Throttle calibration question

[petdocvmd]

I'm building and testing a throttle quad for the 737-700. I've used 100k linear taper slide pots and have a question about calibration.

For now the pots are set up as part of a generic 2-button 3-axis joystick in which the 3rd axis is assigned to flaps via a "ghost" Engine 4 mixture axis.

I calibrated in Windows as best as possible considering throttles are using x/y axes. I set sensitivity to max and null to 0 in FS2k2.

On the FSUIPC joystick page for throttles I went through the steps to set and calibrate the throttles but here's the thing: They each work fine but don't line up very well with each other at the same N1 except at idle or max.

I think I could improve this by physically detaching the pot for 1 lever and moving it back a bit. However this involves pulling some nails, etc - am I missing a way to do this in the software?

Thanks for your advice,

Scott

[Pete Dowson]

On the FSUIPC joystick page for throttles I went through the steps to set and calibrate the throttles but here's the thing: They each work fine but don't line up very well with each other at the same N1 except at idle or max.

I think I could improve this by physically detaching the pot for 1 lever and moving it back a bit. However this involves pulling some nails, etc - am I missing a way to do this in the software?

No. Unless you can get pots which are really very precisely matched I don't think you can ever get them perfectly lined up. (Actually, I don't really think the real throttle levers in the aircraft are necessarily precisely aligned to get the exact same performance from the engines -- at least they don't look that way when I've been in the jump seat, especially with Auto-Throttle engaged, which it invariably is after take-off).

You can use the FSUIPC throttle sync facility to get them all doing the same thing -- that 'cheats' by copying the throttle 1 values to all the others, overriding the separate inputs. That's okay when you don't want differential thrust and all engines are working okay.

To get values the same for each physical position of the throttles, each one would need many calibration points, with a table to convert the input value to the desired output. This is actually a facility provided on EPIC cards -- each input axis is processed through a 256 byte table (the raw inputs are in the range 0-255, so this suits). I could consider something like that for a future version of FSUIPC -- I have been considering adding full axis assignment and calibration facilities in any case (by-passing the FS CFG stuff), and would certainly add this then. But I cannot consider any such addition yet -- not till well after "conquering" FS2004!

Regards,

Pete

[petdocvmd]

No. Unless you can get pots which are really very precisely matched I don't think you can ever get them perfectly lined up. (Actually, I don't really think the real throttle levers in the aircraft are necessarily precisely aligned to get the exact same performance from the engines -- at least they don't look that way when I've been in the jump seat, especially with Auto-Throttle engaged, which it invariably is after take-off).

Understood. And no, I don't want perfect match - just fairly close.

To get values the same for each physical position of the throttles, each one would need many calibration points, with a table to convert the input value to the desired output. This is actually a facility provided on EPIC cards -- each input axis is processed through a 256 byte table (the raw inputs are in the range 0-255, so this suits). I could consider something like that for a future version of FSUIPC -- I have been considering adding full axis assignment and calibration facilities in any case (by-passing the FS CFG stuff), and would certainly add this then. But I cannot consider any such addition yet -- not till well after "conquering" FS2004!

Hmm - is this possibly doable by mere mortals using the fsuipc sdk? If I sample the pot input via Direct Input I could run it through such a table in my app. I don't have the docs in front of me - can this "digitized" data be output to fs2k2 via an axis or offset(s) in fsuipc?

Thanks,

Scott (hoping that fs2004 won't bring radical programming changes )

[Pete Dowson]

Hmm - is this possibly doable by mere mortals using the fsuipc sdk? If I sample the pot input via Direct Input I could run it through such a table in my app. I don't have the docs in front of me - can this "digitized" data be output to fs2k2 via an axis or offset(s) in fsuipc?

Yes. In fact you don't even need to do the direct input bit yourself unless you really want to. (Actually I wouldn't use DX in any case, the ordinary joyGetPosEX calls seem perfectly adequate to me!).

If you check the Programmers Guide for FSUIPC you will see that at offsets 3328-3337 you can read the axis input values -- those read by FS itself, and at 310A you'll find switches to allow these to be disconnected from FS. Hence you can read the values yourself, manipulate them how you like, then write the results to the Throttle control areas at 088C, 0924 et cetera. This is the method you'd use for "fly-by-wire", for example, on an Airbus, where the throttle position only represents a desired "mode" not a specific throttle setting. Take care to keep replenishing the switches at 310A as, for safety reasons, FSUIPC will cancel the disconnection every so often otherwise.

Regards,

Pete

[petdocvmd]

If you check the Programmers Guide for FSUIPC you will see that at offsets 3328-3337 you can read the axis input values -- those read by FS itself, and at 310A you'll find switches to allow these to be disconnected from FS. Hence you can read the values yourself, manipulate them how you like, then write the results to the Throttle control areas at 088C, 0924 et cetera.

Great Pete, thanks! Question: Is the range of values read at (e.g.) 3330 (Throttle 1 axis) the same as that expected at 088C (i.e. -4096 to 16384) assuming calibration done? Also, does FS interpret any part of the range at 088C as reverse thrust or is -4096 idle and 16384 max?

Thanks,

Scott

[Pete Dowson]

Is the range of values read at (e.g.) 3330 (Throttle 1 axis) the same as that expected at 088C (i.e. -4096 to 16384) assuming calibration done? Also, does FS interpret any part of the range at 088C as reverse thrust or is -4096 idle and 16384 max?

Actually the range at 088C et cetera can be -16383 to +16384. All -ve values are reverse thrust. The -4096 I show in the document was just the 'normal limit' for jet reverse as I saw it when I wrote it. In fact the maximum reverse thrust is set in the AIR file and can be more or less than that. If you have an AIR file with the max reverse set to -16383 then you can go just as fast backwards as forwards!

You can read the maximum reverse thrust in 333A. FSUIPC now calibrates reverse to that limit, and this actually changes according to the aircraft you load.

That given, I think you should be able to simply copy the value you read in 3330 into 088C, with any manipulation you want in between.

Why don't you just look at these locations using FSInterrogate? I think you'd then get a feel for what is going on. It'll all become clear .

Regards,

Pete