I would never complain about anything that makes aircraft maintenance / upgrades easier, and AML STCs have dramatically simplified this in FAA-land, but what is the logic behind them?
A generic STC cannot actually properly apply to each aircraft type. This is evidenced by the fact that if you have an STC (AML or non AML) to support a 337, the FAA never looks at it; they just file it. But if you don't have an STC (and are thus doing a Field Approval) they make you go through all kinds of hoops, ensuring that the documentation really does describe your aircraft and what's in it.
EASA have always really hated AML STCs, though one cannot be sure to what extent this was because AML STCs would take a lot of business away from EASA Part 21 companies, and EASA has always been very close to that industry - not least due to having recruited heavily from it! But recently Garmin managed to find a way and now have AML STCs for the GTN boxes (which gives Garmin a massive commercial advantage in Europe). But... even then EASA has the last laugh because they are dragging their heels on approving each version of the firmware.
I think the argument ultimately comes down to how much you trust the people doing the installation and how much common sense they are allowed to have. In the EASA world, this has traditionally been minimal.
You have answered your question yourself :)
Every modern car radio is more difficult to install than an aviation GPS; I've never heard anyone installing a car radio needing an STC.
So the real question is what is the logic behind non-AML STCs.
An STC with an AML is similar in concept to a field approval where there is an existing example STC. The difference is who approves it. In the field approval, each installation is verified by a local FSDO to make sure that the paperwork is done and that the example STC applies to the new application. With the AML, the STC is approved for a list of aircraft as long as all of the installations comply with the same interconnect and the general installation instructions. Here is the wording on the most recent guidance on ADS-B field approvals, note the conditions that are applied:
Can ADS-B Out systems be approved using data approvals other than an STC, including field approvals? Yes, ADS-B Out systems can be approved using data approvals other than an STC if all of the following conditions are met:
a) The ADS-B Out equipment is authorized under TSO-C166b or TSO-C154c;
b) The GNSS position sensor is approved under TSO-C129 or later, TSO-C145a/C146a or later, or TSO-C196 or later;
c) The ADS-B Out equipment (transponder or Universal Access Transceiver (UAT), GNSS position sensor, and interconnect wiring are identical to previously-approved design under type certificate or supplemental type certificate;
d) The installation is performed in accordance with the equipment manufacturer’s installation guidance;
e) The installer verifies the installation in accordance with the guidance of AC 20-165, Chapter 3 and 4. The data from the previously-approved installation may be used to address paragraphs 3-1 c, 3-1 d, 3-3 b (2), 4-1b, 4-1c. A return-to-service operational check flight in accordance with AC 20-165 Section 4-3 is recommended for determining if the installation performance is acceptable;
In other words, if the installer can reference an existing STC and not change the interconnect, it can be field approved.
In the case of the AML STC, the grouping of similar class of aircraft together is in effect going thru the same analysis as it isn't an aircraft by aircraft modification and as long as they are done using the same interconnect (provided in the manufacturers STC install manual), there is no need for additional approval.
I hate to accuse the FAA of being efficient, but in this case it sure makes the installers life easy.
Very interesting; thanks. I did not realise avionics AML SCTs are based around the assumption of similar or identical interconnections. It makes sense, of course. One could have the same stuff in a TB20 or a C421.
On the regulatory side, two additional thoughts:
Other than in the FAA system, there is a fairly hefty fee for each STC. More STC's, more fees, vs one AML STC, one fee. Where it is the same mod, independent of the aircraft type, the notion of a type specific approval is a bit over the top sometimes, and the FAA knows that. However, an STC, by it's very name, is a supplement to a TC, so it has been argued that an STC cannot supplement a whole bunch of TC's , even if it is convenient.
When only a "337" form, and not an STC is used as the method of documenting approval, the "FSDO" office may have reviewed the design compliance of the mod, but it is possible that there has been no aircraft certification discipline involved. This can create compliance gaps, when an aircraft design requirement has been overlooked completely because the 337 form "approval" issuer did not think of it. An AML STC at least ensures that the STC issuer applied the design requirement review to every aircraft type, and assured that there was no design disconnect with that mod on that aircraft type. For this reason, there might be blocks of aircraft not on the AML of the STC. It at least assures that a particular characteristic can be flagged for review at installation, where other mods could also be involved.
