With Airbus FBW system there is no direct link of sidestick and flight control surface movements, everything goes through SEC, ELAC and FAC computers. Computers will actually in regard to flight envelope protection, move flight controls according to pilots command *. So max stick deviation does not always mean max flight control movement. Well actually in Direct Law position of sidestick will correspond to the position of control surfaces. Anyway this thing is quite complicated and am not quite sure how is it simulated in home simulator software?
I have encountered a problem regarding that stick deviation and flight control movements. I mentioned that already in pots instead of transducers at the end of the post. I noticed sort of insufficient stick sensitivity using my linear pots. The stick needs to be pulled to hard and to far from neutral position on takeoff or landing flare. In my opinion much more than with real stick. I remember from Level D sim visit on takeoff I didn’t have to pull stick that much and in general small inputs where needed.
I guess it was a matter of time before I tried the original transducers. I couldn’t help it, I just had to try it even though I had a hunch it won’t be any different! Each axis transducer contains five pots for ELAC and SEC signals and they are all doubled (command and monitor unit). After finding some of potentiometer pin-outs I got the output signal I wasn’t expecting. I’ve gone from linear pot insufficient stick sensitivity to a transducer with to sensitive stick!
I wasn’t really happy with this output signals as stick was extremely sensitive around center and decreases with larger deviations. It looked like this output signals are somewhat logarithmic. It was hard to fly approaches with stick so sensitive around center. Just the opposite from my linear pots solution, lacking sensitivity around and with center! Of course there is also a good probability that I found the “wrong” outputs? I also tested joystick curve software and tried to improve and tweak response curve. It was a little better but still not the way I wanted. I was interested in getting some more information regarding proper stick sensitivity. I guess stick sensitivity in real life changes according to flight phase and envelope?
I recently had a productive debate with a real Airbus captain and that logarithmic output would actually make some sense to him regarding the behavior of the real aircraft. If we talk about normal operation, sidesticks are only used in this relatively small circle around neutral position and in that range the sidestick must be sensitive. Sensitivity should then proportionally decrease nearing to the maximum stick deviation, similar to output signal I got.
We came to a conclusion that perhaps the original sidesticks might be made to work with logarithmic output that aircraft computers know how to interpret. If you send this same logarithmic output signal to a A320 simulator software, made for linear pot joysticks, you get to sensitive sidestick. Vice versa if you use linear output pots that simulator software know how to interpret, on the original sidestick made for logarithmic transducer, you get to small sensitivity in sidestick!?
I guess I need to some more research on this matter or test for some other output signals from ten transducer pots. Or just wait for simulator software to have an option for logarithmic input or some sort of sensitivity curve tool.
I think this response curve should be incorporated in simulator software. Maybe stick sensitivity can be controlled regarding flight envelope by software tweaking response curve. Like I suppose ELAC and SEC do in real airbus.
* By moving the stick left or right you command a roll rate (degrees per second), with greater stick deviation rate will increase (15° / Sec max.). Moving the stick back or forth you command pitch with g load factor (ratio between lift and weight of aircraft). G load is proportional to stick deflection (full back stick max g-load is 2.5 g and full forward -1.0 g).
7 thoughts on “Sidestick sensitivity?”
I recently aquired the Captain Sidestick Unit from MSN488. It came “as removed”, including the transducers. Before, I read your post here about the non-linear behaviour and was already fearing that I would have to find alternative pots, but then I measured the output of the transducers and was positively surprised that it is totally linear. So I now will use the original transducers (I’m still waiting for the BU0836A to connect it to the Sim).
I was wondering about your sticks and where they come from? My transducers come from Vishay Sfernice and are labelled “116SF3BT103W1119 G”. When I googled them, I also noticed there are “116SFZ”-Versions which indeed deliver a sinus output. Maybe your sticks are equipped with these? Has there been a MOD? My stick was manufactured in Okt. ’93.
Thank you for your comment, this is really interesting. My transducers are Sarma FO222 P/N:321000 dating Okt. ’93 also. I am not exactly sure if Vishay/Sfernice is refering to complete transducer or only pots used inside? Did you measured the outputs directly on pot pins or at the end cable pins? I still didn’t check out other possible transducer pot outputs since I use that joystick curve software implementation. I also did not open the transducer yet. I guess I should take time and explore it some more…
Oh, really? You didn’t remove the covers? That was the first thing I did… 🙂
The whole unit is labelled “SARMA F0222” with P/N.321000 MO, but I referred to the labelling on the transducer itself. Check out the data sheet here: http://www.vishay.com/sensors/list/product-54004/
They have 3 signal outputs, like beeing 3 pots in 1. The lower ones on each side have 3 outputs in use, the upper ones only 2 (2 ELACs, 3 SECs). One side, primary, the other for monitoring.
I measured the output directly on the pins, using an oscilliscope. The resistance is clearly directly proportional to the rotation angle. I now soldered wires directly to the pins of one transducer, drilled a hole into the plastic cover and reattached it. I cut off all the other cables coming out of the units.
Yes, it’s funny but I didn’t remove the covers, yet! That used to be something I would always do until I accidentally messed up one of the spoiler pots. Now my first attention are pins. I am guessing I have the same pots inside. It’s true that I didn’t spend much time on sidestick cable pins but it’s also possible that I am using wrong output pins? In schematic manual there are some resistors and back loops drawn. I think I will also try your direct pot output approach and see what I get.
I was thinking, maybe some of that linearity might get lost on push rods to rotational motion linkage?
I was wondering if maybe there’s some electronic stuff hiding between the transducers and the output pins that changes the signal from linear to non-linear.
Concerning linearity getting lost: For the Roll Axis, the ±20° of the stick motion are converted to 30° of transducer lever axis rotation (which is then approxiamtely doubled through the internal gear wheels). I think from a mechanical point of view, for the first 30° of the sinus function, the deviation from linearity should be negligible (<5%).
Finally I removed the transducer covers and tried direct pot output approach. Guess what, I found a linear pot output! I also traced the corresponding pins on the cable cannon plugs. I guess this topic needs additional post and a proper pin-out diagram. Thank you again Jürgen for sharing your transducer pot conclusion.
Oh, wow, now that’s some news! What brand are the transducers?
I’m curious if you find out what happens between the pots and the pins.
Comments are closed.