PFC.DLL problem troubleshooting

[dfournie]

I'm troubleshooting a flight control problem I've seen develop over the last two weeks in my setup. Occaisonally, I experience FS2004 not accepting any inputs from my PFC flight controls. Whatever manuever I'm in (climb, turn, etc) continues without letting me intervene. FS keeps flying OK, but ignores flight control input. After a few seconds FS2004 pauses for just a moment, then allows normal flight control input. I haven't installed anything new. Before I blow away the whole installation and start again, I was wondering if anything like this has been seen?

[Frank.O]

Hi,

what flight controls are you talking about?

A console with yoke and throttle quad and pedals attached to it?

Or a combination of yoke and pedals? If the latter, are they USB?

It may be a different matter entirely but I remember one or two reports from the german FS newsgroup where simmers had trouble with USB flight controls not responding after having been left untouched for a short while (autopilot engaged). It was caused by the computers power management which turned of the USB hub after a while of inactivity. Disabling the power management solved it.

Regards,

Frank

[Pete Dowson]

...I experience FS2004 not accepting any inputs from my PFC flight controls.After a few seconds FS2004 pauses for just a moment, then allows normal flight control input.

It does sound like power management. Are the PFC controls connected direct to a PC COM port, or to USB via an adapter? I found the latter to be much more reliable -- Windows support for "obsolete" ports like COM and LPT seems to be deteriorating -- but you do need to make sure power management is turned off.

I have made some improvements in the serial port handling in PFC.DLL, but there are also large changes for other things and I've not yet finished those nor documentated them. However, you may like to try the attached version 1.912 to see if it helps.

Regards,

Pete

PFCDLL1912test.zip

[dfournie]

Thanks Pete, I'll try it. I have the PFC Beech Yoke, TurboProp power quadrant into the digital I/O Box that connects to the serial port. I've tried it with both a built-in Serial port and a USB to Serial connector. I've seen it happen with both. I'll double check to make sure Power mgmt is off.

[dfournie]

Pete,

It made a fairly significant difference. Using the test pfc.dll and a different USB to Serial converter, the controls seem to be more responsive and appear to have a better digital filtering algorithm. The display stuttering is also gone. I turned off all Power Mgmt in BIOS.

[Pete Dowson]

It made a fairly significant difference. Using the test pfc.dll and a different USB to Serial converter, the controls seem to be more responsive and appear to have a better digital filtering algorithm. The display stuttering is also gone. I turned off all Power Mgmt in BIOS.

Hmm. I can't claim the credit for much if any of that I'm afraid. The changes in serial port handling I mentioned were really only a matter of poll timing adjustments. Most of the other changes in this version are for the new PFC 737NG cockpit, which is why it isn't released yet, still under test! :wink:

Maybe the USB to Serial adapter came with a better driver? I seem to recall that I had a driver once which would stop working after a while, same as you first reported -- they all appear to be FTDI made, I think, but the one which came with my first adapter was well out of date.

Regards,

Pete

[dfournie]

I did actually see a better filtering in the flight controls page. The reason I say this is because of a 'dead spot' in the pitch axis potentiometer of the PFC yoke. It was a separate problem I had been seeing. It was getting rather annoying. I ran a Volt/Ohm Meter across the output pins of the yokes DB-15 connector to verify I wasn't going mad. The dead spot appears to be handled much better (ignored) now for some reason. It works good now so I'm not going to touch it.

[nmthomas]

:D Hi Pete,

Trying your new test fsuipc and pfc. (I use the unreg. version of fsuipc)

So far they perform great, the pfc seems smoother and the only items I noticed were if you tell it not to perform checks it still does them and when you tell it to continue it releases the parking brakes.

I have a cirrus yoke and throttle quad & rudder pedals that are serial port connected. Since my knowledge is very limited about this usb connecting am I correct that what I have can be run through this type of connector and possibly increase my performance standard ??

Excellant work as always Pete, thank you.

Regards, Nick :D :D

[Pete Dowson]

... the only items I noticed were if you tell it not to perform checks it still does them and when you tell it to continue it releases the parking brakes.

Can you be more specific, please? How are you trying to turn them off, and what console are you telling PFC.DLL you are using (I don't mean what you are actually using, just what you've told it).

There's nothing turning any brakes on or off in any of the checking code. It merely checks inputs, there's not one output call in the whole thing. It sounds like you have so button operating that. Please clarify.

