autopilot malfunction altitude hold

[ssp]

I have tried every which way. Initial use of autopilot is fine after takeoff. After climbing to altitude ok. If i change from heading to vor or to change altitude the wheel defaults to -900 ft per min. I attempt to change it back to +500 and then it goes back to -900. I click the autopilot on and off as well as the altitude hold on and off with the hope of resetting it. But this does not work... any suggestions?

[Harerton Dourado - HJG]

Please, I suoppose you're refering to new 707 panel?

If so, please tells what model you're using.

Thanks.

[ssp]

707-120B and 707-320B...

[jimhalinda]

I did a test and am seeing similar results.

Windows XP Pro
B707-320B

Took off, then engaged autopilot. Set for 200 fpm climb. In MAN mode.

Then I switched to HDG mode. No apparent problem. Aircraft turned to the selected heading.

Then switched to VOR (I had pre-tuned for a local VOR). Aircraft started tracking towards the VOR, but vertical speed on the autopilot changed to -900 fpm, and I couldn't change it back.

Drilled a big hole in suburban New York....

So there must be some bug in the autopilot when switching between certain modes. At least from HDG to VOR.

Tried again with the 707-420. Took off, engaged autopilot. Switched to INS mode, no problem. Switched to HDG mode, no problem. Switch to LOC/VOR mode, -900 fpm and I can't get out of it. Switching back to HDG mode didn't help.

Regards,

Jim

[ssp]

Glad to see I am not the only one with this problem. Hopefully someone can come up with the answer and/or fix. The same autopilot is used in the -128 with the same results...

[aerofoto - HJG Admin]

If I change from heading to vor or to change altitude the wheel defaults to -900 ft per min. I attempt to change it back to +500 and then it goes back to -900. I click the autopilot on and off as well as the altitude hold on and off with the hope of resetting it. But this does not work... any suggestions?

OK .... during testing I encountered this once with each of the 2 principal gauge sets during AP controlled flight and using the Mode Selector knob.

However the really odd thing about this is it only ever occurred "THE ONCE ONLY" with each gauge set .... and I've never .... ever .... been able to duplicate/repeat the experience again.

The first time this occurred I was climbing through through FL175. The AP had been engaged almost since departure (I was feeling lazy that day) and I'd trimmed for around 1000FPM climb rate and had HGD HLD engaged on the AP Mode Selector. I then selected VOR LOC as I approached the beam .... and suddenly got an 800-900FPM rate of decent. I mmmediately went back to HDG HLD and also reactivated ALT HLD as well .... but .... this uncommanded decent continued against my will. I then disconnected the AP completly and manually trimmed the aircraft into level flight .... then cautiously manually trimmed up once again for a 1000FPM rate of climb. Having achieved this satifactorily I then re-engaged the AP .... and the aircraft again commenced another 800-900FPM rate of descent. I cancelled ALT HLD and tried mousing the trim wheel, but, it refused to respond. It was then I discovered by shifting the Mode Selector knob to "AUTO" .... I found I could then mouse the trim wheeel desirably and without disconnecting the AP. After doing this all other AP controlled functions performed faultlessly throughout the remainder of the flight and there were no further problems.

As I indicated above. Whatever caused the problem has never .... ever .... occurred again since. Nor have I even been able to cause it to reoccur no matter how I fiddle with the Mode Selector knob during AP controlled flight .... "REAL STRANGE I KNOW"

Even as recently as last night I had the B707-120 & -120 panel aloft for over 2 hours just fiddling with all manner of things, and in particular trying to recreate what happened .... but .... "NO WAY IN THE WORLD" would it repeat it again and everything performed faultlessly.

Slightly off topic, but, worth bearing in mind too: One thing I've noted with B707-120 & -138 (early versions) is they aren't "performers". In the relaworld these aircraft were somewhat "underpowered" .... and I think the new FDE captures this pretty well.

If you're heavy you're not going to be able to climb high or quickly and "stage climbing" is recommended. I observe a lot of speed can be traded during climb .... but .... in level flight acceleration will increase "SLOWLY BUT SURELY". There's also need to plan your fuel and payload carefully .... or .... performance will suffer as a result.

You'll also find the bigger more powerful B707/720 versions much easier to fly. The custom FDE data for each of these seems to reflect their known performance quite well.

OH .... another thing too. Watch your speed during AP controlled flight. Get too fast and the aircraft "may" begin to roll a bit.

Also .... before engaging ALT HLD during climb make sure you're trimmed right down to much less than 500FPM rate of climb through that last 1000FT before your target altitude. A higher climb rate than this will induce a slight fugoid-like motion after ALT HLD kicks in .... which eventally settles down after again to perfect smooth flight around 40 seconds or so

That's what I've learned so far .... and like everyone else I continue learning too.

