Praise, pitch problems and some small questions
-
I could write paragraphs gushing in thanks and appreciation to Just Flight for their work on the F28, and they would all be deserved. Suffice to say that the Fellowship has long been one of my favourite aircraft, the upcoming release of the JF F28 was a huge part in getting me to finally switch to MSFS2020 after doggedly (and largely happily) sticking with FSX, and it was absolutely a Day One purchase for me. While I don't have the numbers, I'd put money on the big majority of my FS hours being at the controls of the Fokker since, and it's been an absolute joy learning its systems, classic oddities and getting back to grips with VOR navigation and procedures after so long being spoiled by watching an autopilot follow a purple line. A brilliant achievement that absolutely earned all the praise and recognition it's got - so thank you!
That said, here is one bit of odd behaviour I've noticed in the past few weeks (with V1.3 of the F28) plus a couple of other questions.
- The autopilot has a strong tendency to oscillate/porpoise in pitch control when trying to hold either a constant altitude or a modest constant climb/descent. It seems to only appear when descending onto a lower altitude during an approach (it's rock steady in the cruise and climb), but then the F28 does a 'nodding donkey' impression all the way round the pattern, drifting between 300-500 feet per minute up and down over about five seconds. It seems more pronounced at lower speeds and lower aircraft weights. Looking at the controls (I have no idea what is happening behind the scenes) the pitch trim wheel and the control column go back and forth, with the column seeming to get ahead of the trim wheel, which then starts trying to back off what it now sees as an overcorrection, which the control column then follows and over-does in the other direction. I suppose you could call it 'autopilot induced oscillation'.
Fast descents in PITCH or IAS mode are held steadily, then the nodding starts as the autopilot eases the descent as the target altitude approaches, then settles into the 300-500 ft/min oscillation when HT mode kicks in.
Switching off the pitch channel switch, getting the aircraft in trim manually and then re-engaging the pitch channel can sometimes cure the nodding, until another large pitch/altitude change is made under autopilot. Most of the time the nodding gradually returns and occasionally the the autopilot goes into a massive pitch-up response as soon as the channel is switched back in.
Some other questions if I may:
-
Is there a key/button input to use for the Autopilot Roll Control? There are ones available for the Pitch Control, and it would be nice to be able to use the Roll Control in a similar way without having to use mouse clicks.
-
Is the response of the autopilot to VOR beam/ILS localiser intercepts correct? The deviation bar starts moving, but the actual intercept response doesn't start until the bar passes the innermost 'dot', which I find even with very acute and 'easy' intercept angles leads to overshoots at typical F28 speeds. This is even more noticeable when doing approach/pattern work when you sometimes have to intercept localisers at steep angles, the sensitivity is very high and the beam is very narrow - by the time the aircraft starts turning in as the bar passes the dot, you're virtually overshooting.
I've been getting around this by using the RMI and the waypoint indicators on the HSI (took me ages to work out what the purple indicators on the HSI were for - essentially an RMI that points direct to the tuned VOR, as opposed to the radial set on the course needle - they aren't mentioned in the manual) to anticipate interception and manually turn the aircraft onto the course.
That may well be how 1970s avionics work and I've been spoiled by GPS RNAV and hyper-sensitive flight control systems, but it makes the workload very high when shooting VOR approaches, especially DME arcs!
-
I also note that when initialing a descent that the initial input to pitch down often results in a pitch up. Same happens with pitching up, the nose will first dip and needs repeated up input to "break" the dip after which it will happily respond correctly to incremental pitch inputs to tune the ascent/descent rate.
Regardless this remains with the 146 my go-to choices to fly. I cannot tell you how keen I am to see the A300..... It's killing me to hold off buying the ini until JF release theirs. If JF can match the quality in these 2 (and the Vulcan) than the pain is worth it :-)
-
@GroverJ41 Thank you for taking the time to fly and write about the F28. It's always great to see people enjoying the products we create and makes the team's effort all worthwhile!
I'll answer each of your questions in a different order than you posted them, but I hope it all makes sense:
(2) We don't have a control assignment linked to the autopilot roll control, as I don't believe there were any obvious control assignments that we could use for it in MSFS. I have made a note of that on our internal bugs/feature requests tracker though to see if there are any control assignments we can use.
(1/3) Questions 1 and 3 are both tied to the same issue, and the answer will depend on whether you have the GPS Navigation option enabled or disabled on the EFB. To give some background, we developed the F28 without any GPS fitted, so we were using the traditional navigational techniques of VORs, NDBs, ILS, etc. The autopilot and flight model was tuned to fly with these techniques and was working very well.
In the lead-up to release, we then decided to add the Working Title GNS530 as an additional navigation option for users who may want some more modern avionics, as well as an option to be able to import flight plans from the MSFS main menu for the autopilot to follow. The difficulty with the Working Title GNS530 is that it overwrites the simulator's core autopilot functions and attempts to force the aircraft to use the GPS's autopilot logic. This wouldn't be as big of an issue if the GNS530 was the only navigation option as we could fine-tune the autopilot and flight model to work with the GNS530 (which I personally feel is much better logic than the default MSFS autopilot logic), but the issue comes when we have to make an aircraft work with two autopilot systems that are using very different logic, and are fighting against each other for control of the aircraft.
So in order to have multiple navigation options available in the aircraft, we have to tune the autopilot and flight model to be compatible with both options. If you are interested, you can test this for yourself and see the difference in the simulator if you complete a flight with the GNS530 selected, and then complete another flight without the GNS530 selected. The autopilot will behave differently in both instances.
