Anyone figured out how to make the Cessna Trim Wheel behave?

[Noel 4,487]

I bought this guy a few yeas ago and while I can trim the aircraft beautifully with it when it comes time to land (a Majestic Q400 for example) I can't seem to get the thing to stop porpoising, or take a sudden dive right before touchdown.   I am using FSIPC to calibrate the axis and as I say I have nice fine control w/ the wheel while in flight.   If yoeu've had good luck making this thing work let me know how!  I go back to using one of the toggle switches on the yoke I use which if fine, but I do like the trim wheel better if I could only make it work during landings!  

[MarkDH 234]

1 hour ago, Noel said:

I bought this guy a few yeas ago and while I can trim the aircraft beautifully with it when it comes time to land (a Majestic Q400 for example) I can't seem to get the thing to stop porpoising, or take a sudden dive right before touchdown.   I am using FSIPC to calibrate the axis and as I say I have nice fine control w/ the wheel while in flight.   If yoeu've had good luck making this thing work let me know how!  I go back to using one of the toggle switches on the yoke I use which if fine, but I do like the trim wheel better if I could only make it work during landings!  

You have probably discovered that the adjustment is too coarse. The Saitek trim wheel only has a 10-bit controller if I remember correctly, which means there are only 1024 positions end to end. According to the SDK, the elevator position has 32769 possible values (+/- 2^14), which presumably means you can get finer control by adjusting the trim in other ways. Short version: you can't fix it 

But wait, there's a slightly longer version . The Saitek trim wheel is actually based around a rotary encoder and if you are prepared to hack it and solder a couple of wires you can use this to your advantage. You will need a controller board such as the Leo Bodnar BBI-32 and you can connect the encoder to that and it will behave like two buttons. You can then map the button 'clicks' to the elevator trim events and gets finer control. This is what I have done in my cockpit. I talk about it in this video and the video description has a couple of links to where I learned how to do it.

Edited April 4, 2019 by MarkDH

[flyforever 267]

Are you talking about the aircraft pitching down after you turn off  the autopilot?

This is a behavior that many aircrafts display due to a poorly programmed trim wheel and autopilot function, that is, the autopilot maintains aircraft control using the trim setting.

If the autopilot trim setting and the trim wheel setting are not the same when you disconnect the autopilot, you will experience this phenomenon.

I believe FSUIPC has a check to reduce this phenomenon.

tony

[Noel 4,487]

3 hours ago, MarkDH said:

You have probably discovered that the adjustment is too coarse. The Saitek trim wheel only has a 10-bit controller if I remember correctly, which means there are only 1024 positions end to end.

There are actually 4095 positions when I test for this using FSUIPC JS calibrator, FWIW.  I'm sure you're right that is playing a role.  Is it safe to assume using a toggle button to adjust trim you're getting the fully quantized 32769?  I know it seems to offer finer control.  The Crash 8 is not fun to hand fly w/ the controllers I have so I depend on the trim toggle a lot for this plane.  I dialed back sensitivity to 1 for the trim wheel and it's helped some.  When the plane has some better airspeed the trim wheel is wonderful--it's only when it's in the final moments when you're at that pre-stall landing speed that it gets dicey.  The controller I have is the rebranded Saitek yoke and it's pretty crude, at least as I have it set up for that plane.  Plus, the aileron turning is notchy--it's not smooth.  I wonder if I can take it apart and spray teflon or something on whatever parts are involved in it.   I did manage to land the thing a moment ago at KDEN and I hand flew it for a good 5 miles to touchdown, but it wasn't real pretty!  I think I'll be able to adjust I just have to not disengage the autopilot until within a mile of TD and be ultra ginger w/ inputs, get it down to that 1/4095 level of resolution, which isn't half bad when you think about what we're doing it would seem.  The yoke is bloody oversensitive too for elevators even though I just took it down to 5 or something.   Might try FSUIPC for that and its slope adjustments.

Edited April 4, 2019 by Noel
why not?

[Noel 4,487]

2 hours ago, flyforever said:

Are you talking about the aircraft pitching down after you turn off  the autopilot?

This is a behavior that many aircrafts display due to a poorly programmed trim wheel and autopilot function, that is, the autopilot maintains aircraft control using the trim setting.

If the autopilot trim setting and the trim wheel setting are not the same when you disconnect the autopilot, you will experience this phenomenon.

I believe FSUIPC has a check to reduce this phenomenon.

tony

Thanks Tony, that is not the issue and I know where that control lives in FSUIPC.