I have a cirrus yoke and throttle quad & rudder pedals that are serial port connected. Since my knowledge is very limited about this usb connecting am I correct that what I have can be run through this type of connector and possibly increase my performance standard ??

Well, I'm not sure. But I've found USB serial port adapters, providing you get a recent driver which doesn't stall like my first one used to, more reliable and generally faster (more "responsive", rather). After all, although the simulated serial port they provide is limited to the set speed, the USB channel itself is much faster. It also is not reliant on the sort of interrupt handling that the real COM ports are -- I tend to think the IRQ handling side of the serial port drivers is one of the problem areas.

On top of that MS and the hardware folks have been concentrating on USB (and firewire) as the connecting method of today and the future, whilst the poor old COM and LPT ports are left to fend for themselves, poorly.

Regards,

Pete

[nmthomas]

Pete, I figured out my problem with the parking brake.

The front page of the pfc.dll that one can bring up with fs9 running has 2 items center the page and turn off the startup checks. I assumed if you uncheck the box the checks would cease.

Nick :oops:

[Pete Dowson]

The front page of the pfc.dll that one can bring up with fs9 running has 2 items center the page and turn off the startup checks. I assumed if you uncheck the box the checks would cease.

Yes, they should do. But I did lock them for some tests. Maybe I forgot to release that lock, I'll check that. Thanks.

Regards,

Pete

[Pete Dowson]

The front page of the pfc.dll that one can bring up with fs9 running has 2 items center the page and turn off the startup checks. I assumed if you uncheck the box the checks would cease.

Here's a version (1.913) with the connection checking unlocked. You will still get them if one of these conditions is true:

1. No COM port specified or it cannot be opened (e.g. something else is using it)

2. The version of FSUIPC installed is earlier than 3.45

3. You have selected the 737NG cockpit as your 'console'.

Good flying!

Pete

PFCDLL1913test.zip

[nmthomas]

:D :D :D Thank you Pete works fine and I will look into a usb comport setup.

:D Nick

[MichaelMcE]

I have made some improvements in the serial port handling in PFC.DLL, but there are also large changes for other things and I've not yet finished those nor documentated them. However, you may like to try the attached version 1.912 to see if it helps.

Regards,

Pete

Pete,

I stumbled across this thread and tried out the 1.912 and 1.913 versions of PFC.dll. The strange thing is that with either of these builds I am unable to perform flight control axis calibration - the calibration screen is non-responsive to flight control inputs. I reverted back to the current 1.90 version and the ability to calibrate works just fine.

-michael

[Pete Dowson]

I stumbled across this thread and tried out the 1.912 and 1.913 versions of PFC.dll. The strange thing is that with either of these builds I am unable to perform flight control axis calibration - the calibration screen is non-responsive to flight control inputs. I reverted back to the current 1.90 version and the ability to calibrate works just fine.

Odd. That's not been changed. What about the throttles?

I'm not sure how I can help at present. What operating system FS version, other things running, and so on? I really wouldn't know where to begin.

I'll be trying to make another interim release next week, but then that's it till late April.

Pete

[MichaelMcE]

I stumbled across this thread and tried out the 1.912 and 1.913 versions of PFC.dll. The strange thing is that with either of these builds I am unable to perform flight control axis calibration - the calibration screen is non-responsive to flight control inputs. I reverted back to the current 1.90 version and the ability to calibrate works just fine.

Odd. That's not been changed. What about the throttles?

I'm not sure how I can help at present. What operating system FS version, other things running, and so on? I really wouldn't know where to begin.

I'll be trying to make another interim release next week, but then that's it till late April.

Pete

Pete,

Oh no - do not spend time on this as it's not a show stopper since all my controls are already calibrated! :wink: I just wanted to let you know that I did detect this and perhaps others that have downloaded it can check to see if they get the same results.

The throttle quad (throttles, speed brake, reverse thrust) show no response either. Yet, when I exit the display and go back into FS, everything works just fine.

I'm running XP Pro, FS9.1, FSUIPC, Autosave, WideFS, Aerosoft MCP747, Ultimate Traffic.

-michael

[Pete Dowson]

Oh no - do not spend time on this as it's not a show stopper since all my controls are already calibrated! :wink: I just wanted to let you know that I did detect this and perhaps others that have downloaded it can check to see if they get the same results.

