Jumpy Airplane cannot trim

[Kent]

Hello Gentlemen:

I have been playing with FMS for a few weeks now and very much enjoy it, however, I am having difficulty trimming my airplane so that it will fly level. I have tried calibrating it many times but without luck. The plane is "jumpy" and needs constant correcting and controlling. I cannot fly very low or come in for a landing as the plane often seems to loose throttle and nose dive into the ground or "wing over" and crash. I can't seem to get it to fly level or straight no matter what I try. It has nothing to do with my frame rate as it happens even without textures, fog, or smooth textures on. I have a Pentium 233 with 163 MB of ram so its not memory related. I also have Windows ME, and a Futaba Skyport Six radio hooked up through my Parallel port. I have noticed when calibating, if I sit and watch the control levels in the "calibration" screen, they seem to bounce somewhat in a cycular fashion beginning with channel eight and working down to channel one, then back to eight and down again. I really enjoy the program and would be extremely pleased if anyone could offer some suggestions for fixing this annoying little problem.

Kindest Regards;

Kent

[Kent]

Hello All:

Thought I would update those of you who have been having the same problem with FMS and its jumpy channels etc. I have been told from a good source that to some extent this problem is software related (i.e. its a problem with FMS itself) that will be fixed in forth-coming editions. Apparently all who use FMS experience some channel or signal motion. Is this true?? However, knowing that a good friend of mine also uses the program and has a very stable and smooth flight environment made me wonder about this. So he brought his radio and cable over and we did some comparisons. All said and done, the problem came down to my Futaba Skyport 6 radio. We alternated between cables, power sources (we have made low voltage transformers that plug into the wall to power the radio so we don't drain our batteries and they work great!!), and radios and found that my Futaba is the culprit (which is really unfortunate since I bought my friend's JR Century VII radio for him for 25 bucks at a swap meet and it works perfectly whereas my 300 dollar Futaba doesn't...hahaha). If anyone needs the pin placements for building the interface cord for the Century VII radio let me know, although they are very similar to the newer JR radios. I don't know why the Futaba radio does this and will let someone who knows much more about electronics specualte on an answer to that one and perhaps pass it on to me. Anyway, the solution to the problem seems to be a very cheap second hand FM radio. My friends channels modulate or jump slightly but nothing near the motion my channels go through with my Futaba hooked up (and I have always been such a fan of Futaba radios ***smile***). Anyway, I hope that all this helps someone other than myself. Would still love to hear any theories as to why the futaba does this and if it can be fixed though.

Warm Regards;

Kent

PS: Anyone have a older JR radio they want to give away or sell for a really cheap price??? haha

[jr8103dt]

I have the same jumpy airplane problem with my XP8103.

[Michael_Moeller]

Quite an interesting thing! There is a certain problem with the current "DLL-Driver" for the interface in FMS. There will be a new VXD driver with the next beta, that will fix (hopefully) the noise problem on the signals, or at least improve the situation.

I have already heard from different places, that there is some "working down" noise on certain configurations. I'am not very sure what is the source for that. It doesn't need to be the radio itself, but some complication with the time-critical puls-length-evaluation inside FMS. This means, that it could be, that the FMS just has some rpoblems with the signals comming from Futaba, but the Futaba signal itself is clean. To check this, it would be interesting to have some measurements with an oscilloscope on the PPM signal, to see, wether this noise is just in the signal, or, some complication reading it.

Michael Möller

[Wayne Tarr]

Yes, I have oscilloscoped just about every pin and connection and found a lot of noise around the PC box. With the adapter away from the PC(not very useful!) there was NO noise.
email me wt@hnpl.net for info!
I've also written manuals in English.

[Shakeel Ahmad]

How about using a bandpass filter centered at the PPM signal frequency or some sort of sheilding.
Can someone suggest what kind of noise is generated by PC?

(Edited by Shakeel Ahmad at 7:54 am on May 31, 2001)

[bsergeev]

There's something wrong with Futaba-s, that's for sure.
I have JR 622 and it works really smooth, no jitter at all!
I also tried an old Futaba Conquest FP-T6NFK with the same cable but with a different connector (6-pin DIN) and it was as jumpy as a little donkey.