With these new panels and revised FDE I think we're all gonna have to "LEARN" how to fly these beast all over again .... right from scratch

Mark C
AKL/NZ

[jimhalinda]

OK, I did some more testing. Switching to LOC/VOR mode seems to work fine, as long as your rate of climb is set to 900 fpm or less.

As soon as I establish an autopilot climb at 1000 fpm or more (with appropriate power settings for a speed of about 280 knots), then switch to LOC/VOR mode, the -900 fpm setting comes on and stays there. And once it happens, disengaging and reengaging the autopilot doesn't seem to help.

This testing I did in the 707-320B. I established an autopilot climb at 100 fpm, then switched back and forth a few times between HDG, MAN, and LOC/VOR. No problems.

Increased climb rate to 200 fpm and tried again, cycling between modes a few times to be sure (even INS mode once or twice).

Then repeated at 300 fpm, 400 fpm, etc. (My passengers were no doubt filling their airsick bags as I rolled back and forth between nav modes!).

No problems up to and including 900 fpm. But once I hit 1000 fpm, then cycled back to LOC/VOR, that was it. Autopilot didn't care, it was going to do -900 fpm no matter what!

One other thing I'm going to try: I'll get into VOR mode at 100 fpm, then steadily increase the climb rate (as power and speed allow) until it's over 1000 fpm. Maybe if you have VOR engaged before hitting that rate of climb it will be OK?

I'll try it and let you know.

Regards,

Jim

[jimhalinda]

OK, did another test. 707-320B again.

Started an autopilot climb at 100 fpm climb rate. Worked my way up to around 500 fpm, then changed to LOC/VOR mode. No problem. Continued increasing rate of climb as power and airspeed allowed. Got up to around 2000 fpm and still no problems. After awhile I started cycling between HDG and LOC/VOR mode, and now no problem, even though I was climbing at 2000 fpm.

Tried one more flight, climbed in HDG mode until I was at around 2000 fpm. Switched to LOC/VOR mode, and again it went to -900 fpm.

Here's what I've learned so far:

1. The first time you engage LOC/VOR mode, if you are climbing at more than 900 fpm, you will get this problem.

2. If you engage LOC/VOR at less than 1000 fpm climb, you won't have a problem immediately, but MAY have a problem switching back to it at higher rates of climb.

Regards,

Jim

[Harerton Dourado - HJG]

Also .... before engaging ALT HLD during climb make sure you're trimmed right down to less than 500FPM rate of climb (through that last 1000FT) before target altitude. Anything more and you may induce a slight fugoid-like motion .... which will settle down after around 30-40 seconds or so

In fact the B707 operating manual says before engaging ALT HOLD the pilot may first level the ACFT manualy.

The performances of all the models I tested so far are great! And so is the smoke on the water injected models! HEAVY SMOKE!

And despite some minor issues with the AP gauge the new panels are simply fantastic. One need to be very careful to operate an aircraft made to be flown with six hands!

Congratulations to George and HJG!

[aerofoto - HJG Admin]

Jim !

I've just had a B707-320C & panel aloft this evenening for just over 1 hour. A nice bird to fly .... and after flying the -120 it was nice to suddenly have a lot of power to spare too ;D

I climbed manually to around 3000FT, engaged the AP, then HDG HLD, and began mousing the trim wheel to set an initial climb rate of 2500FPM.

No problems whatsoever were encountered during my climb .... all the way up to FL330 where I stayed for the duration of most of this evenings flight.

During my AP controlled climb I purposely cycled through all the AP Mode Selection settings (HDG HLD, VOR LOC, MAN, AUTO) whist maintaining a climb rate of not less than 2000-1500FPM all the way up to FL330. I repeated all this 2 or 3 times at least without the slightest problem ever being encountered.

Throughout the cruise at FL330 I once again cycled through all the AP Mode Selection settings several times more .... and again never encountered the slightest bit of trouble. No dutch rolling, fugoid, or uncommanded descent .... everything passed perfectly normally from T/O to touchdown.

So .... I dunno why some folk appear to be suffering problems .... whereas I'm not.

I dunno of if they've maybe got something else going on in FS or if it's just their technique .... in all due respect of course

I most certainly experienced precisely the same uncommanded descent rate you've encountered (in relation to AP Mode Selection settings), but, "ONLY THE ONCE" with each gauge set.

The 2nd time (and last time) it occurred ever I kept the AP connected, then immediately set the Mode Selection knob to "AUTO". Ensuring ALT HLD was "OFF" I then began mousing the trim wheel to re-establish the climb rate I wanted. Having achieved this I then re-enganged HDG HLD or VOR LOC .... and everything continued normally.

