It is somehow expected for the roll steering commands to be suspended when you want to use the ILS, what surprises me is the behavior of the STEC 60-2 constantly hunting for the localizer. I can swear I can manually do a much better job in keeping that cross hair, even in turbulence.
Can anything be done to improve the performance of the STEC A/P when guided by ILS? I can see there’s a ‘SOFT’ label that’s supposed to ‘soften’ the turns, can this affect the auto pilot when it follows the localizer? Or maybe it’s just a slow A/P that can be upgraded to a snappier model….
AlexTB20 wrote:
Can anything be done to improve the performance of the STEC A/P when guided by ILS?
The 55X is similar. I guess that is one of the drawbacks of the rate based AP’s vs attitude based. I get good capture only when there is few wind, the moment there is wind it doesn’t do LOC tracking very well. With GPSS however, tracking is spot on.
The only way to improve that would probably be upgrading to one of their new products coming up which then become attitude based.
Mooney_Driver wrote:
I guess that is one of the drawbacks of the rate based AP’s vs attitude based.
I think this is more a property of the S-TEC A/Ps than of rate based A/Ps in general. In my experience the rate based KAP140 A/P tracks the localizer very well.
@Mooney_Driver, so if I want the existing STEC 60-2 to perform better I can install a GPSS Converter and keep it in HDG mode instead of NAV.
As far as I understand, when in NAV mode without the GPSS converter, the airplane flies to a point and advances to the next one AFTER it gets there, always resulting in an overshoot of the planned turn, as opposed to the GPSS when it always predicts and starts turning BEFORE reaching that point.
So if I understand correctly, I can improve the A/P horizontal precision even without upgrading the Garmin GNS530 to WAAS, just by installing the GPSS converter, as both models, 530 or 530W output the same ARINC interfaced GPS signal?
I have flown back to back ILS and LPV with a G1000 system and the GFC700 autopilot on a day the winds were a strong at altitude. The G1000 GFC700 S turned with the best of them as the winds changed down low. I see no performance difference on any CDI based autopilot, Stec, Century, Bendix King,or Garmin under these circumstances. On the other hand, the lateral control on the G1000/GFC700 combination was a thing of beauty on the LPV, absolutely no S turns, the heading continuously and almost instantly adjusted for wind. It gave me the distinct impression that on the LPV, roll steering was being applied to control the lateral tracking.
On autopilots such as the Stec 60-2, the pilot can control the lateral without using vertical guidance by leaving the autopilot in heading mode and using GPSS. Of course one would need to fly the vertical without GS tracking and would have to hand fly the vertical. If approach mode is used on an LPV, it will use the lateral CDI deflections for guidance and have the same issues with S turns as occur on an ILS.
I have been able to suppress the S turns to some degree if I cheat on where I set the course pointer when I see a deviation begin as evidenced by the DTK and TRK begin to diverge on the GPS.
I suspect that this is simply the result if what each mode can use to track.
When in GPS mode, the system can assume that the GPS position is valid, and steer accordingly. This means it can take into account your ground speed and actual track, even if off the desired track, and knows the exact sensitivity of the “needle”. Roll steering or not, that gives it the edge.
In normal CDI mode, it has no more information than any other autopilot, so it probably uses a normal P-D-I control loop.
It perhaps could do a better hob if it were using the GPS info even in that mode, but I guess it is not allowed to as it is not a primary nav source in that mode
NCYankee wrote:
I have flown back to back ILS and LPV with a G1000 system and the GFC700 autopilot on a day the winds were a strong at altitude. The G1000 GFC700 S turned with the best of them as the winds changed down low. I see no performance difference on any CDI based autopilot, Stec, Century, Bendix King,or Garmin under these circumstances. On the other hand, the lateral control on the G1000/GFC700 combination was a thing of beauty on the LPV, absolutely no S turns, the heading continuously and almost instantly adjusted for wind. It gave me the distinct impression that on the LPV, roll steering was being applied to control the lateral tracking.On autopilots such as the Stec 60-2, the pilot can control the lateral without using vertical guidance by leaving the autopilot in heading mode and using GPSS. Of course one would need to fly the vertical without GS tracking and would have to hand fly the vertical. If approach mode is used on an LPV, it will use the lateral CDI deflections for guidance and have the same issues with S turns as occur on an ILS.
Many thanks for that. It is a very interesting observation. I have an Avidyne IFD, Sandel 3500, and STEC-30 plus GPSS. There has been considerable discussion on the Avidyne forum about how the STEC-30 should be set when flying an LPV or ILS/LOC approach (STEC-30 only has roll steering and alt hold, no vertical nav capability, so the vertical always needs to be hand-flown). Avidyne recommends changing the STEC AP mode from HD (ie using GPSS) to HI-TRK, which uses CDI deviation, when flying both LPV and ILS approaches. Your comments would indicate that leaving the AP in HD+GPSS mode for lateral tracking would give better performance, essentially using GPS for lateral guidance rather than LOC in the case of an ILS. I’m certainly going to go out and compare the two!
I wonder if your autopilot is wired to work from the SN3500, or directly from the IFD540. Most installers would not know about the former case, due to a general lack of familiarity with Sandel products.
The SN3500 contains what is in effect a roll steering converter, taking the GPS’s ARINC429 GPS output and generating the analog roll steering signals from it. Whether the result is better than using the IFD540 directly I don’t know, but anything is possible, and at least the autopilot action would exactly mirror what the SN3500 is indicating.
The GNS and GTN data cards are “locked” in some way (previous threads on this) otherwise you could duplicate them and load up multiple GPSs from one subscription. There is / used to be a hack when programming the GNS (non W) data cards which enabled programming multiple cards from one sub, but that’s all. A search should reveal the details. Some are locked to the card, some are locked to the GPS. Some use a database which is encrypted with a key stored in the GPS…
@Peter, are you talking about this?
I wonder if my Sandel 3308 has the same ‘software’ enabled roll steering…
Yes; it’s been in there, after a specific firmware version, for a number of years.
I have not used it yet because my KLN94 GPS doesn’t have the ARINC429 output which the SN3500 needs to do this.
