It shouldn't unless the settings are so low that even air pressure would effect it, But I doubt that is the case. Show a screen of the controller force settings just for giggles.




Bullet

When this accurs I creep my hand up slowly and this still creates this anomally. Maybe it's a grounding issue? btw my force settings are at defaults. I would show you but I can't figure out how to input imgs.

I just happened to know that it was TV(CRT) that was causing X65F disconnecting/connecting problem.
it looks like the stick is affected by electromagnetic wave the TV generates when is turned on or off.

I notice my body proximity to the X-65F causes the rudder axis to change deflection slightly at 0.9 force setting (I like .47 X and .78 Y on the other axiis at F4 setting). The stick axis also jumps a little when I turn off a fluorescent desk lamp. That sounds like sensitivity of the rudder axis to infrared energy and an issue with RF/EMI shielding, respectively. If the very long USB cables had ferrite cylinders at various points along their length, it might help the stick's X/Y jumping with the lamp turning off. I'm not sure how the rudder deflecting from body proximity could be helped. It sounds like major physical design problems.

Another weird thing is that POV1, pinkie paddle, and button B (possibly others on the stick) cause the rudder to deflect, too. I've determined that's definitely not the result of added twist pressure, especially since button B's effect causes the exact same right-directional deflection of the twist axis: the opposite from pressure applied. The paddle is also symmetric to the twist, so it's not force causing this. That could possibly again be an RF/EMI shielding issue within the wiring or along the cable, but could also be sloppy internal circuits with actual leaking voltage between the paths. It could be a code bug, but it does it to varying degrees on all the drivers.

This is a shame, because I'm tweaking a really bad Falcon AF profile I found on the Saitek site (now it's useable) and am starting to get used to this hyper-responsive, no-delay stick. Oh, I do not like the reduced throw of the throttle, the tiny distance past the afterburner detent, or the excessive resistance to the throttle's motion. If the rudder and RF/EMI issues were not present, I would still be highly compelled to look past the throttle's issues, though. I don't like the rotaries' wiggle or the inability to program the mode switch to send a button command at the same time I switch a mode (like master modes, as I could do on the X52), but that too I could overlook.

I just tried putting my hand around the cable connecting the throttle to the stick. Near the stick end, with my hand nearing and wrapping the cable, the rudder deflects as I near. Definitely some shielding problems with that PS/2 cable. Too bad it's hardwired into the throttle end, because there must be some after-market heavily shielded cables out there, including ones with ferrites along the length. On a related note, I wonder if the use of balanced, double-ended cables with differential opamps (like in pro audio) mighthave helped transmission of signals between the stick and throttle.

By the way, anyone know if the unit is 14-bit resolution?

I currently have the cable between the throttle and stick (the short one we're talking about), around behind my body and between my back and the chair. I calibrated the stick with this arrangement of the cable and my hand actually on the stick and am now getting somewhat better centering and less deflection when hitting button B and the pinkie paddle.

Well. I really wish the consensus on the x65 was better. I know the CH products are way more reliable, but they've got no swag as far as I'm concerned. If there was better support I'd be more willing to throw down $350. But since it's Saitek I know better. So, I want a sexy hotas that is around $350, but there aren't any good solutions out there in that mid-price range. What's up with that?

I ordered some ferrite choke cores to try and get rid of the issues on the twist axis. We'll see how that goes. Aluminum foil has made no difference. Move my hand near the cable or the stick and the rudder deflects. Ditto with just moving the cable around. And the calibration is flakey. Doesn't seem to actually do anything much of the time. Unplugging/replugging the stick works better. Without the drivers installed, I wonder if the windows calibration would work. Hopefully it's RFI/EMI and not infrared that's causing it, as short of wrapping it in neoprene I don't know how you'd mitigate the latter.