I'd like to suggest you purposely try to recreate the problem. Upon successfully doing so try my procedure (setting the Mode Selector to "AUTO" then mousing the trim), then once that's sorted reselect either HDG HLD or VOR LOC as previous. If it does it repeats its bad behaviour yet again then give it the same treatment to sort it out, and upon doing so exit then restart FS again and try the same flight once more.

I wouldn't mind betting it won't repeat this behaviour again as has been my personal experience .... "STRANGE" .... but true

Mark C
AKl/NZ

[dirtybird]

Well like most members in this thread I have experienced all of the VSI issues. I have tried a lot of combinations with the autopilot vertical speed and have come to the conclusion that it is random locking of the autopilot selector. Sometimes the autpilot selector will not budge during the actual problem. The first attempt to solve this issue, involved changing some of the autopilot values in the AIR. CONFIG file. I recomend that if you have the altitude alert inoperative; compare the DC-8 AIR. Config file values with the 707's. That is a good start. This seemed to get the vertical speed adjustment working as well as the VSI scroll/adjustment. Now the only remaining issue is strictly on climbout. The VSI will not hold climb rates at the typical 250 knot below FL10 regulated limits. Upon decent it seems to hold 900 FPM all the way down. Although I prefer-1800 FPM it will be ok if I can have the VSI speed at -900. I know someone out there has to have the first problem I solved in this same thread. Now my question that remains is should I only use the similar autopilot procedure as the DC-8-51? That will climb where I want almost from the point of gear up. Everything else seems fine including the lack of thrust in the 720 . I have used the 707 320 C panel for both of these planes and the same problem exists in both of them. So I know I did the installation procedures right. For the record I have a 1.4GB athlon with 1024 MB and MS XP Home Edition. Any reply or tip is greatly appreciated. I can still fly without the AP but it usually is difficult when flying within busy urban areas.

[christrott]

Guys,

You all need to remember this though - most of these aircraft DID NOT have a climb hold function. They had a PITCH HOLD, a SPEED HOLD and/or MACH HOLD function. To attempt to use the AP to control your rate of climb with any precision (or to modify the CFG to allow it) is to bypass the real operations of the functions. There is no "Altitude Select" or "Altitude Capture" and there is no autothrottle or autoland. These autopilots are extremely rudementary by what you guys are used to.

My suggestion is - go with it. If your AP is giving you odd behaviour in certain modes, then hand fly the airplane in the situations that you may have needed to use that mode. The real APs failed all too often, and most pilots prefered to hand fly the climb and then only engage the autopilot in periods of turbulence and once established at cruise (many times after trimming for cruise themselves). These airplanes are very much "pilot's planes" in their handling and the AP was mainly used to save fuel once in cruise.

If the AP function error is discovered and fix, great, I personally think it just makes the experience all that more realistic. I haven't had a chance to fly the plane yet to tell you that I'm getting the same experience, so I can't add that I have the problem too, but I know that I'd be handling the issue the way I described - like the real pilots did.

[jimhalinda]

C'mon guys, let's call it what it is: a bug. Everybody seems afraid to use the word. It's nothing to be ashamed of, it's just a quirk in the code somewhere.

I've been in software development for 16 years, I know a bug when I see it. I create them all the time!

Sure, it's hard to pin down what the cause is, it doesn't happen all the time, but it's still a bug.

Rationalizing that the real autopilot wasn't reliable anyway doesn't wash with me. Unless George intended to model an unreliable autopilot, which I doubt. To quote his manual: "The rate of clib or descent may be altered using the pitch command wheels". He intended for us to control rate of climb with them. Whether or not this is historically accurate is a different issue.

Now, George is a busy man, and this panel is brilliant work, and for free as well! And I appreciate how hard it can be to track down a bug.

But a bug is what it is.

I can live with it, of course. It won't stop me from flying these 707s, they're just so good with the new panels and dynamics that they're becoming my favorite! I appreciate all the hard work that went into these planes.

Regards,

Jim

[aerofoto - HJG Admin]

C'mon guys, let's call it what it is: a bug. Everybody seems afraid to use the word. It's nothing to be ashamed of, it's just a quirk in the code somewhere.

Call it what "you" will Jim .... but .... I'm simply not now experiencing the problem you're getting .... for whatever reason

As I mentioned above, I most certainly did encounter the same AP problem "once only" with each guage set .... but .... never again since. Nor am I experiencing any other issues either. I wouldn't say so otherwise.

So .... I'm buggered if I know what the hell's going on there

Mark C
AKL/NZ

[Harerton Dourado - HJG]

I bet it´s a compiler issue. But let´s wait for George to look at this issue!