[flyforever 267]

Unless you can change the pot into a multiple turn pot, only software can fix it. If you take apart the unit, you can check to see if the pot is a 60 degree or a 270 degree pot.

If it's a 60 degree pot, each step of the trim wheel total rotation represents 1/60 of 4096.

A 270 degree( most pots are 270 degrees) will give you 1/270 of 4096, which translates into higher resolution.

Now, I am strictly referring to how a slight mechanical rotation of the trim wheel affects the pot rotation. I always felt that trim in FSX or P3D was oversensitive when using a joystick.

If you could do this, you would have to ensure that the trim wheel could rotate more than it currently does.

It's a hack, but most trim airplane trim wheels rotate more than 1/2 turn or so. I don't have the unit, so I am just guessing that the unit does not rotate much.

One option is to add a brand new controller that provides curve tweaks. I am not familiar with Bohener controller, but I am sure his software allows for such tweaks.

There are a few of these joystick calibration freeware out there, but I am not sure if they still work with the latest windows OS.

tony

 

[Freo 83]

I have one, which I've only used once or twice. Gave up, and now gathering dust in a cupboard.

[Noel 4,487]

1 hour ago, flyforever said:

It's a hack, but most trim airplane trim wheels rotate more than 1/2 turn or so. I don't have the unit, so I am just guessing that the unit does not rotate much.

tony

Oh it's a multiple turn pot alright else you couldn't dial 8 increments x 4095 times to go from one pole to the opposite pole.  It actually works really well I think quite possibly it is the Crash 8 that is hard to fly at low airspeeds, and you don't want to be hand flying it any longer than necessary to TD.  I just reconnected the trim wheel after turning it into a means of holding my keyboard in place on top of my yoke.  After setting it up correctly it really does work well.  One other possible problem I was having was my many years old Cessna Yoke and TQ was going bad, so the combination of the plane a low speeds and marginal controllers left me wondering what was happening.  I replaced the Saitek w/ the rebranded Saitek of similar design by Logitech and things seem a lot better such that I was able to hand fly a landing all the way to TD w/ the trim wheel, but it wasn't pretty as I say.  I'll try it with PMDG and see if it will work or not, and will turn off AP within a mile before TD when everything is set up for a proper landing.

Edited April 5, 2019 by Noel

[flyforever 267]

I think your idea of trying the trim wheel with another plane, a stable one like PMDG , might provide some  reference. point. I find it strange that a multiple turn trim wheel should be so rough.

If you can find a generic calibration software that allows you to clean up the behavior curve and, at the same time, create a profile just for the Dash, it might be a good solution.

FSUIPC registered does provide some curve tweaking....

Incidentally, when I built a trim wheel for ailerons and decided to use an axis, I could never level the wings--on any aircraft. Somehow, the joystick and the trim wheel axis competed for control. I finally replaced the pot in the trim wheel with two switches.

Tony

tony

[MarkDH 234]

5 hours ago, flyforever said:

Unless you can change the pot into a multiple turn pot, only software can fix it. If you take apart the unit, you can check to see if the pot is a 60 degree or a 270 degree pot.

It isn't a pot, it's an optical rotary encoder.

[flyforever 267]

thank you for the clarification, Mark.

tony

[Noel 4,487]

6 hours ago, MarkDH said:

It isn't a pot, it's an optical rotary encoder.

It sure feels nice and now actually works pretty good since I changed the slope.  I have the electric trim controlled w/ a rocker switch on the TQ.  I have the button slider set in the middle which assume means you get more than one increment per press, but I'm not sure.  Not sure how I can test for that but if it jumps up to say 4 increments per press then it's not too different from the trim wheel which actually seems to have pretty fine control.  It may have been a red herring from a dying yoke that was giving me trouble.

[LesOReilly 176]

I used Spad.Next to change it into a "increase" and "decrease" to the Trim.  Originally I was using the "axis" method as well, but found that after using an autopilot and then disconnecting the moment you trimmed with the wheel it would jump to the Axis position...Causing pain in the but.....

 

Once changed to the behaviour of increasing or decreasing this helped out a lot.....but only until you may have gotten to a point where you were now fully extended on the Axis then it would not be able to go in that direction......Would have been better if it was just a Rotary Encoder that increased and decreased...The Axis makes it sit in a bin 😞

[MarkDH 234]

1 hour ago, LesOReilly said:

Would have been better if it was just a Rotary Encoder that increased and decreased...The Axis makes it sit in a bin 😞

As I said, it is an encoder internally so you can get to this if you're prepared to do a bit of hacking.