The chokes definitely help. So I ordered another 20. I had difficulty getting the 5mm clamp-on TDK ones to fit. Broke one of the 10 and put probably too many on the other cables when I became frustrated getting them on the stick cable end. These are probably 6mm diameter cables. One large ferrite ring, with the stick-end connector of the cable wrapped around it a bunch of times then taped/covered with non-residue electrical or studio tape would do the job probably better and for 1/10th the price than spending $30 on the little ones. And no risk to the cables. It definitely has the desired effect, though. I'm just getting a little bit of interference remaining when moving the cable or getting my hand near the connector. The twist axis doesn't even deviate outside the deadzone anymore moving the whole cable around at my lowest force setting, which is way below the F4 default. Granted, I have a large deadzone set right now, but it used to deflect over 50% of the full axis just moving it. I'm a little worried these clamp-on ones, being just barely large enough to fit with pliers to squeeze them shut, will damage the PVC insulation on these USB and PS/2 cables. And I also doubt you need them anywhere except right were the cable plugs into the stick. But I will just line the hell out of that cable with them when the rest get here. For $100, obviously a pair of rudder pedals will fix this issue just the same. Saitek definitely screwed up, though. No reason not to have some medium sized choke cores, or at least one on the stick end.

I would've bought this buggy piece of kit if I hadn't been an avid SimHQ lurker. It's a big disappointment because x65 is placed to fill the medium to high priced HOTAS niche, it looks cool, and I would've bit. But I'd much rather find out about problems BEFORE I make the jump rather than afterwards. Thanks all for the discussions and the heads-up!

I don't know about buggy. The STT software and these pr0 format files seem less sophisticated to me than the original dat system for the X52. Saitek had all sorts of promises about how sophisticated the programming would become. Look at my original EECH dat profiles. Holy crap, dude. I'm proud about how they turned out. It's seriously aggravating as hell trying to get anything near that complexity on this new STT. Even the time quantization doesn't work the same for advanced commands. It seems like theoretically it's the exact same features on the STT without a single added one, yet it's more clunky and harder to figure out to get it to do what I want. Like, how do I have a button do one thing with just a quick press and do something else when I hold it for a half second? I've done a few that appear to be doing what I want like that, but it's not easy to achieve and involves tricking the STT to quantize a bunch of the same commands to different values and then pasting them around to where I need, deleting the keys with the wrong time. I just swap them around and keep trying until it does it. Weird. It almost seems like it's an attempt by another group of people to duplicate the original programming but by a bunch of coders that didn't quite know what they were doing. Did MadCatz not get the original dat STT code when they bought Saitek?

The stick and throttle itself are really not buggy hardware, as far as I can tell. USB, midi, and ps/2 cables need ferrite choke cores on them, especially when it comes to controllers at the 10 and 14bit resolutions. I don't know what the x-65f is, but it's clearly some very fine, low-intensity, highly intricate electrical signals going through. There almost needs to be balanced XLR connections going through these, with DSPs translating the signal back to what the computer can use. I wonder if this is why Logitech programmed in the "hysteresis" into the G940, as a sort of signal redundancy to prevent interference from altering the signal parameters. It would have added few bucks to the cables, tops. It's just those little cylinders you see on computer cables, is all. Until I moved from DJing on vinyl and CDs back in Vegas to buying my first computer DJ controller (10bit, then a 14bit one), I had no idea what those things were. They can be the difference between hell and perfect operation. In the case of the X-65f, one single large rubberized ferrite choke core for a couple dollars will likely fix the issue for users. Wrap the stick end cable around it a few times, plug it into the stick. Another single clam-on at the end of the one that goes into your computer would be a nice insurance just in case there is other interference I didn't notice in the rest of the cabling.

A word on the deadzones: they don't function like you'd normally think. They are just a pressure buffer. They reject a portion of the initial force before deflection starts to register. Since there's no stick movement at all, it's not like substantial deadzones are unnatural or show dead play in the middle of the stick's motion. It's just hand pressure that can occur -- like your palm just resting on the hand rest -- that doesn't do anything. So deadzones are absolutely necessary. Unless you've got them in the sim, there has to be something in the Saitek control panel set for this. As the interference has been reduced drastically with the choke cores, I can make these deadzones smaller. It's also important to point out that the deadzones do not scale in width as you change forces with the F1-F4 console panel box. A setting to change the width automatically might actually be nice (1.0:1.0 to 2.0:1.0 from F1 to F4), but as it is you just need to realize that while a very small deadzone is required at F1 (even when I have it reduced further, especially in the twist force), you'll need to be extra careful when you drop the force down on-the-fly. Conversely, if you start out at F4 or F3 with a more substantial deadzone and then later change to F1 while in-sim, you're rejecting a lot of initial pressure unnecessarily. So right now you either have to compromise or optimize the deadzone for the particular force you will mostly be using. And like I said, my deadzones are finally getting down to reasonable ones.