The easiest solution for us would be to remove the Working Title GNS530 from the product and replace it with another GPS unit which doesn't attempt to overwrite the simulator core autopilot logic (such as the KLN 90B), but we don't want to be removing features from a product that people may have purchased the product purposely to use, so we are in a bit of a tricky situation with what we can do.
The oscillations you have reported are something we can replicate on our end, but they typically only occur at low aircraft weights (which is why you are noticing them when approaching the destination airport). It can be nullified if you apply a gentle amount of opposite pressure on the controls in the opposite direction to the oscillation. We are actively looking into this and it is at the very top of our priority list for the F28, but unfortunately, our efforts so far haven't proven successful. Once a permanent solution is found, we will be sure to include a fix in an update to the F28 Professional.
The VOR/ILS beam intercept points are also controlled by the simulator's autopilot logic and/or the Working Title logic, so this will likely also vary depending on what navigation option is selected on the EFB. I would need to check this but as a side note. I believe the real F28 also has a relatively small maximum intercept angle of approximately 45 degrees, so to avoid overshooting, you would need to be within 45 degrees of the intended intercept angle.
Hopefully that answers your questions, but please do let me know if I've skipped over anything or any clarification is needed.
Mark - Just Flight
-
Thank you for the detailed and helpful reply, Mark.
Yeah, I should have said that this is all on 'steam gauge' navigation, not with the GNS530 switched in. I did do a flight with the GPS to see if that made any difference (I remember reading posts - probably from you - last year that talked about the layers of autopilot logic from JF, WT and MSFS). At the moment on my F28 the autopilot doesn't work at all with the GNS530, but I'm very confident that that's because I've got a bit of a tangle of mods going on, and I never fly the Fokker with GPS anyway, so I'm happy to go without it.
I fully appreciate what a job it must be for developers - especially ones dealing in older, quirkier systems and shooting for a high level of detail and realism - to work with/around the core needs and limitations of the simulation, so all I'll say is that I am keenly hoping you come up with a workable solution. If you don't, it won't stop me spending a lot of virtual hours in the F28 anyway.
I guess the same applies to putting in a command for the Roll Control - unlike the Pitch, most aircraft don't have an on-command, spring-centred roll input like the F28 so no one at Asobo thought to make it available, which leaves you rather stuck.
As I said, I'm quite happy to chalk the rather 'dumb' beam intercept to the nature of 1970s avionics. It's just forced me to learn how to estimate turning circles, start-of-turn offsets, closing angles etc. in my own head while doing approach work, which I'm sure is how the F28 captains would have done it back in the day when they had to come off an arc to intercept a localiser at right angles.
Thanks again for the reply. And to everyone at JF for the F28!
-
@GroverJ41 Just to correct something I said in my previous response, we do actually have control assignments for the autopilot roll controls, they are:
INC MAX BANC (bank right)
DEC MAX BANK (bank left)I've just realised that those aren't currently listed in the manual so I will make sure we get that updated in with the next F28 update.
With regards to the GPS not working, there are two Working Title GNS530 options available via the Content Manager in the simulator. One of them is much older and has a lower version number, and the other is the most up-to-date version with a higher version number. It sounds counterproductive, but in Content Manager, search "GNS530" and uninstall the one that the simulator allows you to uninstall. Once uninstalled, the F28 will "force" the correct version of the GNS530 to show, and it should then work correctly.
Mark - Just Flight
-
Hey @Mark I was thinking about this issue you have with GPS and non-GPS cockpit. Don't you think it might be easier to get some kind of a 2 installations-option? like when installing the bird you have to select if you want the version with or without the GPS? will that allow you to completely delete it from the system when flying without it? I know some other addons have this option in the installation process. At the end it is allowing you to work on 2 separates copy of you product? but maybe it is more complicate than this and I am just delusional :P
-
Having two separate aircraft folders for the F28, one with the GPS fitted, and one without, would allow us to develop two separate flight models, one suited to each option. Unfortunately, the problems with that would be the time and work required to develop, test, and maintain two separate flight models for each of the four F28 variants, and it would also remove the ability to hot-swap the GPS option whilst in the simulator which many users like the option of having. There are also uncertainties around how this may work with the marketplace version of the F28 that doesn't have an installer like the version purchased from our website.
It is a good idea though and I have logged it on our internal bugs/feature request tracker for us to discuss internally. We will be having these sorts of discussions internally and making decisions once we get around to integrating the UNS-1 into the F28.
Mark - Just Flight
-
Hey,
So cool to see that you are still actively taking feedbacks from users for this beautiful addon you sold us!
UNS-1 might offer the opportunity to just disable the GPS thing... I know some simmer are using it, but do we agree that in real life it was conventional only or UNS-1 equipped, but not a GPS?
Nevertheless, I do understand you cannot just remove a feature that some user might have paid for.
Maybe offering the option that you evaluating the most expensive to develop can become something like an "extra-pack" to pay for... for the extra-work you had to do. Such as:
F28 GPS fitted : extra pay
F28 UNS-1 fitted : extra pay
F28 conventional : standard priceit might be a solution too.
Hope you will find something to solve this issue.
Fingers crossed!Cheers,
-
I personally would fully support removing the GPS from the aircraft with the introduction of the UNS. It would be more accurate to the time period when F28s were most in use and it would maintain the use of modern procedures and waypoints for those that want to do it that way.
The "right" answer seems to be to go the route that some other developers have been going and coding a fully custom autopilot vs relying on the default MSFS one, but I can empathize with a desire to not spend development money on that at this juncture. Everything I've heard about the default MSFS autopilot sounds like it's just hot garbage.
