarts at the server end. The date and time are when the reading was taken ...

Have a look at nzfi.inspire.net.nz .... I finally got there. Should probably add more $50 cameras ....

Or you could use the Rolls-Royce approach when they made the Merlin. Flog the engine till something breaks. Find what broke, strengthen it, bolt things up again, flog the engine etc etc. Repeat until time-between-breaks is considered acceptable. They ended up with a pretty reasonable engine I understand ..... :-)

Me - I am just talking about things as they are, not as they might be in a more sensible (as defined by who) world. Jabiru have a TC. If CASA think that's wrong then they should revoke it. Then CASA could explain which bits of the TC hoops it failed. Otherwise CASA should just shut up ....

Some philosophical points. CASA regulates Type Certificates. If an engine passes the rules to get a TC, CASA's responsibilty stops there. If some people at CASA think the engine is mechanically unreliable, but it has passed the TC stuff, that is just too bad. The opinion of a CASA staffer is no better that yours or mine. For example, noone keeps score on how many Lycs/Contis need a hone/ring before TBO etc. The current action is effectively "reviewal/revocation of a TC" but not based on any of the criteria for getting a TC. Otherwise known as "making up the rules as we go" which is done "because we can". Madness. Effectively CASA have "cancelled" the TC with vague talk of unreliability - perhaps CASA will "reinstate" it one day with equally vague talk about things being better. Don't hold your breath - that would amount to CASA endorsement and expose CASA to liability if the fixes weren't. This whole thing makes a farce of the TC process. It's a bit like eating sugar laden crap food which we know will make us fat and prone to diabetes etc. Because it is legal, eating it is a personal choice thing. Until it becomes illegal then the regulator has no place in this. Like flying behind a Jabiru. As it happens - I chose not to continue doing this, but it was a personal choice. The only rational thing for CASA to do here is to say something like "oops - we had no business doing this - our own rules have no provision for it". Then they should just apologise all over the place and shut up on the subject. Then the market will decide whether to (buy/upgrade to) Rodney's latest Type Certificated brain fart. IMO his track record is not all that flash, but it is not CASA's problem and CASA had no business being there so long as the TC is there.

I can't see the relevance of this to the current thread. There was an STC - ages ago - for early 450 kg Jabs with an 912UL (80hp) - owner was Tony Grills. The aircraft had a load capacity of one pilot, his dog, a credit card and 3 cups of gas. I think there were two aircraft converted. I think they might have been at Rockhampton and I don't know if they are still there and/or flying. There is a 160C at YCAB that was done via an EO, not an STC. I won't bore you with the details, but there will never be an STC based on that conversion - or another EO as the rules have changed. Dafydd L has pointed this out elsewhere. A 544kg 160C with a 912ULS is getting a bit marginal weight-wise too (you could ask me how I know). Can't do the 600kg thing as the stall speed is a tad high. There are other Jabs with Rotaxes in Oz ( I know of a J170 and a J400), but they are not factory builds. The J160C that I (foolishly) imported to NZ and (even more foolishly) Rotaxified was a factory build, but over here all ultralights are treated like Aus 19- aircraft. LSA's excepted of course .... Lord knows what would happen if an Australian wanted to buy it off me - CASA would have a bit of a fit. Effectively it would be a 24- turned into a 19- which is a total nono as far as I know. It flies nicely though. The time/cost are all sunk and in the long term we're all dead anyway :-) So I guess that strictly speaking you are right - there are legal 24- Rotax conversions in Australian skies, but don't hold your breath waiting to see any more.

I'm a Kiwi who Rotaxified his Jab, so I really have no skin in this game, but just out of curiosity - do you think I'm making sense? :-) Data logging - and dropping the results into a graph package - and video recording - are all just so easy these days. If any regulator was getting nervous she could phone up and demand a Skype walkaround. etc etc.

I need help here with the "huge amount of data analysis" notion. Let's look at a hypothetical 2-hour run. Logged every 5 seconds that would be 1500 data samples. Maybe 20 things would be logged ... Drop them into a spreadsheet and plot them as a time series. If there was anything odd, it would take about 5 minutes to identify that - if the data were presented intelligently. If there was nothing dodgy in the graphed data, then job is done for that run. You wouldn't sit down and watch a multichannel video (with sound) of an engine running in a test cell for two hours on the offchance that you might see something that wasn't in the (extensive and trouble free) data record ...... I still reckon the quickest way out of all this is for RS to bury the hatchet with IB, merge some of IB's good ideas into the Jab certificate and go with that. Assuming of course that IB's good ideas work out to be right :-) Can't be any worse that some of RS's previous tries.

Some of this certification stuff is a bit confusing to me. The certification process requires the on-site presence of CASA engineering staff at a gazillion dollars per hour with most of that being spent doing either nothing much or hardly anything. This is century 21 and video is soooo easy to generate in large quantities and data is sooo easy to log in vast quantities. So one might work on the idea that the process is all videoed (from as many angles as you want) plus all operational data (T's, P's, RPM, phases of the moon etc) is gathered at 5-second intervals. CASA take all of that and look for anything strange in it ... it would not be hard to sift out any bits of data that look a bit strange and go to the video to find out more. Assuming the results are unexceptional, two people should be able to knock that off in a couple of days. Even public servants with smoko/lunch breaks .... To keep people happy there could be the odd random auditing visit on site. If there are fifty sessions of two hours, then someone could check out an uneventful session in about 15 minutes. Hence my two man-days estimate. Over the years there have been a number of changes - hydraulic lifters, flywheel attachment, through bolt stuff, change ignition coils and probably more that I don't know about. Some of these would appear quite significant. One assumes the engine certification did not have to be redone from scratch (?). So there must be a point at which the post-mod engine becomes a "new" one and we (presumably) aren't there yet. So if Jabiru took some/all of the Camit changes (and the Stiff/Bent differences were resolved), how hard would it be to call the result a "Jabiru 2200/3300"?

Have a look at http://arp-bolts.com/p/arpultratorque.php

Why bother reinventing the wheel? ARP say - use our lubricant and know what a torque value means. It's what they do .....