I'm still considering getting rudder pedals, but that's now because of axis pollution and unintended deflection of X & Y when trying to yaw in helo sims when I have hover hold turned on. To be fair, I believe even the RAH-66 had a pinkie button on the SAC related to yaw for the purpose of trimming and axis isolation. So it's not just endemic to force sensing on the X-65f. With the interference issues finally being obviously mitigated (the changes with the choke cores are not subtle), my issues with force sensing are general now to the technology as a whole. I don't think much of it for general flying purposes in fixed-wing aircraft. I doubt Airbus 300 series aircraft will ever move away from normal joysticks to force sensing. It does help with aiming the gun in aircraft that don't require constant stick pressure to maintain attitude, but just simple aerobatics and landings are more sloppy. Aiming is surreal under the aforementioned circumstance, but AFCS Trim Update helicopter flying is definitely where this pressure/force sensing thing comes into its own. While I can see why a little bit of deflection was added on the F-16 and RA-66 sticks to help with axis pollution, it really works wonderfully mimicking this AFCS chopper stuff regardless.

The perfect implementation would have adjustable throw AND pressure sensing for a helo attitude hold/ attitude command side stick. That way, when you go to a Wings Leveling flight law for landing, you could have a lot of throw, low sensitivity, and the stick would behave almost like a conventional joystick so that stick position reflect aircraft attitude. When you were in normal flight, it would have little or no throw. But that's asking a lot. Pressure sticks can do things normal joysticks cannot as-is, and vice versa. And I suppose the same is true with force feedback joysticks, as well. For the price I paid, I’d really like to try the G940 to have a better sense of what that actually means, but that apparently has some severe issues that require serious modding to remedy, and it does not appear that resolves all of them. Considering my interest in GlovePIE/PPJoy for this helo auto trimming stuff, I think I will keep the X-65f and simply eventually have my X52 sent to me from storage for situations where it is not appropriate: driving sims, Search & Rescue 4, etc. Hopefully, my JFA-18 traps improve. But I was fine with the X52 on that. Probably has something to do with the real flight laws of the Superbug when gear is down. Not identical to, say, an F-16’s in holding attitude.

I think I'm mostly done with modding someone's F4AF profile and created a single-mode EECH profile. The latter is still missing a few things like time acceleration I think, but most of what I want is on there. I haven't bothered trying to figure out pinkie-shift with the STT software. A S&R4 profile is about 2/3 the way finished, but I don't know if I'll get around to completing it with the trim button (num 0?) not functioning and the odd mention of turning the AFCS off without a command to actually do it. Seems like an unfinished stinker. I still need to do profiles at least for JFA-18, Flaming Cliffs 1, Comanche Gold, and WWII Fighters (easy).

Edit:

Two more days with JFA-18 and I'm starting to get the hang of traps with this. Maybe Janes does attitude hold to 1G with the gear down? The key is no deadzones within Janes. Deadzones reduced quite a lot in the Saitek panel. They're minimal, now. One notch of nonlinearity in Saitek panel for X, Y, and twist (I use it for every sim, probably not necessary with Janes). The ferrite choke cores. And lots of practice. I also have the stick tilted back with something under the front of it a little over an inch tall (Boxer screw/hex set red/black) that forces it to center offset so hand pressure doesn't cause as much back Y without me trying. The very gentle internal axis curve of JFA-18 works quite well as-is with pressure/force sensing and this configuration.

http://img195.imageshack.us/img195/3718/x65fdeadzones.jpg

Here's what I settled on for forces a week ago:

Kgf & N m

F1 (great for Time Accel)
1.88
2.50
2.10

F2
1.25
1.88
1.80

F3 (just trapped fine three times in JFA-18)
0.78
1.25
1.40

F4
0.47
0.78
0.90

Edit: Made the axiis totally linear. Found I didn't have enough authority at low pressures in WWII Fighters. A reset button for those would be nice.