[Wayne Tarr]

I have come up with some more oscilloscope info from testing many more radios, and it does seem that some radios put out a jumpy signal, which indicates they are in need of a signal retune. New radios probably work better than old ones for this reason. Still, I have interfaces with bandpass filters on them now, and it does improve the external noise problems quite well.

Wayne Tarr

[Isaac]

In my computer, FMS is also *very* "jumpy" (Pentium MMX 166MHz 64Mb RAM Voodoo3 3000). I use a Futaba Conquest 4 channel.

I watched the signal in my oscilloscope (Tektronix THS730A 200MHz 1GS/s), and found it very steady, no jitter. Even after the transistor, in the parallel port pin, the signal is absolutly steady.

I think this effect is due to interrupt latency in Windows, it is: when the signal arrives to the parallel port pin, it generates an interrupt signal to the processor. The processor needs a litlle time to start serving the interrupt, and this time is a bit different each time the interrupt is generated. When the interrupt service routine is executed for measuring the time, an unknown ammount of added time has passed, so the time measuring for the pulses is variable.

I think that in a faster computer, the interrupt latency is lesser, and in a very fast computer it can be so small that it won't interfere anymore.

So, before blaming your Futaba transmitter, look at your computer's speed.

Best regards,
Isaac

[bsergeev]

My PC has 850MHz CeleronII processor and 512MB memory. JR is as smooth as a baby's butt, while Futaba is as jumpy as a bighorn sheep in Rocky Mountain.
Probably, in some cases, problems arise from the noisy signal and in other -- from PC/Windows.

[GeorgieBoy]

I also have a Futaba radio and am experiencing some jumpiness with some planes. It seems to affect the high performance planes mostly.

I am running a P4 1.3 GHz processor so a slow computer is not the issue. Hopefully the new DLL will help.

[boby]

This may help. The March 2002 issue of RCM&E magazine has an article about making a dummy transmitter specifically for FMS. Cost less than £50 !!!!!

[heliphil50]

FMS doesn't like futabas mine is not too bad with the futaba but I have borrowed a friends JR and wow brilliant (Thanks Brian) no jitters at all I've ordered the interface

[jfitter]

I hope this helps pin down the jumpy airplane problem. I have the same problem. I am a turbine flyer and mainly use the F4 model with the thrust increased to 60N - really exciting when it's jumpy!! I have more-or-less fixed/worked around the problem and here are the results of experiments.
Forget interface/dll/noise problems - there are no problems. I use a computer encoder I/my company developed and thus I am able to communicate with FMS directly through the encoder's serial interface and make firmware changes to the encoder for the purposes of experimenting. The converters are 10bit thus the positional output (for ppm) is 11bit resolution with 1bit jitter (plenty of resolution). When I feed FMS with this the "jumpy airplane" problem persists. Suspecting the jitter, software filters were applied in the encoder (2nd order low pass), but as expected the problem was not cured. A few other approaches were explored, including varying the trim rates, all with no success which only left the physics engine. If this were the culprit then changes to the model data should improve things - it did!!
To get the F4 (and other airplanes) to fly properly (almost) it is necessary to reduce the moments of inertia about all three primary axes to very small values. Some of the other parameters were tweaked to more realistic values also, but the primary fix was the MOI change. The planes do not fly absolutely correctly just yet but the improvement is spectacular.
What pre-empted this approach was the observation that the "jumpy airplane" appeared to behave as if it were over damped on the 3 primary axes. This could occur if the interaction between the MOI and the lift vectors followed a first order response rather than the expected 2nd order response. Reducing the MOI's to negligible values does not improve the type of response but does remove much of the damping. The result is an airplane which no longer feels over damped, but does still suffer from an incorrect response about the primary axes - ie. it does not jump about but has an unrealistic feel about it.
Since the MOI values required for correct flight dynamics are unrealistically small this indicates to me that there is some problem with the physics engine which needs attending to.
Looking for noise with cro's, trying different transmitters, etc. will not solve this problem because it is not caused by any of these things - I know because I have already tried all of these.
If anyone would like to experiment with the modified F4 they are welcome - just send an email to me and ask.
I would like to see this problem solved because I think that FMS has the most potential of any product of this type that I have ever seen. I enjoy using it and I enjoy it's technical open-ness. Solving this problem, I believe, would result in the best handling simulator available - none of the others I have used actually "feel" like real models and I have 30 years of competitive flying to draw upon for that comment.