Sometimes the reasons have little to do with the actual aircraft, but more to do with the regulatory structure above the approval.
When I installed the 430 in my Comanche it came with STC for (I think) a Piper Arrow or a Cherokee 6. When it was done the papers + the 337 were sent to the FAA office at Heathrow, the inspector refused it and asked for data, I didn't know what he wanted and still don't know. After becoming a bit hot under the collar he came over, set in the aircraft for 15min opperated it and signed the 337. It seems that having an STC does not always help.
Ben
FAA office at Heathrow? I'd be interested in when that was. As far as I have been able to tell they have not done avionics field approvals in years.
If you have an STC but it doesn't cover your aircraft type (or it does cover the type but e.g. for the wrong airframe serial number range) you can use the contents of the STC to support a field approval.
The design data used to get an FAA STC can also be used to apply for an EASA STC, nowadays.
A 430 is not a Major Mod anyway (FAA or EASA) unless (AIUI) driving an autopilot.
Every modern car radio is more difficult to install than an aviation GPS; I've never heard anyone installing a car radio needing an STC.
The consequences of a car radio not working are likely to be less troubling than the consequences of the GPS failing during a GPS approach though.
Nobody would argue that a say TB20 is more complex than a 1970s car - it isn't. My VW Scirocco is way more complex than the most modern GA aircraft.
And avionics are mostly really easy to install. They have to be otherwise the vast majority of installers would not be able to cope! The procedure for installing almost anything is to get the installation manual(s), and start from the back where the wiring diagrams are. You look for the one which most closely matches what you are trying to do and go for it. There is about a 95% chance it will work - or will work well enough for the average customer to not notice any missing functionality.
The 1990s kind of boxes (the KX155 radios, KLN94 GPS, KI525 HSI system, etc) could be installed that way by an absolute monkey and provided he knew how to use a crimping tool, the job would work. I have never played around with the more integrated stuff (G1000 etc) but it's probably similar, though the installer needs special "codes" to access the setup menus.
Crucially, NO understanding of electronics or how the actual signals represent the data is required. It is just wiring, with some knowledge of the function you are connecting up.
There are a few avionics people (I know 1 or 2 in the UK) who actually understand the stuff at the "signal" level (e.g. look at the XYZ/400Hz-ref heading signals with a scope and check if the heading represented is right) but they are very busy and do not work on their own so the chance of getting a decent job done by their company is well below 100%.
The problem is that to comply with ICAO, you have a certification system for components and modifications which somehow has to be complied with.
The way EASA and FAA operate it differs but it differs more in implementation than in the basic philosophy.
The FAA tends to let the installer use his judgement, whereas EASA tends to assume he is useless and likes to have more prescriptive instructions.
I doubt say a GPS failing in flight is any more troublesome than drive by wire or brake by wire failing on a motorway with 100MPH.
The transmission controller of a friend's car decided that the first gear was appropriate on the motorway with 100MPH. This ended up in an uncontrollable excursion to the neighbouring lane. Out of sheer luck, nobody was there.
In contrast, I experienced two GPS crashes (interestingly, at approximately the same spot in south England on two different flights), a TSO C129 unit. It caused some higher workload trying to restart the unit, but that's it, no particular luck involved.
The problem is, the "monkey installer" approach by EASA doesn't work. In the end, you need to have someone who actually understands electronics. I had two difficult problems with my aircraft. One was the glide slope indicator jumping all over the place, and generally indicating too high, the other was the "aerobatics programme" of the autopilot. For both problems I went several times to the most highly awarded (by EASA) avionics company in the region (they even have a DOA [Design Organisation Approval, not Dead on Arrival]), they always told me everything was ok. In the end I had to find the problems myself. Had I relied on the glide slope or the auto pilot in IMC, things would have gotten ugly. But the paperwork was ok.
I had some help from a friend who was a very experienced avionics guy, but he sent his license back when EASA came along with their excessive paperwork orgy, and decided to quit aviation for greener pastures.
The same will undoubtedly happen on the mechanical side (if it hasn't already), thanks to Part-M, a remarkably bad piece of legislation.
The EASA regime is attracting the wrong people, those that know how to operate a pen and not those who know what a screwdriver is. And companies are wasting huge amounts of money on getting and keeping DOA's, instead of education of their employees.