I'll need to fix it before I make a general release, some time in April maybe. I'll have a quick look next week when I add a few other changes I've been promising. Even if I can't fix it then maybe I can add some diagnostics to try to track it down.

I'm running XP Pro, FS9.1, FSUIPC, Autosave, WideFS, Aerosoft MCP747, Ultimate Traffic.

Nothing there any different to what I'm running. Could you send me your PFC.INI file just in case it's some combination of options? petedowson@btconnect.com.

Thanks,

Pete

[nmthomas]

Hi Pete,

Thought I would check back after reading the threads and I can tell you that using 3.469 and 1.913 with a rewired currus yoke and throttle quadrant rudder pedals all from pfc also using a 2 engined jet controller that it all works great.

System is XP 3.0 and 1 gig ram. I have reset all the controls using pfc 1.913 and have no problems. To top it off I unchecked the suppress for possible interference from game ports on the first fsuipc page and the autothrottle works like advertised. No problems at all.

When you finalize I might stay with these. Thank you for a great piece of work. Have a good holiday riding trains.

Regards, Nick :D :D

[Pete Dowson]

To top it off I unchecked the suppress for possible interference from game ports on the first fsuipc page and the autothrottle works like advertised. No problems at all.

Your throttles are behaving well then. I never really could understand why some (like mine) give no measured jitter and others wobble all the time. I'm sure there cannot be such variations in manufactured quality. I bet it's down to environment -- dirt, humidity, temperature and quality of power supply.

When you finalize I might stay with these.

Well FSUIPC 3.47, released tomorrow, is just a little different from 3.469 -- a fix for FS98-style visibility setting in FS2004 (still used at least by Squawkbox 2.3), added controls to disconnect throttles in case we can't resolve the issue on the 767.

Changes planned for PFC.DLL are mainly in answer to additional requests, like having throttle equalisation for three and four-engined aircraft.

Regards,

Pete

[nmthomas]

Pete,

I forgot to say that the programmed autopilot disconnect will not disconnect the AP it plays the sound but no disconnect this is on the LevelD767 works fine on other aircraft.

I don't have a registered version of fsuipc so I cannot reprogram the keys to the autopilot but I can get by with the Z key or the mouse.

Have a good day.

Nick

[Pete Dowson]

I don't have a registered version of fsuipc so I cannot reprogram the keys to the autopilot but I can get by with the Z key or the mouse.

The Z key operates the Level D767 A/P??

Pete

[MichaelMcE]

I don't have a registered version of fsuipc so I cannot reprogram the keys to the autopilot but I can get by with the Z key or the mouse.

The Z key operates the Level D767 A/P??

Pete

Yes, it does. Let me explain. The default LVLD keymaps have the Z key as the A/P ON/OFF toggle. There are three other keys that are mapped to CMD L, CMD C, CMD R (Left, Center, Right A/Ps).

What I did was assign the PFC A/P Button to Keypress 'Z'. The result is this:

(a) If the LVLD A/P is OFF: then the A/P is turned ON and the Center A/P is ENGAGED. I do not know why the Center a/p is engaged - I would have thought if anything that the Left would have been. (??)

(b) If the LVLD A/P is ON and any one of the three A/Ps are ENGAGED: then the A/P is turned OFF and consequently the ENABLED A/P is disabled. At this point, an A/P disconnect warning is displayed and remains displayed. A second press of the A/P button is necessary to clear the warning. This puts the A/P state back in the state described in (a).

Note: LVLD support mentioned that in programming a button to toggle the a/p we would need to press it twice to completely disengage and turn off the a/p. LVLD support also mentioned a A/P 'soft' disconnect custom control. By default it is not mapped to any key. I did not have the time to try that out to see what functionality it provides. If it is purely a disconnect function that I would prefer staying with the current settings since they allow me to turn ON and ENGAGE as well as Disengage/turn OFF the A/P.

-michael

[MichaelMcE]

I'm running XP Pro, FS9.1, FSUIPC, Autosave, WideFS, Aerosoft MCP747, Ultimate Traffic.

Nothing there any different to what I'm running. Could you send me your PFC.INI file just in case it's some combination of options? petedowson@btconnect.com.

Thanks,

Pete

Sure. It will be very late tonight before I get home to do so.

-michael