The main challenge there tends to be the hacking of the hole in the wing, constructing the inspection cover for it, etc.
Wouldn’t this innocuous looking bit of kit sit somewhere where aluminum surgery is not required?
https://www.gps.co.uk/garmin-gmu11-magnetometer/p-0-3066/
If not it looks like a KLN94 with a resolver and BRNAV only. My IFR minima SEP are quite high, so approach approval is not a must have. If I ever needed a low approach it would be ILS.
The resolver could in fact be a GI275 in CDI mode.
I reckon most of these go into the tail.
It could go into the wing, if you can get access. It is actually a good Q why the King fluxgates go under an inspection cover. Well, they need to be rotated more or less correctly after installation. Maybe the Garmin one can be installed in any orientation?
The KLN94 will do GPS approaches just fine. It just doesn’t do LPV.
Maybe the Garmin one can be installed in any orientation?
The GMU11 needs +/- 0.5 degree from flight centerline and +/- 3 degree pitch/roll.
Outboard wing panel next to aileron linkage, ELT shelf, or behind the luggage space are all good spots that don’t involve surgery.
A_and_C wrote:
I find it hard to believe that the KLN94 is not approved for IFR ( NP ) approaches. It is a slide in replacement for the KLN89b that with the correct software level is IFR approved.
It is IFR “approvable” at the right software version for NP GPS approaches if the installation complies with AC 20-138AR. An AFMS is also required for the approval.
Graham wrote:
For the ignorant…. what is roll steering (and how else does an AP steer a heading?)
Roll steering (AKA GPSS) is a means of controlling the autopilot where a precise bank angle is communicated to the autopilot. It does not depend on a CDI deviation or an OBS setting to determine when a turn or course correction needs to occur. The GPS just provides the bank angle as a command to the autopilot. This allows courses that are not defined by a CDI to be flown, such as a DME arc, a holding pattern, a procedure turn and to lead turns complying with turn anticipation. Also since the GPS knows the desired track, the track offset, the current track, the GS and if there is ADAHRS input, then the heading and the TAS is also available so it can predict turns better and set an appropriate holding pattern. Even without the ADAHRS, the GPS will provide the bank angle needed to maintain the track, which is the wind corrected course and this is computed every 0.2 second for a WAAS GPS. So you don’t get the annoying S turns after a heading vector when you resume navigation as the autopilot re-determines the wind correction. With roll steering, the autopilot immediately turns to the wind corrected course.
I’d say there is a bit of ambiguity around this stuff – see e.g. here. Most of the GA planes which have roll steering in some form don’t have any airdata system (an ADC200 or a functional equivalent), so their avionics can’t possibly do properly precomputed turns.
Also there is no reason why a “CDI” cannot show the right thing on say a DME arc… after all, it can do it on a straight line, and a straight line is just a special case of an arc 
The roll steering setup is just a hack, whereby a GPS (whose database has holds, arcs, etc) can bypass all the “classic IFR” front panel junk and drive the autopilot directly by telling it what roll angle is needed, via an ARINC429 data stream (32 bit packets, each containing a ~20 bit signed binary integer value whose meaning is determined from an 8-bit “label”).
In a classic GPS/VOR/LOC setup, the autopilot is driven from the wiring on the back of the “CDI” where the same signal which does the CDI L-R deviation bar also drives the autopilot. This works fine for all enroute stuff and flies the 90 degree turns in a normal GPS IAP well enough (it is what I have been flying with perfectly happily since 2002) but when the GPS manufacturers started implementing stuff like holds, they gave up on outputting the right CDI signals (AFAIK) and moved to the roll steering via ARINC429. That means the CDI may not be indicating what the plane is actually doing. Whether this is actually true I don’t know; it could be bloody confusing, but I know for a fact that the roll steering converters which produce a faked heading bug (and the AP flies in HDG mode the whole time, not NAV) do produce turns which the CDI doesn’t reflect.
The industry also moved to running the GPS solution at 5Hz for WAAS whereas the KLN94 is 1Hz. But one could build a box which converts the KLN94’s RS232 “aviation data stream” to a 5Hz ARINC429 roll steering data stream; I believe the required roll angle information can be derived from the current leg track and the current track. The 5Hz is certainly not a problem, using some quadratic interpolation of the 1Hz data.
Peter wrote:
This works fine for all enroute stuff and flies the 90 degree turns in a normal GPS IAP well enough (it is what I have been flying with perfectly happily since 2002) but when the GPS manufacturers started implementing stuff like holds, they gave up on outputting the right CDI signals (AFAIK) and moved to the roll steering via ARINC429.
That is not what happened. Roll steering was available on the KLN90B and many of the non WAAS GPS systems such as the Garmin GPS 155. DME arcs were supported if they were in the database. When WAAS came along. holding patterns and PT were an addition. With the WAAS unit and airdata with heading input, the shape of the holding pattern is not an oval racetrack.
The early autopilots do not have a roll steering input, and Stec invented the converter that converted a bank angle into a heading error voltage. Later autopilots such as the Stec 55X and the KFC225 have a roll steering input via ARINC 429. The KFC225 also has a roll steering kluge that is supposed to work with the KLN94, but performs terribly.
For those of us who have roll steering available with our autopilots, I have yet to hear anyone wanting to go back to CDI steering. So last century :)
I do now recall the KLN90B has some ARINC429 ports but when it was being installed there wasn’t any autopilot which could accept ARINC429 roll steering – AFAIK.
The earliest AP I know of with roll steering is the KFC225, and that one had a short OEM life due to a (presumed) firmware bug which trashed the servos pretty fast – unless a mod was done to the servos, but you can’t do it on a certified plane, and HBK had no idea how to fix it properly because anybody who knew anything left c. 2003. The AP business then had zero innovation until Garmin came along many years later.
Interesting poll on avionics shops in southern UK.
1. I got one response on options, and not unreasonably, the shop was not interested in replacing the CDI with a resolver, but quoted a high four figure sum for installation of a GNC355 and a GI275, with an estimated install time of three weeks. They might be able to help in June. Am trying to understand an 80 to 100 hour install. With the hardware this would amount to around 20-25% of the value of the aircraft, with effectively only adding around half of that to the second hand value.
2. Another shop said they were too busy to quote and to come back in July!
3. Am waiting on two other shops, and I have had another shop suggested to me.
No shop has suggested whether the removed kit gets any credit. Do not understand if that also applies to avionics work in the USA. The kit is old, functional, and appears on Avionics shops as worth around $3k plus. It does seem a shame that this might not receive any credit, say 30% or 50% of value.
RobertL18C wrote:
It does seem a shame that this might not receive any credit, say 30% or 50% of value.
Problem is, cash is king and holding old stuff taken in part-ex ties up cash which can takes ages to realise if and when a buyer for the removed stuff can be found. I was also reluctant to offer a part-ex except for very specific items, and instead used to offer to sell equipment on the owner’s behalf and take a percentage when it sold. When I sold my business we were still holding several hundred items of old removed equipment which just wouldn’t shift.
